UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Paramagnetic organometallic complexes of chromium, iron and cobalt Leznoff, Daniel Bernard 1997

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

Item Metadata

Download

Media
831-ubc_1997-250938.pdf [ 21.73MB ]
Metadata
JSON: 831-1.0059492.json
JSON-LD: 831-1.0059492-ld.json
RDF/XML (Pretty): 831-1.0059492-rdf.xml
RDF/JSON: 831-1.0059492-rdf.json
Turtle: 831-1.0059492-turtle.txt
N-Triples: 831-1.0059492-rdf-ntriples.txt
Original Record: 831-1.0059492-source.json
Full Text
831-1.0059492-fulltext.txt
Citation
831-1.0059492.ris

Full Text

PARAMAGNETIC ORGANOMETALLIC COMPLEXES OF CHROMIUM, IRON AND COBALT by DANIEL BERNARD LEZNOFF B.Sc. (Hon.), York University, 1992. A thesis submitted in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMISTRY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA May 1997 © Daniel Bernard Leznoff, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of e^fflivrW The University of British Columbia Vancouver, Canada Date DE-6 (2/88) ABSTRACT The chemistry of the amidodiphosphine ligand "N(SiMe2CH2PPh2)2 with chromium, iron and cobalt metal centres in the +2 and +3 oxidation states is explored. Metal-chloride complexes MCl[N(SiMe2CH2PPh2>2] are high-spin and tetrahedral for iron(II) and cobalt(II) and dinuclear with bridging chlorides for chromium(II). Alkyla t ion reactions give high-spin, square-planar CrR[N(SiMe2CH2PPh2)2] (R = M e , C H 2 P h , SiMes 2H); the benzyl fragment i n Cr(r] 2-CH2Ph)[N(SiMe2CH2PPh2)2 ] is r i 2 -bound . The analogous cobalt(II) a l k y l complexes, CoR[N(SiMe2CH.2PPh2)2] (R = Me, CH2Ph, CH2SiMe3) , are square-planar and low-spin; the benzyl fragment is r i 1 -bound . CrMe[N(SiMe2CH2PPh2)2] reacts wi th H2 to give {[(Ph2PCH2SiMe2)2N]Cr)2(p-H)2, which is isostructural with the bridging chloride. Variable temperature magnetic susceptibility measurements give J = -12.4 cm" 1 for the chloride and J = -139 c m - 1 for the hydride. Reaction of CrMe[N(SiMe2CH2PPh2)2l with PhSSPh yields a rare five-coordinate chromium(III) complex CrMe(SPh)[N(SiMe2CH2PPh2)2]. The one-electron oxidation reaction of CrR[N(SiMe 2CH 2PPh 2)2] with alkyl halides gives Cr(R)X[N(SiMe2CH2PPh2)2]. Bu lky alkyl g roups s t a b i l i z e c h r o m i u m ( I I I ) d i a l k y l c o m p l e x e s , an e x a m p l e b e i n g Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2]. Reac t ion o f CoX[N(SiMe2CH2PPh2)2] with PI1CH2X gives five-coordinate CoX2[N(SiMe2CH.2PPh2)2] complexes. Solid state variable temperature magnetic susceptibility measurements confirm the presence of an S = 1 cobalt(III) centre with a zero field splitting parameter of D = 32.6 cm" 1 . The S = 3/2 cobalt(II) complex CoI[N(SiMe2CH2PPh2)2] exhibits zero-field splitting with D = 20 cm" 1 . Alkyl complexes of cobalt(III) can not be prepared; Co(R)X[N(SiMe2CH2PPh2)2] decomposes via Co(III)-C bond homolytic cleavage, loss of alkyl r ad ica l and format ion of CoX[N(SiMe2C H.2PPh2)2 ]. Addition o f Mel to CoX[N(SiMe2CH2PPh2)2] g ives CoIX[N(SiMe2CH2PPh2)-(SiMe2CH2PPh2CH2Ph)], a i i zwitterionic complex, by intermolecular reaction of RX with fluxional phosphine arms. Conversely, Mel reacts with CoMe[N(SiMe2CH2PPh2)2] to give CoI[N(SiMe2CH2PPh2)2], likely via Co(Me)I[N(SiMe2CH2PPh2)2]. Reaction of CoMe[N(SiMe 2CH 2PPh 2) 2] with P h C H 2 X gives a mixture of CoX2[N(SiMe 2 CH 2 PPh 2 ) 2 ] and CoX2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)]. A mechanism to account for intramolecular methyl-transfer, based on methyl-phosphoranyl radical formation from an organometallic radical cage, is presented. Alkylation of the iron complex FeCl[N(SiMe2CH2PPh2)2] gives low-spin Fe(rt5-C5H5)[N(SiMe2CH2PPh2)2], isostructural with the analogous low-spin chromium(II) complex, Cr(ri5-C5H5)[N(SiMe2CH2PPh2)2] and high-spin, tetrahedral FeR[N(SiMe 2CH 2PPh2)2] (R = CH 2 SiMe3 , CH(SiMe3)2). The reaction of iron(II) complexes with alkyl halides yields a mixture of zwitterions as well as FeX2[N(SiMe2CH2PPh2)2]; the iron(III) complex can be prepared independently from F e X 3 . FeBr2[N(SiMe2CH2PPh2)2] is trigonal-bipyramidal and is an example of a spin-admixed S = 3/2, 5/2 iron(III) system. • Anti - Co[PhP(CH2SiMe2NSiMe2CH2)2PPh] (Co[P2N 2]) is low-spin and square-planar; reaction with I 2 gives five-coordinate anti - CoI[P2N2]. Syn- and anti - Co[P2N 2] react with 0 2 to yield cobalt phosphine-oxide complexes. iii T A B L E OF CONTENTS ABSTRACT ii T A B L E OF CONTENTS iv LIST OF TABLES xvi LIST OF FIGURES xix GLOSSARY OF TERMS xxiv A C K N O W L E D G E M E N T S xxx DEDICATION xxxi QUOTATION xxxii Chapter 1: PARAMAGNETIC ORGANOMETALLIC COMPLEXES - A GENERAL INTRODUCTION 1 1.1 Coordination vs. Organometallic Chemistry 1 1.2 History of Paramagnetic Organometallic Research 4 (i) Stable 17- and 19-electron systems 4 (ii) Unstable 17-electron metal-centred radicals 5 (iii) Compounds containing multiple unpaired electrons 5 (iv) Homoleptic Paramagnetic Organometallic Compounds 6 1.3 Magnetism: A Brief Tutorial 6 1.4 Characterization of Paramagnetic Organometallic Compounds 16 (i) Nuclear magnetic resonance spectroscopy 17 (ii) Electron spin resonance spectroscopy 19 (iii) Other techniques 23 (iv) Magnetic susceptibility measurements 24 (v) X-ray crystallography 25 1.5 Chemistry of the Amidodiphosphine Ligand System: Scope of the Thesis 26 iv 1.6 References 28 Chapter 2: CHROMIUM(II) AND COBALT(II) HALIDE AND A L K Y L COMPLEXES 33 2.1 Introduction 33 2.2 Synthesis, Structure and Reactivity of Chromium(II) and Cobalt(II) Halide Complexes 35 (i) Synthesis and structure of {[(Ph2PCH2SiMe2)2N]Cr} 2(u.-Cl)2 35 (ii) Reactivity of {[(Ph2PCH2SiMe2)2N]Cr}2(JI-C1)2 (1) with donor ligands 38 (iii) Synthesis and structure of CoX[N(SiMe2CH2PPh2)2] (X = C1, Br, I) 40 (iv) Reactivity of CoCl[N(SiMe2CH2PPh2)2] (2) with donor ligands 44 2.3 Synthesis and Characterization of Chromium(II) and Cobalt(II) Alkyl Complexes 44 (i) Synthesis and structure of CrMe[N(SiMe2CH2PPh2)2] (5) 44 (ii) Synthesis and structure of CoMe[N(SiMe2CH2PPh2)2] (6) 47 (iii) Paramagnetic lK NMR spectrum of CoMe[N(SiMe2CH2PPh2)2] (6) 50 (iv) Synthesis and structure of M(CH2Ph)[N(SiMe2CH2PPh2)2] (M = Cr (7), Co (8)) 52 (v) Synthesis of M(CH2SiMe3)[N(SiMe2CH2PPh2)2] (M = Cr (9), Co (10)) 57 (vi) Synthesis of M(Ti5-C5H5)[N(SiMe2CH2PPh2)2] and structure of the Cr(II) derivative 58 (vii) Paramagnetic A H NMR spectra of M(C5H 5)[N(SiMe2CH 2PPh2)2] complexes 62 (viii) Electronic structure and magnetism of CrR[N(SiMe2CH2PPh2)2] 63 2.4 Synthesis and Structure of Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13) 67 2.5 Paramagnetic Hydrides of Chromium(II) and Cobalt(II) 73 (i) Synthesis and structure of {[(Ph2PCH2SiMe2)2N]Cr}2(p-H)2 73 (ii) Variable temperature magnetic susceptibility of {[(Ph2PCH2SiMe2)2N]Cr }2(p-X)2 76 (iii) Synthesis and decomposition of Cr (CH2CH 3 ) [N(SiMe 2 CH2PPh2)2] (15) 82 (iv) Attempted synthesis of a cobalt(II) hydride complex 83 2.6 Reactivity of Chromium(II) and Cobalt(II) Methyl Complexes with Small Molecules 84 (i) Ethylene 84 (ii) Carbon Monoxide 85 (iii) Other substrates 86 2.7 Summary and Conclusions 87 2.8 Experimental 88 2.8.1 General Procedures 88 2.8.2 Materials , 89 2.8.3 Molecular Orbital Calculations 89 2.8.4 Synthesis and Reactivity of Complexes 90 (i) {[(Ph 2PCH2SiMe 2)2N]Cr}2(p-Cl)2 (1) 90 (ii) Cr(py)Cl[N(SiMe 2CH 2PPh2)2] U«py) 90 (iii) CoCl[N(SiMe2CH 2 PPh 2 ) 2 ] (2) 91 vi (iv) CoBr[N(SiMe2CH2PPh2)2] (3) 91 (v) CrMe[N(SiMe2CH2PPh2)2] (5) 91 (vi) CoMe[N(SiMe2CH2PPh2)2] (6) 92 (vii) Cr(ri2-CH2Ph)[N(SiMe2CH2PPh2)2] (7) 92 (viii) Co(CH2Ph)[N(SiMe2CH2PPh2)2] (8) 93 (ix) Cr(CH2SiMe3)[N(SiMe2CH2PPh2)2] (9) 93 (x) Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] (10) 94 (xi) Cr(ri5-C5H5)[N(SiMe2CH2PPh2)2] (11) 94 (xii) Co(C 5 H 5 )[N(SiMe 2 CH 2 PPh 2 ) 2 ] (12) 95 (xiii) Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13) 95 (xiv) {[(Ph 2PCH 2SiMe 2) 2N]Cr} 2(u:-H) 2 (14) 96 (xv) Attempted Synthesis of Cr(CH2CH3)[N(SiMe2CH2PPh2)2] (15) 96 (xvi) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with H2(D2)/PEt3 96 (xvii) Reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with CO 97 (xviii) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with CO 97 (xix) Conditions of reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with ethylene 97 (xx) Conditions of reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with ethylene 98 (xxi) Reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with acetylene 98 (xxii) Reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with phenylacetylene 98 vii (xxiii) Reaction of CoMe[N(SiMe2CH2PPfi2)2] (6) with phenylacetylene 98 (xxiv) Titration of {[(Ph2PCH2SiMe2)2N]Cr}2(p-Cl)2 (1) with donor ligands 99 2.8.5 Variable temperature magnetic susceptibility analyses 99 2.9 References 100 Chapter 3: ONE-ELECTRON OXIDATION REACTIONS OF Cr(II) COMPLEXES. . . 108 3.1 Introduction 108 3.2 Reaction of Cr(II) Chloride Complex 1 with AgBPh4 109 (i) Synthesis and structure of CrCl2(THF)[N(SiMe2CH2Ph 2) 2] (19).. 109 (ii) Synthesis of a cyclopentadienyl Cr(III) complex (22) 113 3.3 Reaction of Chromium(II) Methyl Complex 5 with PhSSPh 114 3.4 Reaction of Chromium(II) Complexes with Alkyl Halides 119 (i) Introduction 119 (ii) Reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with methyl iodide and bromide 120 (iii) Synthesis of a chromium(III) dialkyl complex 124 (iv) Survey of alkyl halide reactivity with chromium(II) complexes 130 3.5 Reactivity of Chromium(III) Complexes with Ethylene and Hydrogen 134 3.6 Comparison of Five-coordinate Cr(III) and Ir(III) Complexes 135 3.7 Summary and Conclusions 138 3.8 Experimental 139 3.8.1 General Procedures and Materials 139 3.8.2 Synthesis and Reactivity of Complexes 139 (i) CrCl2(THF) [N(SiMe2CH 2PPh 2) 2] (19) 139 viii (ii) C r C p C l [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (22) 140 (iii) CrMe(SPh) [N(SiMe 2 CH 2 PPh2)2] (23) 140 (iv) G (Me)Br [ [N(S iMe 2 CH 2 PPh2)2 ] (24) 141 (v) Cr (Me)I [N(SiMe 2 CH 2 PPh2)2] (25) 141 (vi) Attempted synthesis of C r R 2 [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (R = M e , C H 2 P h ) 142 (vii) C r (CH 2 SiMe3)Cl [N(S iMe2CH 2 PPh2)2 ] (26) 142 (viii) Attempted synthesis of C r ( C H 2 S i M e 3 ) R [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (R = M e C H 2 P h ) 143 (ix) C r ( C H 2 S i M e 3 ) 2 [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (27) 143 (x) Reaction of C r M e [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (5) with P h C H 2 C l and C F 3 C H 2 1 143 (xi) Reaction of CrMe[N(S iMe2CH 2 PPh 2 )2 ] (5) with 2-methylal lyl chlor ide 144 (xii) Reaction of Cr ( r | 5 -C 5 H 5 ) [N(S iMe2CH2PPh 2 ) 2 ] (11) w i t h P h C H 2 C l 144 (xiii) Reaction of { [ ( P h 2 P C H 2 S i M e 2 ) 2 N ] 2 C r } 2 ( u . - C l ) 2 (1) with P h C H 2 C l and ClCH 2 C02Me 145 (xiv) Reaction of { [ ( P h 2 P C H 2 S i M e 2 ) 2 N ] 2 C r } 2 ( p : - C l ) 2 (1) with M e l 145 (xv) Reaction of C r ( M e ) I [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (25) w i t h H 2 146 (xvi) Conditions of attempted reaction of C r ( R ) X [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] with ethylene 146 (xvii) Reaction of C r ( C H 2 S i M e 3 ) 2 [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (27) with ethylene 146 ix 3.9 References 147 Chapter 4: REACTIVITY OF COBALT(II) HALIDE AND A L K Y L COMPLEXES WITH A L K Y L HALIDES 153 4.1 Introduction 153 4.2 Reactivity of Cobalt(II) Halide Complexes with Benzyl Halides 155 (i) Synthesis and characterization of cobalt(ffl) dihalide complexes.... 155 (ii) Variable-temperature magnetic susceptibility of cobalt(III) dibromide 36 161 (iii) Kinetic and mechanistic points 164 4.3 Reactivity of Cobalt(II) Halide Complexes with Methyl Iodide 168 4.4 Reactivity of Cobalt(II) Methyl Complex 6 with Methyl Halides 171 (i) Synthesis of cobalt(II) halide complexes by redox substitution 171 (ii) Variable temperature magnetic susceptibility of cobalt(II) iodide complex 4 173 4.5 Reactivity of Cobalt(II) Methyl Complex 6 with Benzyl Halides 175 (i) Synthesis and structure of a zwitterionic cobalt(II) complex 175 (ii) Mechanism of intramolecular zwitterion formation 179 4.6 Summary and Comparison with Chromium(II) Redox Chemistry 183 4.7 Experimental 187 4.7.1 General Procedures and Materials 187 4.7.2 Variable Temperature Magnetics Measurements 187 4.7.3 Kinetic Measurements 187 4.7.4 Synthesis and Reactivity of Complexes 188 (i) CoCl2 [N(SiMe 2 CH 2 PPh2)2] (34) 188 (ii) CoBrCl[N(SiMe 2CH 2PPh2)2] (35) 188 (iii) CoBr2[N(SiMe 2CH 2PPh 2) 2] (36) 189 (iv) Reaction of CoBr[N(SiMe2CH2PPh2)2] (3) with Benzyl Bromide in the Presence of TEMPO 190 (v) Reaction of CoCl[N(SiMe2CH2PPh2)2] (2) with Mel. . . . 190 (vi) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with Mel: Synthesis of CoI[N(SiMe2CH2PPh2)2] (4) 190 (vii) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) withMeBr 191 (viii) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with Benzyl Chloride: Synthesis of CoCl2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (41)... 191 (ix) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with Benzyl Bromide: Synthesis of CoBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (42)... 192 (x) Reaction of Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] (10) with Benzyl Bromide 192 (xi) Reaction of Co(CH2Ph)[N(SiMe2CH2PPh2)2] (8) with Benzyl Bromide: 193 (xii) Reaction of CoX2[N(SiMe2CH2PPh2)2] (34, 36) with alkylating agents 193 4.8 References 194 Chapter 5: IRON(II) AND IRON(III) HALIDE AND A L K Y L C O M P L E X E S 200 5.1 Introduction 200 5.2 Synthesis and Reactivity of Iron(II) Halide Complexes 201 (i) Synthesis and characterization of FeX[N(SiMe2CH2PPh2)2] (X = Cl, Br) complexes 201 xi (ii) Reaction of FeCl[N(SiMe2CH2PPh2)2] (43) with donor ligands 202 5.3 Synthesis and Structure of a Bis-Ligand Iron(II) Complex 205 5.4 Synthesis and Characterization of Iron(II) Alkyl Complexes 210 (i) Attempted synthesis of iron methyl and benzyl complexes 210 (ii) Synthesis and structure of Fe(ri5-C5H5)[N(SiMe2CH2PPh2)2] (46) 212 (iii) Synthesis and structure of FeR*[N(SiMe2CH2PPh2)2] (R* = bulky alkyl) 213 (iv) Reaction of Fe{CH(SiMe3)2} [N(SiMe2CH2PPh2)2] (48) with small molecules 220 5.5 Reaction of Iron(II) Halide and Alkyl Complexes with Benzyl Halides 221 (i) Reaction of FeCl[N(SiMe2CH2PPh2)2] (43) with benzyl bromide 221 (ii) Reaction of Fe{ CH(SiMe3)2} [N(SiMe2CH2PPh2)2] (48) with benzyl bromide 223 5.6 Synthesis, Characterization and Magnetic Properties of Iron(III) Dihalide Complexes 225 5.7 Summary and Conclusions 237 5.8 Experimental 238 5.8.1 General Procedures and Materials 238 5.8.2 Synthesis and Reactivity of Complexes 238 (i) FeCl[N(SiMe2CH2PPh2)2] (43) 238 (ii) FeBr[N(SiMe 2 CH 2 PPh 2 ) 2 ] (44) 239 (iii) Reaction of FeCl[N(SiMe2CH2PPh2)2] (43) with pyridine 239 (iv) Fe[N(SiMe 2 CH 2 PPh 2 ) 2 ] 2 (45) 239 xii (v) Attempted Synthesis of FeMe[N(SiMe2CH2PPh2)2].. .240 (vi) Attempted Synthesis of Fe(CH2Ph)[N(SiMe2CH2PPh2)2] 240 (vii) Fe(Ti5-C5H5)[N(SiMe2CH2PPh2)2] (46) 241 (viii) Fe(CH 2SiMe3)[N(SiMe2CH2PPh 2)2] (47) 241 (ix) Fe{CH(SiMe3)2} [N(SiMe2CH2PPh2)2] (48) 242 (x) Reaction of F e { C H ( S i M e 3 ) 2 } -[N(SiMe2CH2PPh2)2] (48) with Ethylene 242 (xi) Reaction of Fe{ C H ( S i M e 3 ) 2 } -[N(SiMe 2 CH 2 PPh 2 )2] (48) with H 2 242 (xii) Reaction of FeX[N(SiMe 2 CH 2 PPh2)2] ( X = C l , Br) with Benzyl Bromide 242 (xiii) Reaction o f Fe{CH(SiMe 3)2)-[N(SiMe2CH2PPh2)2] (48) with Benzyl Bromide 243 (xiv) Reaction of FeBr2[N(SiMe2CH2PPh2)2] (49) with Alkyll i thium reagents 244 (xv) FeBr2[N(SiMe2CH2PPh2)2] (49) 244 (xvi) FeCl2[N(SiMe2CH2PPh2)2] (52) 244 (xvii) FeI2[N(SiMe2CH2PPh2)2] (53) 245 (xviii) FeClI[N(SiMe2CH2PPh2)2] (54) 245 5.8.3 Titration of FeCl[N(SiMe2CH 2PPh 2)2] (43) with pyridine 246 5.9 References 246 Chapter 6: EXTENSIONS, EXPANSIONS AND EXTRAPOLATIONS ...251 6.1 Introduction 251 6.2 Extensions Using Current Compounds 252 (i) Cationic, paramagnetic organometallic complexes 252 xiii (ii) Reduction of MX[N(SiMe2CH2PPh2)2] complexes 253 (iii) Organic radical generators and other organic applications 254 (iv) Other reactions 255 (v) Modification of the phosphine in the amidodiphosphine ancillary ligand 256 6.3 Expansion to Other Metals 256 (i) Molybdenum(III) 256 (ii) Manganese(II) and other metals 258 6.4 Extrapolations to a Second Generation Macrocyclic Amidophosphine Ligand 259 (i) Synthesis and structure of paramagnetic P 2 N 2 metal complexes.. .260 (ii) Reactivity of cobalt and chromium P 2 N 2 complexes 262 6.5 Conclusions and Summary 266 6.6 Experimental 268 6.6.1 General Procedures and Materials 268 6.6.2 Synthesis and Reactivity of Complexes 268 (i) CrCl[N(SiMe 2CH 2P I 'Pr 2) 2] (55) 268 (ii) MoCl 2(THF)[N(SiMe 2CH 2PPh2)2] (56) 269 (iii) Attempted preparation of MnX[N(SiMe2CH 2PPh2) 2]....269 (iv) Anti - Cr[P 2N 2] (58) 269 (v) Anti - Co[P2N2] (59) 270 (vi) Syn - Co[P 2N 2] (60) 270 (vii) Syn - MoCl[P 2N 2] (61) 271 (viii) Anti - CoI[P2N2] (62) 271 (ix) Reaction of anti - Co[P2N2] with Dioxygen Anti - Co[(PO)2N2] (63) 271 (x) Reaction of anti - Co[P2N2] with Sulfur xiv Anti - Co[(PS)2N2] (64) 272 6.7 References 272 Appendix 1: X-ray Crystal Structure Data 276 Appendix 2: Solid-State Variable Temperature Magnetic Susceptibility Data 308 Appendix 3: Cartesian Coordinates for Model Compounds used for ZINDO-MO-ROHF Calculations 312 Appendix 4: Kinetic Data for Co(II) One-Electron Oxidation Reaction 316 xv List of Tables Table Title Page Table 1.1 Spin-only magnetic moments for various numbers of unpaired 10 electrons. Table 1.2 Comparison of electronically similar paramagnetic vs. 27 diamagnetic metal centres. Table 2.1 Selected bond lengths and angles in {[(Ph2PCH2SiMe2)2N]Cr}2- 36 (H-C1) 2 (1). Table 2.2 UV-vis spectral data, equilibrium constants (equation 2.2) and 39 solution magnetic moments for {[(Ph2PCH2SiMe2)2N]Cr)2-(|J.-C1)2 (1) and its adducts with various ligands. Table 2.3 UV-vis spectral data for CoX[N(SiMe2CH2PPh2)2]. 41 Table 2.4 Selected bond lengths and angles in CoI[N(SiMe2CH2PPh2)2] 43 (4). Table 2.5 Selected bond lengths and angles in MCH3[N(SiMe2CH2PPh2)2] 48 (M = Cr, Co). Table2.6 Selected bond lengths and angles in 54 MCH2Ph[N(SiMe2CH2PPh2)2] (M = Cr, Co). Table 2.7 Selected bond lengths and angles in Cr(rj5- 61 C5H5)[N(SiMe2CH2PPh2)2] (11). Table 2.8 ZINDO relative total energy calculations for the simplified 65 complexes CrR(NH2)(PH3)2 in the indicated geometries. All values given are in kJ/mol. Table2.9 Selected bond lengths and angles in 71 Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13). Table 2.10 Selected bond lengths and angles in {[Ph2PCH2SiMe2)2N]Cr}2- 75 (|i-H)2 (14). xvi Table3.1 Selected bond lengths and angles in 112 CrCl2(THF)[N(SiMe2CH2PPh2)2](19). Table3.2 Selected bond lengths and angles in 117 GMe(SPh)[N(SiMe2CH2PPh2)2] (23). Table3.3 Selected bond lengths and angles in 122 CrMeBr[N(SiMe2CH2PPh2)2] (24). Table3.4 Selected bond lengths and angles in 127 Cr(CH2SiMe3)Cl[N(SiMe2CH2PPh2)2] (26). Table3.5 Selected bond lengths and angles in 130 Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27). Table 4.1 Selected bond lengths and angles in CoBr2[N(SiMe2CH2PPh2)2] 159 (36). Table 4.2 UV-vis spectral data for CoXY[N(SiMe2CH2PPh2)2]. 164 Table4.3 Selected bond lengths and angles in 176 CoCl2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (41). Table 5.1 Selected bond lengths and angles in Fe[N(SiMe2CH2PPh2)2]2 208 (45). Table 5.2 Selected bond lengths and angles in Fe(T| 5- 214 C5H5)[N(SiMe2CH2PPh2)2] (46). Table5.3 Selected bond lengths and angles in 218 Fe{CH(SiMe3)2]}[N(SiMe2CH2PPh2)2](48). Table 5.4 Selected bond lengths and angles in FeBr2[N(SiMe2CH2PPh2)2] 227 (49). Table 6.1 Selected bond lengths and angles in anti - Co[P2N2] (59). 261 Table 6.2 Selected bond lengths and angles in anti - CoI[P2N2] (62). 264 Table A1.1 Crystallographic data. 276 XVll Table A1.2 to A1.23 Table A2.1 to A2.5 Tables A3.1, -A.3*2j .A.3»3j A3.5, A3.6. Table A3.4 Table A4.1 Table A4.2 Table A4.3 Final atomic coordinates (fractional) and 5 e q (or U e q x 103) (A2) 298 to 307 for reported crystal structures (twenty two sets). Magnetic susceptibility, moment and temperature data for 308 to 311 complexes 1,14, 36, 4 and 59. Cartesian coordinates for model compounds used for ZINDO- 312 to 315 MO-ROHF calculations (nine sets). ZINDO relative total energy calculations for the simplified 314 complex CoMe(NH2)(PFl3)2 in the indicated geometries. All values given are in kJ/mol. Raw UV-vis data for CoCl[PNP] (2, 10"3 M) reaction with 316 PhCH2Br (0.083 M). kobs data at different concentrations of benzyl bromide. 317 k0bs data at different concentrations of benzyl chloride. 318 xviii List of Figures Figure Caption Page Figure 1.1 Paramagnetic Organometallic Complexes: A bridge between 3 classical Werner-type coordination compounds and low-valent organometallic compounds. Figure 1.2 Zeeman splitting of a S = 1/2 Kramers' doublet. 8 Figure 1.3 D-orbital splitting diagrams for d 6 (a) high-spin octahedral; (b) 11 low-spin octahedral; (c) high-spin tetrahedral systems. A D is the octahedral crystal field splitting parameter; the tetrahedral analogue is A t. The number of unpaired electrons is n. Figure 1.4 D-orbital splitting diagrams for intermediate-spin d 6 (a) square 12 planar; (b) square pyramid; (c) trigonal bipyramid geometry. The number of unpaired electrons is n. Figure 1.5 (a) Zeeman splitting of an S = 3/2 state (b) with zero-field 15 splitting. Figure 1.6 (a) Effect of a magnetic field on an S = 1/2 Kramers' doublet, (b) 20 Effect of large zero-field splitting and a magnetic field on an S = 3/2 system. Expected ESR transitions are shown as solid arrows. D is the zero-field splitting separation. Figure 1.7 Effect of a magnetic field on an S = 1 system with (a) no zero- 21 field splitting (b) small zero-field splitting (c) large zero-field splitting. ESR transitions are shown as solid arrows: (a) one transition; (b) two transitions; (c) no transitions (within spectrometer range). D is the zero-field splitting separation. Figure 1.8 Factors affecting NMR vs. ESR spectrum observation. 22 xix Figure 2.1 Molecular structure (ORTEP) and numbering scheme for 36 {[(Ph2PCH2SiMe2)2N]Cr}2(p-Cl)2 (1). Phenyl substituents on phosphorus have been removed for clarity. Figure 2.2 Molecular structure (ORTEP, 50% ellipsoids) and numbering 43 scheme for CoI[N(SiMe2CH2PPh2)2] (4). Figure 2.3 Molecular structure (ORTEP) and numbering scheme for 46 CrMe[N(SiMe2CH2PPh2)2] (5). Figure 2.4 Molecular structure (ORTEP, 50% ellipsoids) and numbering 50 scheme for CoMe[N(SiMe2CH2PPh2)2] (6). Figure 2.5 lH NMR spectrum of CoMe[N(SiMe2CH2PPh2)2] (6) in 51 C 6D 6(*). Figure 2.6 Molecular structure (ORTEP) and numbering scheme for 53 MCH2Ph[N(SiMe2CH2PPh2)2] (M = Cr (7), Co (8)). Figure 2.7 Molecular structure (ORTEP) and numbering scheme for Cr(rj5- 61 C5H5)[N(SiMe2CH2PPh2)2] (11). Figure 2.8 lH NMR spectrum of Co(C5H5)[N(SiMe2CH2PPh2)2] (12) in 62 Q>D6(*). Figure 2.9 Simplified coordination sphere models for complexes 2, 5 and 11 65 (left to right) used in ZINDO ROHF MO-energy calculations. Figure 2.10 Graph of calculated relative energies vs. spin multiplicity. 66 Figure 2.11 Molecular structure (ORTEP) and numbering scheme for 71 Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13). Phenyl substituents on phosphorus have been removed for clarity. Figure 2.12 Molecular structure (ORTEP) and numbering scheme for 75 {[Ph2PCH2SiMe2)2N]Cr}2(p-H)2 (14). Phenyl substituents on phosphorus have been removed for clarity. xx Figure 2.13 Magnetic susceptibility (per mole of dimer) vs. temperature plot 78 for chloride 1. The line was generated using the Heisenberg dimer model with J = -12.4 cm"1, g = 1.99 and P = 0.007. Figure 2.14 Magnetic susceptibility (per mole of dimer) vs. temperature plot 78 for hydride 14. The line was generated using the Heisenberg dimer model with J = -139 cm"1, g = 1.98 and P = 0.0017. Experimental data were corrected for TIP = 300 x 10"6 cirPmoH. Figure 3.1 Molecular structure (ORTEP) and numbering scheme for 112 CrCl 2(THF)[N(SiMe2CH 2PPh2)2] (19). Figure 3.2 Molecular structure (ORTEP) and numbering scheme for 116 CrMe(SPh)[N(SiMe2CH2PPh2)2] (23). Figure 3.3 Molecular structure (ORTEP) and numbering scheme for 121 CrMeBr[N(SiMe2CH2PPh2)2] (24). Figure 3.4 Molecular structure (ORTEP) and numbering scheme for 126 Cr(CH2SiMe3)Cl[N(SiMe2CH2PPh2)2] (26). Figure 3.5 Molecular structure (ORTEP) and numbering scheme for 129 Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27). Figure 3.6 (a) High-spin Cr(III) and (b) low-spin Ir(III) d-orbital occupancy 136 in a trigonal-bipyramidal geometry; in both cases Jahn-Teller distortion is necessary. Figure 4.1 ^H NMR spectrum of CoCl 2[N(SiMe 2CH 2PPh 2) 2] (34) in 158 C 6D 6(*). Figure 4.2 Molecular structure (ORTEP) and numbering scheme for 159 CoBr2[N(SiMe2CH2PPh2)2] (36). Figure 4.3 Graph of magnetic moment vs. temperature for 162 CoBr2[N(SiMe2CH2PPh2)2] (36). xxi Figure 4.4 Graph of magnetic moment vs. temperature for 174 CoI[N(SiMe2CH2PPh2)2] (4). Figure 4.5 Molecular structure (ORTEP) and numbering scheme for 176 CoCl2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (41). Figure 5.1 Spectrophotometric data for the titration of 204 FeCl[N(SiMe2CH2PPh2)2] (43) with pyridine. Data are obtained at 25 °C. Spectrum #7 from the bottom, where the isosbestic point is broken, is assumed to be the first endpoint. Figure 5.2 Spectrophotometric data for the titration of 204 FeCl(py)[N(SiMe2CH2PPh2)2] (43»py) with pyridine. Data are obtained at 25 °C. The endpoint is at the top spectrum, where [py] = 1.64 M. Figure 5.3 Molecular structure (CHEM 3D®) and numbering scheme for 208 Fe[N(SiMe2CH2PPh2)2]2 (4 5). Phenyl substituents on phosphorus have been removed for clarity. Figure 5.4 Molecular structure (ORTEP) and numbering scheme for Fe(rj5- 214 C5H5)[N(SiMe2CH2PPh2)2] (46). Figure 5.5 Mass spectrum of Fe{CH(SiMe3)2} [N(SiMe2CH2PPh2)2] (48). 216 Figure 5.6 Molecular structure (ORTEP) and numbering scheme for 218 Fe {CH(SiMe3)2} [N(SiMe2CH2PPh2)2] (48). Figure 5.7 Molecular structure (ORTEP) and numbering scheme for 227 FeBr2[N(SiMe2CH2PPh2)2] (49). Figure 5.8 UV-vis spectra for FeX2[N(SiMe2CH2PPh2)2]. X = Cl (52, top), 229 X = Br (49, bottom). Figure 5.9 Variable Temperature ! H NMR spectra for 231 FeBr2[N(SiMe2CH2PPh2)2] (49) from 6-12 ppm. Spectrum (a) 60 °C, (b) 40 °C and (c) 25 °C. xxii Figure 5.10 Graph of magnetic moment vs. temperature for 234 FeBr2[N(SiMe2CH2PPh2)2] (49). Figure 5.11 Energy levels for iron(III) electronic states in a tetragonal field. 235 (a) A/£ is large and negative; (b) A/£ is large and positive; (c) A/C, < 1, spin admixed system. Taken from reference 73. Figure 6.1 lH NMR spectrum of MoCl2(THF)[N(SiMe2CH2PPh2)2] (56) in 257 C 6D 6(*). Figure 6.2 Molecular structure (ORTEP) and numbering scheme for anti - 261 Co[P2N2] (59). Figure 6.3 Molecular structure (ORTEP) and numbering scheme for anti - 264 CoI[P2N2] (62). Figure A4.1 Raw UV-vis data for CoCl[PNP] (2) + PhCH2Br (0.083 M). 316 Figure A4.2 Pseudo-first order rate plot of raw data (Table A4.1). 317 Figure A4.3 Second-order rate plot for CoCl[PNP] (2) + PhCIi^Br; kob s vs. 318 [PhCH2Br]. xxiii GLOSSARY OF TERMS The following abbreviations, most of which are commonly found in the literature, used in this thesis. i H proton 2 H or D deuterium 3 1 P {1H} observe phosphorus while decoupling proton A absorbance A angstrom (10"10 m) acac acetylacetonate Anal. analysis Arf 2,5-C6H3FMe atm atmosphere B magnetic induction BAr f {B[3,5-(F3C)2C6H3]4}-B e q equivalent isotropic thermal parameter [(8/3)7t2Ueq] bipy bipyridine B.M. Bohr magneton br broad "Bu w-butyl group, -CH2CH2CH2CH3 'Bu tertiary butyl group, -C(CH3)3 Bz benzyl group, -CH2PI1 C Curie constant Calcd. calculated CCD charge coupled device Chem 3D® molecular modelling program for the Macintosh computer cm"1 wave number xxiv Cp cyclopentadienyl group, [C5H5]" Cp' substituted Cp ligand, e.g. [CsFL^Me]-Cp* pentamethylcyclopentadienyl group, {C5(CH3)5}-CT charge transfer cyclam 1,4,8,11 -tetraazacyclotetradecane D axial zero-field splitting parameter d doublet DC Diet Coke deg (or °) degrees depe 1,2-bis(diethylphosphino)ethane dippe l,2-bis(diisopropylphosphino)ethane dippp l,3-bis(diisopropylphosphino)propane DME 1,2-dimethoxyethane DMF N,N-dimethylformamide dmpe 1,2-bis(dimethylphosphino)ethane dmpm bis(dimethylphosphino)methane d n number of d-electrons d n n-deuterated dppe 1,2-bis(diphenylphosphino)ethane dppm bis(diphenylphosphino)methane dppp l,3-bis(diphenylphosphino)propane EAN effective atomic number EI electron impact en ethylenediamine, N H 2 C H 2 C H 2 N H 2 equiv equivalent(s) ESR electron spin resonance Et ethyl group, -CFE2CH3 XXV eV electron Vol t E X A F S x-ray absorption fine structure F fit quality factor G Gauss g grams, or Lande splitting factor (g-value) G C - M S Gas Chromatography/Mass spectrometry H external applied magnetic field h Planck's constant F f M P A hexamethylphosphoramide H z Hertz, seconds"1 I N D O intermediate neglect of differential overlap IR infrared J magnetic exchange coupling constant K degrees Ke lv in or equilibrium constant k rate constant or Boltzmann's constant k J kiloJoules kobs observed rate constant L neutral two-electron donor ligand M magnetization M central metal atom (or "molar", when referring to concentration) M + parent ion m meta, or gram mass of sample in 1 m L solvent (Evans' method) M e methyl group, - C H 3 mle mass/charge (mass spectroscopy unit) Mes mesityl group, -2,4,6-Me3C6H2 Mes* super-mesityl group, -2,4,6-fBu3C6H2 mg milligram(s) xxv i MHz megaHertz mL millilitre mmol millimole MO molecular orbital mol mole MS mass spectrometry m s spin quantum level N Avogadro's number nm nanometers NMR nuclear magnetic resonance o ortho OD optical density OEP octaethylporphyrin ORTEP Oakridge Thermal Ellipsoid Plotting Program P fraction of paramagnetic impurity p para Pc phthalocyanine Ph phenyl group, -C6H5 phen phenanthroline [PNP] amidodiphosphine ligand, "N(SiMe2CH2PPh2)2 [P2N2] diamidodiphosphine hgand, [PhP(CH2SiMe2NSiMe2CH2)2PPh] [(PO)2N2] [PhP(0)(CH2SiMe2NSiMe2CH2)2P(0)Ph] P-P chelating diphosphine ligand ppm parts per million *Pr isopropyl group, -CH(CH3)2 [(PS)2N2] [PhP(S)(CH2SiMe2NSiMe2CH2)2P(S)Ph] PVC polyvinyl chloride xxvii py pyridine R hydrocarbyl substituent R* bulky hydrocarbyl substituent ROHF restricted open Hartree-Fock rt room temperature S total spin quantum number s singlet or seconds sh shoulder (UV-vis spectroscopy) solv solvent SOMO singly occupied molecular orbital SQUID superconducting quantum interference device T temperature t triplet TEMPO 2,2,6,6-tetramethyl-1-piperidinyloxy THF ether tetrahydrofuran TIP temperature independent paramagnetism TMEDA 1,1,2,2-N,N,N',N'-tetramethylethylenediamine tmtaa tetramethyl-dibenzotetraaza[ 14] annulene Tns Toranosuke TPP tetraphenylporphyrin triphos MeC(CH2PPh2)3 U eq equivalent isotropic thermal parameter UV-vis ultraviolet-visible v br very broad VSM vibrating sample magnetometer VT variable temperature Zeeman coefficient xxviii w weak X halide substituent x-dn Complex x.y has n number of A H atoms replaced by 2 H atoms ZFS zero field splitting ZINDO Zerner intermediate neglect of differential overlap P electronic Bohr-magneton, A. spin-orbit coupling constant A lODq, crystal field splitting parameter A x x = o, octahedral 10 Dq; x = t, tetrahedral 10 Dq 5 chemical shift £ extinction coefficient Xx x magnetic susceptibility (x = v, volume; x = g, gram; x = m, mol i"ln n-hapto (hapticity) Av change in frequency (Hz, Evans' method) v 0 spectrometer frequency (MHz, Evans' method) v x x vibrational band for bond xx |4-X bridging X-ligand |4eff effective magnetic moment m.o. spin-only magnetic moment t tau value for five-coordinate complexes C spin-orbit coupling constant; X = ± CJ2S [ 15]aneN4 1,5,9,13-tetraazacyclopentadecane °C degrees Celsius xxix ACKNOWLEDGEMENTS I would first and foremost like to thank my supervisor, Professor Michael Fryzuk, for his guidance, encouragement and patience during my course of study. Many thanks go to past and present members of the Fryzuk group, including Garth Giesbrecht, Paul Duval, Michael Bowdridge, Sam Johnson, Laleh Jafapour, Fr6d6ric Naud, Erin Ma, Shane Mao and Drs. Jason Love, Annie Schneider, Jim Kickham, Guy Clentsmith, Murugesapillai Mylvaganam, Xiaoliang Gao and Martin Stefan for their friendship, support and illuminating discussions. It is highly interesting to note the novel and exciting interactions I have had associating with such a talented group of people. I greatly appreciate Professor Robert Thompson for his invaluable aid; I have learned a great deal from him. Thanks also to Xia and Dave Summers of the Thompson group for helping me with magnetic measurements. I appreciate the help of the UBC Department of Chemistry support staff, in particular Mr. Marshall Lapawa (mass spectroscopy), Mr. Peter Borda (elemental analysis), Ms. Marietta Austria ( ! H NMR spectroscopy), Ms. Elizabeth Varty (illustrator) and Mr. Steve Rak (glassblowing). A very special thanks is warranted to Dr. Steven Rettig, who was invaluable in producing top-quality crystal structure solutions. CCD structures were solved by Dr. Victor Young (University of Minnesota) and Dr. Glenn Yap (University of Windsor); many thanks to them as well. I am grateful for financial support from NSERC (NSERC 1967 Scholarship) and the University of British Columbia (I. W. Killam Fellowship). Daniel Bernard Leznoff xxx To my parents and Junko with love and respect xxxi One gains knowledge from doubt, doubt from knowledge. Torahiko Terada, physicist. xxxii Chapter One Paramagnetic Organometallic Complexes - A General Introduction 1.1 Coordination vs. Organometallic Chemistry It has been said that modern coordination chemistry began in 1893 with the work of Alfred Werner, who postulated the existence of first and second coordination spheres around transition metal centres to explain the different coloured isomers of octahedral cobalt(III) compounds.1 Werner was awarded the Nobel Prize in 1913 in recognition of his seminal role in opening the field of coordination chemistry. Werner and his contemporaries made particular use of complexes containing cobalt and chromium metal centres in the +2 and +3 oxidation states; hence these valences could be described as "classical". This predominance was due to the fact that the +3 centres were kinetically inert while the +2 centres were kinetically labile.2 Coordination complexes, combinations of metal centres surrounded by ligands, have been a prime feature of inorganic chemistry ever since. Archetypal examples of such complexes include the hexaaqua 1 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction series of the first-row transition metals, M ( H 2 0 ) 6 n + (n = 2,3), and complexes such as [Cr(NH3)6]3+ and [C0CI4]2".1'2 Generally, a-donor ligands such as H2O and NH3 are featured in classical coordination compounds. The bonding and spectra in such complexes were interpreted using crystal field and ligand field theory;3 the stabilities of such complexes were also interpreted in this l ight. 1 , 2 In addition, due to the prevalence of first-row transition metals in classical coordination chemistry, many of these complexes are paramagnetic.4 Organometallic chemistry is defined as the study of compounds containing a metal-carbon bond. 2 , 5 Although compounds such as Zeise's salt, K[(C2H4)PtCl3] and Ni(CO)4 were prepared as early as 1827 and 1890 respectively,5 modern transition-metal organometallic chemistry can be said to have begun with the discovery 6 , 7 and subsequent bonding description8 of ferrocene in 1951, for which Wilkinson and Fischer won the Nobel Prize in 1973. Typical transition-metal organometallic compounds include the aforementioned ferrocene, bis(benzene)chromium, and metal carbonyls. Flistorically, low-valent transition metal centres with 7t-acid ligands like CO and PR3 have dominated the organometallic literature.5 ,9 Bonding in organometallic complexes is usually explained using molecular orbital theory and organometallic stability is governed by the effective atomic number (EAN) or so-called "18-electron rule". 5 , 9 , 1 0 As a result, the vast majority of organometallic complexes reported are diamagnetic.1 1 , 1 2 Although the two fields of organometallic and coordination chemistry have been presented as perhaps disparate and contrasting fields, it is completely accurate to state that organometallic chemistry is really the coordination chemistry of compounds with metal-carbon bonds. That is, organometallic complexes are special cases of coordination complexes where at least one ligand is bound to the metal through a carbon atom.9 That said, the general contrasts between the two fields remain valid. Paramagnetism in coordination complexes is unexceptional; in organometallic complexes, it is unusual. Hence, paramagnetic organometallic complexes, or paramagnetic coordination complexes with a metal-carbon bond, represent a bridge between the two extremes defined by Werner-type complexes and low-valent organometallic complexes (Figure 1.1). 2 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction Werner - Coordination Compounds T H 3 N 1 3 + r i2-CI H 3 N V ,..NH3 S N. Co H 3 N ^ X N H 3 C r ^ \ C | L H 3N J [ u • Paramagnetic Organometallic Complexes Fe o c ^ ' x ^ r ^ f f ^ r cx / C 0 O C ^ I ^ C O N / CO Organometallic Compounds Figure 1.1 Paramagnetic Organometallic Complexes: A bridge between classical Werner-type coordination compounds and low-valent organometallic compounds. The synthesis and reactivity of paramagnetic metal alkyl complexes, in particular, reactions of the metal-carbon bond, are of interest from both an industrial view regarding catalytic systems, and from an academic view in terms of the paucity of such compounds.13 Reactivity patterns unique to the presence of unpaired electron density at the metal centre could be utilized; for example, radical-based reactivity unavailable to diamagnetic organometallic systems could be accessed.14 There is precedent that paramagnetic, highly reactive complexes are important as active catalysts15"17 in industrial systems, the best example being the chromium-catalyzed 3 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction polymerization of ethylene. 1 8" 2 0 Hence an examination of paramagnetic organometallic species is expected to be fruitful on many fronts. 1.2 History of Paramagnetic Organometallic Research Odd-electron organometallic compounds were once viewed as isolated exceptions to the 18-electron rule but explorations into such necessarily paramagnetic systems has led to a refuting of such a statement.14 Research concerning paramagnetic organometallic systems can be divided into several general areas, namely stable and unstable 17- and 19-electron systems, stable cyclopentadienyl ring-type stabilized systems and non-Cp/non-CO systems. The overview below is by no means comprehensive and further relevant background (in particular, chromium, iron and cobalt systems) w i l l be presented in the appropriate chapter. (i) Stable 17- and 19-electron systems The concept of stable 17-electron systems is no longer revolutionary. 2 1 Stable 19-electron systems, while not as common as the 17-electron systems, have been synthesized 2 2 " 2 4 and their role as intermediates in organometallic reaction mechanisms has been acknowledged. 2 5 ' 2 6 Even among the earliest literature one can find examples of 17- and 19-electron systems which, at the time, were considered unusual exceptions to the eighteen-electron rule. The metal-carbonyl V(CO)6 is a classic example of a 17-electron compound and has been wel l studied as a result. 2 7 Ferrocinium salts [Cp2Fe] + X" are also 17-electron compounds . 2 8 , 2 9 The standard example of a 19-electron compound, also paramagnetic, is cobaltocene, C p 2 C o . 3 0 , 3 1 Other examples of stable 17-electron systems include M n ( C O ) 3 L 2 (L = bulky phosphine) , 3 2 CpCr(CO)2PPh3, 3 3 Cp'Cr(NO)(L)X (Cp' = Cp, Cp*, L = 2 electron donor, X = h a l i d e ) 3 4 , 3 5 and R e ( C O ) 3 ( P P h 3 ) 2 . 3 6 , 3 7 4 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction (ii) Unstable 17-electron metal-centred radicals A substantial body of literature pertains to the generation and subsequent reactivity of unstable 17-electron metal-centred radicals. Methods of generation and reactivity of these organometallic radicals have been reviewed.14'25'26 A large class of these compounds are dimers containing metal-metal bonds; the radical monomers are generated thermally or photochemically and are unstable with respect to reformation of the dimer (equation 1.1) 2 6 , 3 8 ~ 4 2 hv L n M — M L n ^ 2 M L n • M L n = M ( C O ) 5 (M = Mn, Re), C p M ( C O ) 3 (M = Cr, Mo, W) C p M ( C O ) 2 (M = Fe, Re, Os), C o ( C O ) 4 , CpNi(CO) (iii) Compounds containing multiple unpaired electrons In comparison to compounds with one unpaired electron (17- and 19-electron systems, d 1 and d^  systems and other S = 1/2 metal centres) organometallic compounds containing more than one unpaired electron (S > 1/2) are less well studied. An excellent recent review by Poli of paramagnetic organometallic compounds13 showed that the majority of paramagnetic organometallic literature is based on complexes containing Cp-type ligands (either one or two). Consequently, compounds prepared in this thesis by and large avoid the use of Cp as an ancillary ligand. In the absence of the Cp fragment as a stabilizing ligand, chelating phosphines have been used to stabilize paramagnetic organometallic systems. Compounds containing bis(diphenylphosphino)methane (dppm) were generally found to be octahedral, low-spin systems.43 Hermes44 and Girolami prepared a series of high-spin (S > 1/2) metal alkyls using the sterically demanding chelating diphosphine bis(diisopropylphosphino)ethane (dippe) with Cr(II)45 Fe(II)46 and Cr(III).47 The chemistry of chromium(III) has been examined in detail using Cp-based ligands by Theopold48 and recently Gambarotta has prepared chromium(II) alkyls stabilized 5 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction byTMEDA. 4 9 One final series of compounds deserves special comment, namely the aquo alkyls [RCr(H20)5] 2 + , 5 0 which serve as a unique bridge between Werner-type complexes and organometallic chemistry. (iv) Homoleptic Paramagnetic Organometallic Compounds The use of very bulky alkyl groups has allowed for the isolation of highly coordinately and electronically unsaturated homoleptic metal alkyls, some of which are paramagnetic.51 Examples include MR 3 (R = CH(SiMe3)2, M = Ti (d1), V (d2), Cr (d3)),52-53 M(CH 2SiMe 3) 4 (M = V (d1), Cr (d2))5 4-5 5 and Mn(CH2SiMe3)2 (d5). The series of M(l-norbornyl)4 (M = Ti, Zr, Hf, V, Cr, Mn, Fe, Co) 5 6 complexes deserves special mention since they are rare examples of highly oxidized metal-alkyl complexes. These are the only known Co(IV) and Fe(IV) alkyl complexes, which also have the unique distinction of being tetrahedral and low-spin (section 1.3).56"58 Anionic paramagnetic homoleptic complexes also exist and are particularly prevalent in the case of chromium where complexes such as CrPh6Li3(Et20)2.5 have been well studied. 5 1 , 5 9 ' 6° The bulky aryl group 2,4,6-trimethylphenyl, or mesityl, has been used to synthesize very low-coordinate homoleptic aryl complexes with metal centres such as V(III) and Cr(IV) (d2), Cr(III) (d3), Cr(II) (d4), Mn(lT) (trimeric, d5), Fe(II) (d6) and Co(II) (d7).60 1.3 Magnetism: A Brief Tutorial6 1 6 3 This thesis is concerned with paramagnetic organometallic compounds and hence the measurement and explanation of the magnetic behaviour of the compounds synthesized is an important feature. In order to set the stage with regard to magnetic effects, a discussion of the theory of magnetism in transition metal complexes is necessary. Rather than use quantum mechanics and matrix mathematics the discussion below is an attempt to "paint" a physical picture 6 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction of some of the magnetic phenomena observed in transition metal complexes. A detailed theoretical explanation of magnetic behaviour is beyond the scope of this thesis and can be explored in various treatises on the subject.4'63"65 Magnetic effects are generated by the motion of charged particles. In transition metals, the intrinsic spin and the orbital motion of the electron (orbital angular momentum) both contribute to the overall paramagnetism of a metal centre. In an external magnetic field H , the magnetic dipoles in the compound align with the field, causing a magnetization M . The magnetization of the substance results in an induced field in the substance, called the magnetic induction B . The magnetic induction B hence describes the behaviour of a substance in an external magnetic field, and is given by:62*63 B = H + 4TUM [ 1 . 2 ] where H is the applied field strength. What is usually measured is %v» the magnetic susceptibility per unit volume which, if M is isotropic, is given by: Xv = M / H [ 1 . 3 ] This volume susceptibility can be converted to gram (%g) and molar (%m) susceptibilities using sample density and molecular weight, respectively. There are two types of fundamental magnetic behaviour exhibited by metal complexes: diamagnetism and paramagnetism.61"63 All substances exhibit diamagnetic behaviour, which is characterized by the field induced by circulations of pairs of electrons. Diamagnetism is a small effect characterized by a small, negative % value. Paramagnetism typically results from magnetic effects due to the spin and orbital motion of unpaired electrons, is much more potent than diamagnetism and is characterized by large, positive % values. In this thesis, the term "paramagnetic" applies to systems with unpaired spins, while "diamagnetic" refers to spin-paired compounds. 7 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction It is important to realize from the outset that magnetism is a bulk property. Calculations concerning magnetism are based on the idea that one or more energy levels of a system are populated and the distribution of that population, governed by the Boltzmann distribution, defines the susceptibility at any given temperature. The key lies in defining the appropriate energy levels to be populated. In the most simple case, a system with one unpaired electron (S = 1/2), there are only two levels of interest. An S = 1/2 state is composed of two quantum levels, namely m s = +1/2 and m s = -1/2. This doublet of states is referred to as a Kramers' doublet. The defining feature of a Kramers' doublet is that its double degeneracy can only be removed by an external magnetic field.61"63 This splitting of a Kramers' doublet is called the Zeeman splitting effect (Figure 1.2) and is the source of magnetic effects, as well as being key to both NMR and ESR spectroscopy. Using the equation for the energy difference between the m s = +1/2 and -1/2 states, the equation for the spin angular momentum of the electron and the Boltzmann distribution for thermal population of the two levels, and assuming that thermal energy (kT) is greater than the Zeeman sphtting (usually the case for first-row transition metals) gives the expression: + 1/2 - 1/2 Increasing H > Figure 1.2 Zeeman sphtting of a S = 1/2 Kramers' doublet. Xm -NAg2j32 4kT [1.4] 8 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction where N A is Avogadro's number, g is a proportionality constant (approximately 2.0 for an electron with no orbital angular momentum), (5 is the Bohr magneton of the electron and k is the Boltzmann constant. This equation can be rewritten as: Xm - Y C = 4k~^ ~ [1.5] This is the common form of the Curie law which predicts that the susceptibility is inversely proportional to the temperature; hence a plot of % vs. 1/T should yield a straight line. This specific formula has a more general form: N A g 2 3 2 Xm = - i f e ^ S t S + 1) [1.6] It is common to employ a quantity called the effective magnetic moment (u,eff), which is defined as: u e f f = (3k /N A (32 ) l/2 ( X m T )V2 = 2.828 (Xm T)V2 B.M. [1.7] Assuming no orbital contribution to the magnetic moment, values for "spin-only" magnetic moments (u,s.0.) can be calculated from: Hs.o. = g[S(S + i y j V 2 B.M. [1.8] which is obtained by combining equations 1.5 and 1.6. The values measured experimentally for p:eff can therefore be correlated to the number of unpaired electrons in the system, an extremely useful piece of information. Table 1.1 shows the expected values for spin-only magnetic moments. Note that this assumption of zero orbital angular momentum is useful as a first approximation for first-row transition metals but breaks down with heavier metals and the lanthanides. 9 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction Table 1.1 Spin-only magnetic moments for various numbers of unpaired electrons. Number of unpaired electrons S M-eff (spin-only) B.M. 1 1/2 1.73 2 1 2.83 3 3/2 3.87 4 2 4.90 5 5/2 5.92 The number of unpaired electrons in a transition metal complex is related to the splitting of its d-orbitals, which can be qualitatively predicted by crystal field theory. In an octahedral complex, the d-orbitals split into two levels as shown in Figure 1.3. The orbitals are filled by obeying Hund's rule, so in a d 1 , d 2 or d 3 complex it is obvious that there will be one, two and three unpaired electrons, respectively. In the d 4 to d 7 cases, however, there are two possible ways to put the electrons into the orbitals. If the electrons are located such that the maximum number of unpaired spins result, this situation is referred to as a high-spin system. If the electrons pair up as much as possible a low-spin system results. Whether a high or a low-spin system is observed is related to the pairing energy of the complex, that is, whether the energy gained by putting electrons in the lower energy level is offset by the energy required to pair the electrons (pairing energy). The energy difference between the T 2 g and E g levels (lODq or A) is a function of the metal and ligands present. As an example, for a d 6 octahedral complex, a high-spin system yields 4 unpaired electrons (S = 2, Figure 1.3(a)), while a low-spin system (S = 0, Figure 1.3(b)) will be completely spin-paired. In a tetrahedral geometry, the d-orbitals split as shown in Figure 1.3(c) and again the same choices exist. In practical terms, however, almost all tetrahedral complexes are high-spin due to the much smaller 10 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction energy difference between lower and upper levels. The aforementioned cobalt and iron(IV) norbornyl compounds are unique examples of low-spin tetrahedral complexes.56"58 d x 2 . y 2 d z 2 -t-4-f-n=4 (a) Ao d x y d x z d y z +i4_ |4_ |4_i4| i f . n=0 (b) - t - - M - ^xy d x z d At 4f4^ n=4 (c) d x 2 . y 2 d z 2 Figure 1.3 A schematic of the d-orbital splitting diagrams for d 6 (a) high-spin octahedral; (b) low-spin octahedral; (c) high-spin tetrahedral systems. A Q is the octahedral crystal field splitting parameter; the tetrahedral analogue is A T . The number of unpaired electrons is n. For regular tetrahedral and octahedral complexes only high and low-spin systems can result. It is impossible to have the intermediate-spin case, S = 1, in such geometries. However, in square planar or five-coordinate systems such a situation becomes possible, assuming one d-orbital becomes so high in energy that it cannot be populated at all. This is shown in Figure 1.4 for square planar, square pyramidal and trigonal bipyramidal ligand arrays and a d 6 system. Note that intermediate-spin systems are much less common than high and low-spin systems; this may be due to the need for a tuned ligand set, but may also be due to the relative scarcity of compounds in the appropriate geometry. 11 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction dz2 .-4- d xy dz2 ~~T- d \ ^ - f - - r - d x y dx2.y2 xy \ ^ dz2 Mf 4f d XZ d; M f # d x z d Mj- 4 f d x z d; n=2 n=2 n=2 (a) (b) (c) Figure 1.4 A schematic of the d-orbital sphtting diagrams for intermediate-spin d 6 (a) square planar; (b) square pyramidal; (c) trigonal bipyramidal geometry. The number of unpaired electrons is n. The above discussion makes no mention of any orbital angular momentum contribution to the magnetic moment. A quick survey of experimental data for first-row transition metals shows that in some cases the spin-only formula correctly predicts the moment, while in others the prediction is less accurate. The inconsistencies are due to orbital contributions. When can such contributions be expected? In most free ions, the magnetic moment has a full contribution from orbital angular momentum but upon application of a ligand field, such contribution is "quenched". If an electron can occupy degenerate orbitals that permit circulation of the electron about an axis, orbital angular momentum will not be fully quenched. For example, d 1 octahedral systems (Ti(III), V(IV)) have a large orbital contribution to their moment since the single electron is in the dXy, d x z and d y z triad, which are symmetry related. Hence the single electron can circulate throughout the three d-orbitals and an orbital angular momentum contribution to the magnetic moment is observed. This is referred to as a first-order orbital angular momentum contribution. 12 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction On the other hand, a d 3 octahedral system (Cr(III)) half-fills the lower three d-orbitals, consequently exchange is not possible and spin-only p^ ff values to first-order result. A low-spin d 7 system also has one electron in two degenerate orbitals but this time the unpaired electron is in the d z 2 , d x 2 . y 2 set, which is not symmetry related and hence spin-only moments would be expected. In fact, there is an orbital angular momentum contribution to the magnetic moment here but it is second-order in nature. The ground state has no orbital angular momentum but quantum-mechanical mixing of low-lying excited states into the ground state allows for an orbital contribution to the moment. This is observed for high-spin Co(II) tetrahedral systems and low-spin Co(II) square-planar systems as well. Usually second order effects are substantially less in magnitude than first-order contributions. As well, magnetic moments with first-order angular momentum contributions are strongly temperature dependent, while moments with second-order contributions generally obey the Curie law. The magnitude of the coupling of spin and orbital angular momentum is described by the spin-orbit coupling constant X, which has an empirical value for each metal in a given oxidation state. Another parameter, is also used and is defined as X, = ±(£ / 2S), where S is the spin quantum number of the system. Less than half-filled electronic shells have positive constants, while greater than half-filled shells have negative A. values. As a result, orbital contributions to Ti(III) d 1 systems will reduce the spin-only moment, while orbital contributions to Co(II) d 7 systems will raise the observed moment. The magnitude of X increases with oxidation state and atomic number. As a result, second and third row transition metal magnetochemistry becomes more and more dominated by spin-orbit coupling effects and the spin-only assumption becomes invalid. The addition of factors such as spin-orbit coupling and orbital angular momentum contributions to the moment invalidate the Curie law. A much more general treatment of magnetic behaviour vs. temperature was provided by Van Vleck, who derived the equation that predicts magnetic susceptibility vs. temperature given the appropriate energy level diagram to be populated: 13 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction kT 2W n ( 2 ) ] exp ( kT n X = N [ 1 . 9 ] The terms W n refer to the nth energy level being populated. The numerical superscripts refer to the energy series: E n = W n(°) + HWnW + H ^ 2 ) + ... where W is a Zeeman coefficient, the first term is the energy of the nth level independent of field, the second term is the first-order Zeeman term (effect of the magnetic field on the nth level's energy) and then higher terms. The Curie law (equation 1.4) can be obtained from the Van Vleck equation in the situation where W n(°) = Wn( 2) = 0 (no second order effects) and W n (^ is the energy associated with the S = 1/2 Zeeman splitting. In the case where systems with orbital contributions are included, the levels that are produced as a result of spin-orbit coupling (J-states) are substituted into E n and very complicated equations that are a function of X and T are obtained. The second-order term in the Van Vleck equation leads to another phenomenon observed in paramagnetic systems, namely temperature independent paramagnetism (TIP). In this situation, the ground state is mixed with unpopulated excited states due to the action of the external magnetic field. This paramagnetism is additive over the expected magnetic susceptibility and is considered a correction. Importantly, this "temperature independence" refers to the fact that the magnetic susceptibility is independent of temperature, as opposed to the moment. Hence, TIP is of greater importance at high temperatures where its relative contribution increases. The magnitude of TIP is usually fairly small and can be accounted for in modelling of data using standard equations which depend on A, the crystal field splitting parameter. 14 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction The discussion to this point has assumed that, for first-row transition metals, essentially spin-only magnetism with a correction for orbital angular momentum contributions will yield an accurate description of the expected magnetic moment. However, the presence of spin-orbit coupling and the ligand field can affect the spin quantum number (S value) of the metal. It was stated that the degeneracy of a Kramers' doublet could only be removed by a magnetic field. However, in S > 1/2 systems, the spin quantum number S contains more than one Kramers' doublet. For example, an S = 3/2 system contains m s = ±1/2 and ±3/2 doublets. This fourfold degeneracy may be removed in the absence of a magnetic field by a non-cubic ligand field to yield zero-field splitting of the S = 3/2 system (Figure 1.5). (a) (b) Figure 1.5 (a) Zeeman sphtting of an S = 3/2 state (b) with zero-field splitting. The origin of this splitting is quantum-mechanical in nature; essentially each doublet interacts with the ligand field differently and splitting occurs. Hence, compounds with lower symmetry and more anisotropic ligand fields would be expected to show large zero-field splitting behaviour. Spin-orbit coupling is also a mechanism to introduce zero-field splitting of degenerate levels. The magnitude of the splitting is designated D and can be obtained from variable 15 References begin on page 28 Chapter I: Paramagnetic Organometallic Complexes - A General Introduction temperature magnetic susceptibility data by fitting the data to the Van Vleck treatment of the system. In this case the energy levels, separated by D, as well as their separate Zeeman splitting effects are factored into the Van Vleck equation to yield complex equations that are functions of D, g and T. Experimentally, zero-field splitting behaviour is characterized by a smooth drop in magnetic moment at low temperatures. Large zero-field splitting is common in tetrahedral Co(II) complexes. Finally, it should be noted that all of the above discussion is based on the assumption of a magnetically dilute compound, that is, only one isolated metal centre is being considered. Magnetically concentrated compounds, containing more than one paramagnetic centre, can also be treated using the Van Vleck equation and the appropriate energy levels. Interacting metal centres give rise to interesting magnetic behaviour such as antiferromagnetism, ferromagnetism and other phenomena.63'65 Recently, molecular magnetism has emerged as an important new frontier in magnetochemistry and is based on the interaction of paramagnetic fragments in an extended molecule.65'66 Fragments can include transition metal centres, lanthanides and paramagnetic ligands. 1.4 Characterization of Paramagnetic Organometallic Compounds62'67 In part, the study of paramagnetic organometallic complexes has been hampered by the difficulty of their characterization relative to diamagnetic systems.14,44 ,48 The standard tools of organometallic chemistry, in particular proton and multinuclear magnetic resonance spectroscopy, are much less applicable to paramagnetic systems. Although ESR spectroscopy is often considered a valuable tool in the study of paramagnetic systems, as will be seen, this is also of limited use. The challenge of characterization and the tools available to meet that challenge are outlined below. 16 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction (i) Nuclear magnetic resonance spectroscopy Undoubtedly the most vital tool used in the characterization of diamagnetic organometallic complexes is NMR spectroscopy.67 Unfortunately, in many cases, paramagnetic compounds show no observable NMR spectrum at all. If observable, the resonances of the spectrum are very broadened and highly shifted from their diamagnetic values. Coupling information is usually lost. A typical 1 H NMR spectrum of a paramagnetic species could have resonances in the range from +200 to -200 ppm! The broadening and shifting of resonances in paramagnetic NMR spectra have different origins. The broadening observed is due to fast relaxation of the protons. Fast relaxation broadens peaks due to the relationship of the excited state lifetime (x) with the uncertainty of the energy of transition (AE), given by:67 xAE - hlln [1.10] Hence, if x is short (i.e., fast relaxation), the uncertainty in the transition energy is large and broad lines are observed. If x is longer (i.e. slow relaxation), the uncertainty is small and sharp lines are observed (barring saturation effects at very long x). What causes relaxation? There are many mechanisms for relaxation of nuclei and most involve interactions between the nucleus and fluctuating magnetic fields in the sample.62'67 Interactions between the nucleus and its surroundings are termed "spin-lattice" relaxation and could include fluctuations due to tumbling and dipole-dipole interactions with other nuclei. Unpaired electron-nuclear dipole interactions are very large and this is the source of fast nuclear relaxation in paramagnetic compounds. The unpaired electrons can relax nuclei via either dipole-dipole interactions or through-bond coupling in some cases. As a result, *H NMR spectra are broadened or unobservable, depending on the extent of the relaxation and the size of x. A key point here is that the efficiency of this relaxation mechanism depends on the relaxation of the electron. It has been shown62 ,68 that rapidly relaxing electrons are less efficient at relaxing protons; electrons with long excited state lifetimes relax 17 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction protons well.69 Unfortunately, nuclear relaxation effects in paramagnetic compounds are generally too strong to allow for the observation of other nuclei such as phosphorus or carbon, especially in the usual cases where the phosphorus nucleus is bound directly to the metal centre. 2 H NMR spectroscopy has been shown to be a useful procedure for obtaining sharper paramagnetic spectra as the deuterium nucleus relaxes at a substantially slower rate than the proton, allowing for sharper lines. 6 8- 7 0- 7 5 The shifting of resonances two different sources: pseudo-contact and contact (Fermi) shifts.62'67'68 Pseudo-contact shifting is a through-space process whereby the magnetic field of the unpaired electron(s) alters the local field around the nucleus in question and causes a shift. The key point is that since this is a through-space effect, resonances due to protons closer to the metal centre through-space will be shifted more than ones more remote. This effect weakens rapidly with distance; equations describing pseudo-contact shifts include 1/r3 terms, where r is the distance between unpaired electron and shifting nucleus. Contact, or Fermi shifts are based on through-bond effects. These include either direct derealization of the unpaired electrons to the nuclei being observed and also spin-polarization, which is essentially indirect derealization of unpaired spin. The magnitude and direction of the contact shifts are hence potentially a measure of the unpaired electron density distribution throughout a compound and as a result much effort has gone into theoretical prediction and experimental determination of such shifts.76 This is not a trivial matter since only the isotropic shift is observed; the pseudo-contact component of the shift must be accounted for before the contact shift can be evaluated.62'68 Given this situation, how does one assign a paramagnetic *H NMR spectrum? Ideally, selective deuteration at every proton site would unambiguously assign the spectrum (each peak would disappear one by one) but this is usually impractical. Fortunately, integration of peaks to obtain the relative number of protons associated with each resonance is still valid, hence the integration ratio can be obtained. Sometimes that information is sufficient to assign some peaks although extremely broad peaks are difficult to accurately integrate. However, the inherent width 18 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction and isotropic shift of each resonance provides a clue as to its assignment. As stated above, resonances due to protons that are closer through-space to the metal centre will be generally be broader and more shifted than more remote protons' resonances. Using this, and knowing the structure of the complex in question, it is often possible to crudely predict which proton resonances will be broadest and correlate that with the observed spectrum. The mechanism of contact shifting can be useful as well, for example, in assigning phenyl-proton resonances. Due to electron spin derealization into phenyl rings (e.g. PPh2) the resonances due to the ortho and para-protons are often shifted in one direction and the meta-proton resonances in the other.62,68 Finally, variable temperature NMR spectroscopy is a useful technique to gain information, thanks to the Curie law, which says that the magnetic susceptibility varies inversely with temperature (section 1.3). For a molecule that obeys the Curie law, raising the temperature reduces the magnetic susceptibility and hence the lH NMR spectral peaks will sharpen and shift towards their diamagnetic values. Consequently, plots of 8 vs. 1/T are often used in paramagnetic NMR research; the y-intercept is the theoretical diamagnetic value of 5. As well, integration of peaks at higher temperature is often more accurate due to the increased sharpness of the lines. Using integration, width and variable temperature data in concert can often lead to a plausible assignment of a paramagnetic spectrum. Even without an assignment, however, an observable spectrum still can act as a fingerprint for the compound. (ii) Electron spin resonance spectroscopy ESR spectroscopy detects unpaired electrons and their coupling to nearby spin-active nuclei. Hence it would seem that ESR would be critically useful in the characterization of paramagnetic complexes and in fact for certain systems, notably Mn(II) and Cu(II) this is indeed the case. However, despite a naive view that "if it is paramagnetic it has an ESR spectrum", a 19 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction large number of paramagnetic complexes are in fact ESR-silent. There are two predominant reasons for this situation. The ESR signal is based on the excitation of an electron from the m s = -1/2 to m s = +1/2 energy level of the magnetically split S = 1/2 Kramers' doublet (Figure 1.6), in exact analogy to NMR. 6 7 For S = 1/2 systems the energy diagram is hence straightforward and theoretically an ESR signal should be observable for such systems. Figure 1.6 (a) Effect of a magnetic field on an S = 1/2 Kramers' doublet, (b) Effect of large zero-field sphtting and a magnetic field on an S = 3/2 system. Expected ESR transitions are shown as solid arrows. D is the zero-field splitting separation. However, in the case of S > 1/2 systems there are more doublets to be considered and zero-field splitting plays a key role in whether ESR spectra will be observed or not. Often, for multi-unpaired electron systems, zero-field sphtting reduces the degeneracy of the spin system. For odd-electron systems (S = 3/2, 5/2) the ground state will still be a degenerate Kramers' doublet, usually S = 1/2, and so an ESR spectrum can still be observed (Figure 1.6(b)).62'67 However, for even-electron systems (S = 1, 2) the ground state that results from zero-field splitting is S = 0, a non-20 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction degenerate state that cannot be split by a magnetic field (Figure 1.7). As a result, unless the zero-field splitting is relatively small and the S = 0 state is close to the S = 1 state (Figure 1.7(b)), an ESR spectrum will not be observed using a standard X-band ESR spectrometer since the transition required is usually outside the energy range of the spectrometer (Figure 1.7(c)). Examples where this situation prevails include high-spin Cr(II) (S = 2, d4) and high-spin Fe(II) (S = 2, d 6 ) . 6 2 , 6 7 Figure 1.7 Effect of a magnetic field on an S = 1 system with (a) no zero-field sphtting (b) small zero-field splitting (c) large zero-field splitting. ESR transitions are shown as solid arrows: (a) one transition; (b) two transitions; (c) no transitions (within spectrometer range). D is the zero-field sphtting separation. In addition to the zero-field splitting problem facing even-electron systems, all multi-unpaired electron systems have to deal with potentially fast electron relaxation. The principle is the same as for broadening of resonances in NMR as described above. Fast electronic relaxation causes a broadening or complete loss of signal in complexes. Important electronic relaxation mechanisms include interactions with other nuclei and also concentration effects such as intermolecular spin-spin interactions.62,67 Magnetic field anisotropy results in greater internal magnetic fluctuations and hence causes faster relaxation. This effectively means that compounds 21 References begin on page 28 Chapter I: Paramagnetic Organometallic Complexes - A General Introduction with lower symmetry will relax faster. As well, compounds with large zero-field splitting will relax faster than those without; the symmetry and zero-field splitting are of course related issues. As a result, metal centres with an odd number of unpaired electrons, for example S = 3/2 Co(II), may not show an observable ESR spectrum.62,67 Cooling of the sample to liquid nitrogen or liquid helium temperatures reduces the relaxation time and in some cases a signal can then be observed, but this is not always the case. Obviously, there are numerous complications to be considered if the nucleus in question is not a S = 1/2 centre; in many cases an ESR spectrum will not be observable. E S R NMR Small Z F S => Slow relaxing e" Fast relaxing 1 H • Y E S NO Large Z F S = > Fast relaxing e" —> Slow relaxing 1 H NO Y E S Figure 1.8 Factors affecting NMR vs. ESR spectrum observation. It should be obvious that the relaxation properties that favour ESR signal detection will oppose NMR signal detection. Recall that fast electronic relaxation is not efficient at relaxing protons, hence an NMR spectrum might be observable. Fast electronic relaxation, however, will preclude observation of an ESR spectrum. Compounds with large zero-field splitting (ZFS) will have rapidly relaxing electrons and hence an NMR but no ESR spectrum is expected. The two methods are generally complementary (Figure 1.8).62,67 22 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction (iii) Other techniques Compared with NMR spectroscopy, mass spectrometry, infra-red and UV-vis spectroscopy and elemental analysis are generally of limited use in characterizing diamagnetic organometallic complexes. From mass spectra, observation of an assignable molecular ion (M+) peak with a characteristic isotope pattern is perhaps the most useful information that can be gleaned. Fragments of the M + peak that have lost easily definable groups (CH3, halide, PPh2 etc.) can also be of aid. Samples, however, must be stable under the ionization conditions and be volatile enough to generate a signal. Transition-metal complexes are generally highly coloured and the colours observed depend on the ligands in a given complex, the oxidation state of the metal, and the geometry around the metal. Hence, UV-visible spectroscopy is a potentially useful probe of these factors. While a UV-vis spectrum of a complex will not yield information about its molecular composition, the oxidation state and spin-state of the central metal and perhaps the ligand geometry can sometimes be ascertained. Infrared spectroscopy is of limited use in the study of a number of organometallic systems due to the complexity of the spectrum. There are three notable exceptions to this limitation, namely the characterization of metal-carbonyl, nitrosyl and metal-hydride containing systems. The vCO, vNO and vM-H stretching frequencies generally lie in unobscured areas and the location of such bands provide valuable structural information. In the case of metal carbonyls, bridging vs. terminal metal-carbonyls can be distinguished. Metal-acyl systems, both r)1 andr|2, can be identified. Metal-hydride stretching frequencies also indicate whether the hydrides are terminal or bridging metals. Deuteration and examination of the shifted vM-D stretching frequency is a useful corroboration of the original assignment. Elemental (combustion) analysis of a new compound provides a useful clue as to its composition. Compared with diamagnetic organic or organometallic systems, for paramagnetic 23 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction organometallic complexes the elemental analysis could be among the only information available and is hence a vital characterization tool. (iv) Magnetic susceptibility measurements The measurement of the magnetic susceptibility and calculation of the magnetic moment (as described in section 1.3) provides important information as to the oxidation and spin-state of the compound. In the case of multinuclear complexes, a room-temperature magnetic moment in the absence of variable-temperature data is not particularly useful, but for simple mononuclear systems, where the Curie law is obeyed, a single room-temperature measurement may suffice. The spin and oxidation state may then be obtained from comparison of the calculated magnetic moment with the expected values for a given number of unpaired electrons. Note that this procedure assumes validity of the spin-only formula (see section 1.3). Measurements of the magnetic moment in both the solid-state and in solution are important. Solid-state susceptibilities of complexes may be obtained using VSM and SQUID magnetometers61,63,64 but these instruments require substantial effort to operate. In routine cases, a simple Gouy balance may be used to obtain the magnetic susceptibility. The Gouy balance operates on the principle that paramagnetic compounds are attracted to a magnetic field while diamagnetic ones are repelled. A tube packed with the compound in question is placed between the poles of an electromagnet and the difference in the weight of the tube with the magnet on and off yields a value of the magnetic susceptibility.61,63,64 Evans' Method7 7 , 7 8 is an NMR technique for the measurement of magnetic susceptibilities in solution and is based on the idea that the paramagnetic shift induced by a paramagnetic complex on a diamagnetic standard can be related to the magnetic susceptibility. Essentially, a sample containing a known concentration of the paramagnetic substance is prepared, and a diamagnetic standard is added. Into the sample is also inserted a sealed tube that contains only the diamagnetic 24 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction standard. Upon measurement of the spectrum, two peaks for the diamagnetic standard are observed: one is paramagnetically shifted and the other is not. The frequency difference between the two peaks can be related to the magnetic susceptibility of the complex by: 3 Av Xg = —— + Xo [ L U ] 4 TC v 0 m where Av is the shift difference in Hertz, vQ is the spectrometer frequency in Hertz and m is the mass in grams of compound in one mL of solvent. Xo 1 S the diamagnetic susceptibility of the solvent and is usually ignored for paramagnetic transition metal complexes. This method is widely used to determine magnetic susceptibilities of paramagnetic complexes in solution. (v) X-ray crystallography X-ray crystallography does not depend on whether unpaired electrons are present or not, merely on whether crystals of sufficient quality and size can be prepared. Despite the usefulness of the other techniques mentioned it is safe to say that without a crystal structure of a given complex its definitive characterization remains in doubt, particularly if the compound is the first of its type. Consequently, a ligand system that induces high-quality crystals is invaluable in the study of paramagnetic systems. The fact that there are over twenty crystal structures in this thesis reinforces the importance of this technique to the discipline. The structures are presented in most cases as Oak Ridge Thermal Ellipsoid Plot (ORTEP) diagrams, in which the ellipsoids represent a 33% probability distribution of electron density around the nucleus (unless otherwise noted). With the advent of new detector technology in CCD (charge coupled device) X-ray diffractometers and the consequent reduction in required crystal size, the study of paramagnetic organometallic complexes will surely become much less arduous. 25 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction 1.5 Chemistry of the Amidodiphosphine Ligand System: Scope of the Thesis This thesis examines the organometallic chemistry of the first-row transition metals chromium, iron and cobalt in the +2 and +3 oxidation states, using an amidodiphosphine ligand system. As stated earlier, the organometallic chemistry of these Werner-type oxidation-states is generally less well developed than the comparable low oxidation state chemistry. Mononuclear cobalt(II), iron(III) and chromium(III) complexes are paramagnetic under any circumstances (d7, d 5 and d3) and chromium(II) and iron(II) can be paramagnetic (d4 and d6) given the appropriate ligand environment. The ligand of choice for most of this thesis is a tridentate chelating mixed-donor ligand, -N(SiMe2CH2PPh2)2. This monoanionic amidodiphosphine ligand [PNP] has the advantage of straightforward synthesis from commercially available starting materials.79 Specifically, reaction of lithium diphenylphosphide with l,3-bis(chloromethyl)-l,l,3,3-tetramethyldisilazane gives HN(SiMe2CH2PPh2)2, which can be deprotonated with "BuLi to give the lithium salt of the ligand (Scheme 1.1). This can then be used in simple metathesis reactions with metal halide salts. M e 2 .Si M e 2 Si . CI H CI 1,3-bis(chloromethyl)-tetramethyldisilazane Scheme 1.1 2 L iPPh 2 THF P h 2 P M e 2 M e 2 - S i w S L I T I H PPrv 3 L iPR 2 T H F L n M X R 2 P "BuLi hexanes M e 2 M e 2 - S i w S L N I L i P R 2 R = Me, Ph, 'Pr, fBu 26 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction The amidodiphosphine ligand is also conducive to formation of highly crystalline complexes, an important requirement especially for paramagnetic systems. This is partially due to the presence of the diphenyl substituents on the phosphines; ligands with other substituents (Me, *'Pr, *Bu) can be prepared but the crystallinity of their complexes is much less general than the phenyl derivative. Table 1.2 Comparison of electronically similar paramagnetic vs. diamagnetic metal centres. Metal Centre (this thesis), dn, Spin State Diamagnetic Metal for Comparison (published)80 Cr(II) d 4 high spin Cr(III) d 3 high-spin Fe(II) d 6 high-spin Co(II) d 7 high-spin Co(II) d 7 low-spin Co(ffl) d 6 interm.-spin Ni(II), Pd(II), Pt(II),79 d 8 Rh(III), Ir(III)81-83 d 6 Ru(II),84 d6 No comparable system Ni(II), Pd(II), Pt(II)79 d 8 Rh(III), Ir(III)81-83 d 6 One more important point supporting the use of this ligand is undoubtedly the fact that the chemistry of this ligand on diamagnetic metal centres has been well explored in the Fryzuk laboratory.80 The ligand was originally designed as a hard/soft donor mismatch, in which the phosphines favour late transition metals and the amides early transition metals. Consequently, unusual reactivity patterns and patterns of stability were envisioned and realized in many cases.80 This bulk of knowledge with regard to -N(SiMe2CH2PPh2)2 coordinated to diamagnetic systems is invaluable as a reference point to the stability, structure and reactivity of paramagnetic complexes (Table 1.2). In particular, high-spin Cr(II) (d4) is comparable to Ni, Pd and Pt(II) d 8 systems;79 both are expected to be square planar, but whether the reactivity is similar or not remains to be 27 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction seen. Other examples of comparable systems include high-spin Cr(III) (d3) and Rh and Ir(III) (d6);81"83 the Cr(III) centre has three half-filled d-orbitals while the Rh and Ir(III) centres have three all-filled d-orbitals. A comparison of complexes in this manner could reveal interesting reactivity and stability trends. The thesis is organized into five chapters, excluding the general introduction. Chapter 2 introduces the chromium(II) and cobalt(II) halide and alkyl complexes stabilized by "N(SiMe2CH2PPh2)2- Chapter 3 explores the redox chemistry of the chromium(II) halides and alkyls with disulfides and alkyl halides (RX) and the five-coordinate Cr(III) complexes that result. Chapter 4 expands on the alkyl halide reaction, now with cobalt(II) halides and alkyls. Comparisons between chromium and cobalt are made. Chapter 5 examines iron(II) and iron(III) chemistry with •N(SiMe2CH2PPh2)2 and the final chapter extends the thesis to new directions involving novel ligand systems, different metal centres and exciting reactivities. 1.6 References (1) Huheey, J. E. Inorganic Chemistry: Principles of Structure and Reactivity; 3rd ed.; Harper & Row: New York, 1983, Ch. 9. (2) Douglas, B. E.; McDaniel, D. H.; Alexander, J. J. Concepts and Models in Inorganic Chemistry; 2nd ed.; John Wiley & Sons: USA, 1983, Ch. 7. (3) Figgis, B. N. In Comprehensive Coordination Chemistry; G. Wilkinson, R. D. Gillar and J. McCleverty, Eds.; Pergamon Press: Oxford, 1987; Vol. 1; Ch. 6. (4) Figgis, B. N.; Lewis, J. Prog. Inorg. Chem. 1964, 6, 37. (5) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles and Applications of Organotransition Metal Chemistry; University Science Books: Mill Valley, CA, 1987, Ch. 1. 28 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction (6) Kealy, T. J.; Pauson, P. J. Nature 1951,168, 1039. (7) Miller, S. A.; Tebboth, J. A.; Tremaine, J. F. / . Chem. Soc. 1952, 632. (8) Wilkinson, G.; Rosenblum, M.; Whiting, M. C.; Woodward, R. B. / . Am. Chem. Soc. 1952, 74, 2125. (9) Crabtree, R. Ff. The Organometallic Chemistry of the Transition Metals; 2nd ed.; John Wiley & Sons: USA, 1994. (10) Tolman, C. A. Chem. Soc. Rev. 1972,1, 337. (11) Comprehensive Organometallic Chemistry; Wilkinson, G.; Stone, F. G. A.; Abel, E. W., Eds.; Pergamon Press: Oxford, 1982. (12) Comprehensive Organometallic Chemistry II; Wilkinson, G.; Stone, F. G. A.; Abel, E. W., Eds.; Pergamon Press: Oxford, 1995. (13) Poli, R. Chem. Rev. 1996, 96, 2135. (14) Organometallic Radical Processes; Trogler, W. C , Ed.; Elsevier: Amsterdam, 1990; Vol. 22. (15) Theopold, K. H.; Heintz, R. A.; Noh, S. K.; Thomas, B. J. In Homogeneous Transition Metal Catalyzed Reactions; W. R. Moser and D. W. Slocum, Ed.; American Chemical Society: Washington, DC, 1992; pp 591. (16) Thomas, B. J.; Noh, S. K.; Schulte, G. K.; Sendlinger, S. C ; Theopold, K. H. / . Am. Chem. Soc. 1991,113, 893. (17) Bhandari, G.; Kim, Y.; McFarland, J. M.; Rheingold, A. L.; Theopold, K. H. Organometallics 1995,14, 738. (18) Clark, A. Catal. Rev. 1969, 3, 145. (19) Karol, F. J.; Karapinka, G. L.; Wu,C; Dow, A. W.; Johnson, R. N.; Carrick, W. L. / . Polym. Sci., Polym. Chem. Ed. 1973,10, 2621. (20) Karol, F. J.; Brown, G. L.; Davison, J. M. / . Polym. Sci., Polym. Chem. Ed. 1973,11, 413. (21) Baird, M. C. Chem. Rev. 1988, 88, 1217. 29 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction (22) Schut, D. M.; Keana, K. J.; Tyler, D. R.; Rieger, P. H. / . Am. Chem. Soc. 1995,117, 8939. (23) Tyler, D. R. Acc. Chem. Res. 1991, 24, 325. (24) Tyler, D. R.; Mao, F. Coord. Chem. Rev. 1990, 97, 119. (25) Astruc, D. Chem. Rev. 1988, 88, 1189. (26) Tyler, D. R. Prog. Inorg. Chem. 1988, 36, 125. (27) Baird, M. C. In Organometallic Radical Processes; W. C. Trogler, Ed.; Elsevier: Amsterdam, 1990; Vol. 22; pp 49. (28) Morrison, W. H.; Hendrickson, D. N. / . Chem. Phys. 1973, 59, 380. (29) Cowan, D. O.; Levana, C ; Park, J.; Kaufman, F. Acc. Chem. Res. 1973, 6, 1. (30) Wilkinson, G. / . Am. Chem. Soc. 1952, 74, 6148. (31) Pfab, W.; Fischer, E. O. Z. Anorg. Allg. Chem. 1953,274, 316. (32) McCullen, S. B.; Walker, H. W.; Brown, T. L. / . Am. Chem. Soc. 1982,104, 4007. (33) Cooley, N. A.; Watson, K. A.; Fortier, S.; Baird, M. C. Organometallics 1986,5, 2563. (34) Legzdins, P.; McNeil, W. S.; Shaw, M. J. Organometallics 1994,13, 562. (35) Legzdins, P.; McNeil, W. S.; Batchelor, R. J.; Einstein, F. W. B. J. Am. Chem. Soc. 1994,116, 6021. (36) Moelwyn-Hughes, J. T.; Garner, A. W. B.; Gordon, N. J. / . Organomet. Chem. 1970, 26, 373. (37) Singleton, E.; Moelwyn-Hughes, J. T.; Garner, A. W. B. / . Organomet. Chem. 1970, 21, 449. (38) Meyer, T. J.; Caspar, J. V. Chem. Rev. 1985, 85, 187. (39) Brown, T. L. In Organometallic Radical Processes; W. C. Trogler, Ed.; Elsevier: Amsterdam, 1990; Vol. 22; pp 67. (40) Scott, S. L.; Espenson, J. H.; Zhu, Z. / . Am. Chem. Soc. 1993,115, 1789. (41) Lee, K.-W.; Brown, T. L. / . Am. Chem. Soc. 1987,109, 3269. (42) Huber, T. A.; Macartney, D. H.; Baird, M. C. Organometallics 1995,14, 592. 30 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction (43) Gixolami, G. S.; Wilkinson, G.; Galas, A. M. R.; Thornton-Pett, M.; Hursthouse, M. B. /. Chem. Soc, Dalton Trans. 1985, 1339. (44) Hermes, A. R. Ph.D. Thesis, University of Illinois at Urbana-Champaign, 1988. (45) Hermes, A. R.; Morris, R. J.; Girolami, G. S. Organometallics 1988, 7, 2372. (46) Hermes, A. R.; Girolami, G. S. Organometallics 1987, 6, 763. (47) Hermes, A. R.; Girolami, G. S. Inorg. Chem. 1990, 29, 313. (48) Theopold, K. H. Acc. Chem. Res. 1990,25, 263. (49) Hao, S.; Song, J.-I.; Berno, P.; Gambarotta, S. Organometallics 1994,13, 1326. (50) Espenson, J. H. Acc. Chem. Res. 1992, 25, 222. (51) Davidson, P. J.; Lappert, M. F.; Pearce, R. Chem. Rev. 1976, 76, 219. (52) Davidson, P. J.; Lappert, M. F.; Pearce, R. Acc. Chem. Res. 1974, 7, 209. (53) Barker, G. K.; Lappert, M. F. / . Organomet. Chem. 1974, 76, C45. (54) Yagupsky, G.; Mowat, W.; Shortland, A.; Wilkinson, G. / . Chem. Soc, Chem. Commun. 1970, 1369. (55) Mowat, W.; Shortland, A.; Yagupsky, G.; Hill, N. J.; Yagupsky, M.; Wilkinson, G. / . Chem. Soc, Dalton Trans. 1972, 533. (56) Bower, B. K.; Tennent, H. G. / . Am. Chem. Soc. 1972, 94, 2512. (57) Byrne, E. K.; Theopold, K. H. / . Am. Chem. Soc 1989, 111, 3887. (58) Byrne, E. K.; Richeson, D. S.; Theopold, K. H. / . Chem. Soc, Chem. Commun. 1986, 1491. (59) Schrock, R. R.; Parshall, G. W. Chem. Rev. 1976, 76, 243. (60) Koschmieder, S. U.; Wilkinson, G. Polyhedron 1991,10, 135. (61) Earnshaw, A. Introduction to Magnetochemisry; Academic Press: New York, 1968. (62) Drago, R. S. Physical Methods for Chemists; 3rd ed.; W. B. Saunders: USA, 1996. (63) Carlin, R. L. Magnetochemistry; Springer-Verlag: Heidelberg, 1986. (64) O'Connor, C. J. Prog. Inorg. Chem. 1982, 29, 203. (65) Kahn, O. Molecular Magnetism; VCH: New York, 1993. 31 References begin on page 28 Chapter 1: Paramagnetic Organometallic Complexes - A General Introduction (66) Miller, J. S.; Epstein, A. J. Angew. Chem., Int. Ed. Engl. 1994, 33, 385. (67) Ebsworth, E. A. V.; Rankin, D. W. H.; Cradock, S. Structural Methods in Inorganic Chemistry; 2nd ed.; Blackwell Scientific Publications: Oxford, 1991. (68) Lamar, G. N.; Horrocks, W. D.; Holm, R. H. NMR of Paramagnetic Molecules; Academic Press: New York, 1973. (69) Murthy, N. N.; Karlin, K. D.; Bertini, I.; Luchinat, C. / . Am. Chem. Soc. 1997,119, 2156. (70) Johnson, A.; Everett, G. W. / . Am. Chem. Soc. 1972, 94, 1419. (71) Wheeler, W. D.; Kaizaki, S.; Legg, J. I. Inorg. Chem. 1982, 21, 3248. (72) Johnson, A. R.; Wanandi, P. W.; Cummins, C. C ; Davis, W. M. Organometallics 1994, 13, 2907. (73) Wanandi, P. W.; Davis, W. M.; Cummins, C. C. / . Am. Chem. Soc. 1995,117, 2110. (74) Laplaza, C. E.; Johnson, M. J. A.; Peters, J. C ; Odom, A. L.; Kim, E.; Cummins, C. C ; George, G. N.; Pickering, I. J. / . Am. Chem. Soc. 1996,118, 8623. (75) Hill, D. H.; Parvez, M. A.; Sen, A. / . Am. Chem. Soc. 1994,116, 2889. (76) McGarvey, B. R. Inorg. Chem. 1995, 34, 6000. (77) Evans, D. F. / . Chem. Soc. 1959, 2003. (78) Sur, S. K. / . Magn. Res. 1989, 82, 169. (79) Fryzuk, M. D.; MacNeil, P. A.; Rettig, S. J.; Secco, A. S.; Trotter, J. Organometallics 1982, 1, 918. (80) Fryzuk, M. D. Can. J. Chem. 1992, 70, 2839. (81) Fryzuk, M. D.; MacNeil, P. A.; Rettig, S. J. Organometallics 1986, 5, 2469. (82) Fryzuk, M. D.; MacNeil, P. A.; Ball, R. G. / . Am. Chem. Soc. 1986,108, 6414. (83) Fryzuk, M. D.; MacNeil, P. A.; Massey, R. L.; Ball, R. G. / . Organomet. Chem. 1989, 368, 231. (84) Fryzuk, M. D.; Montgomery, C. D.; Rettig, S. J. Organometallics 1991,10, 467. 32 References begin on page 28 Chapter Two Chromium(II) and Cobalt(II) Halide and Alkyl Complexes 2.1 Introduction One goal of this thesis is to bridge the gap between classical coordination chemistry and organometallic chemistry via the synthesis and examination of paramagnetic organometallic complexes.1 A good case in point involves complexes of chromium. Even though the coordination chemistry of this element encompasses oxidation states ranging from - 2 to +6, 2 a the most commonly studied derivatives are Werner-type complexes of Cr(II) and Cr(III) and most of these paramagnetic species display octahedral geometries. By contrast, the organometallic chemistry of chromium has concentrated on diamagnetic, formally Cr(O) carbonyl complexes, such as Fischer carbenes and arene half-sandwich derivatives.3,4 A series of quadruply bonded Cr(II) dimers has been well studied, but these too are essentially diamagnetic.23 ,5"7 One notable example of a paramagnetic chromium organometallic complex is the series of formally Cr(I) 33 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes radicals such as CpCr(C0)3 and CpCr(NO)X(L).8 ,9 This limited research into paramagnetic organometallic chromium complexes is surprising since chromium catalysts are used industrially in ethylene polymerization,10-12 where the active site is suspected to be a paramagnetic, highly coordinatively and electronically unsaturated chromium(III) centre.13,14 The situation for cobalt is similar in that most organometallic chemistry has utilized diamagnetic low-valent Co(0) and Co(I) systems, with classical coordination chemistry concentrating on higher-valent, paramagnetic Co(II) and Co(III) complexes. The discovery that Vitamin B12 contains a cobalt(III)-carbon bond focused organometallic chemists on Co(III)-C bond formation; this is the only naturally occurring organometallic compound. Cobalt(III) alkyl chemistry will be discussed in more detail in Chapter 4. In general, however, organometallic cobalt chemistry has focused on the use of carbonyl-containing species such as HCo(CO)4 and Co2(CO)8, which figure prominently in catalytic cycles involving H2, CO and alkenes, such as hydroformylation of olefins (oxo process) and the hydrogenation of CO to a variety of products. 2 1 3 , 1 5 In contrast, high-valent cobalt chemistry formed the basis of Werner's groundbreaking ideas in coordination chemistry.20 Again, with the exception of Vitamin B12-related model work, little organometallic chemistry of cobalt (II) and (III) systems has been examined. This chapter will present the organometallic chemistry of chromium(II) and cobalt(II) with an amidodiphosphine ancillary ligand and compare the chemistry of the two metals. First, the preparation of precursor ligand-metal halide complexes is described, then the conversion of these derivatives into species with metal-alkyl bonds is examined. The physical and chemical properties of the complexes are investigated and their reactivity explored. 34 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes 2.2 Synthesis, Structure and Reactivity of Chromium(II) and Cobalt(II) Halide Complexes (i) Synthesis and structure of {[(Ph2PCH2SiMe2)2N]Cr}2(u.-Cl)216 A convenient starting material for the synthesis of chromium(II) paramagnetic complexes is Crd2*THF.17 Addition of one equiv of LiN(SiMe2CH2PPh2)218 to a THF suspension of CrCl2*THF resulted in the rapid formation of a dark blue solution. Removal of the THF and recrystallization from toluene afforded light blue crystals of empirical formula CrCl[N(SiMe2CH2PPh2)2] (1) in high yield (equation 2.1). The reaction also proceeds when using an anhydrous CrCl2 suspension in THF but the yield is substantially lower; an alkane-soluble green byproduct, likely Cr[N(SiMeCH2PPh2)2]2 contaminates this latter reaction. The *H NMR spectrum of 1 shows only broadened resonances from which little structural information could be determined. In addition, all of the Cr(II) compounds reported here are ESR silent both in solution and in frozen glass. However, crystals of 1 were obtained by slow evaporation of a saturated toluene solution; the results of the solid-state structure determination by X-ray crystallography are shown in Figure 2.1 and selected bond lengths and angles in Table 2.1. THF CrCI 2 'THF + L iN(S iMe 2 CH 2 PPh 2 ) 2 {CrCI[N(SiMe 2CH 2PPh 2) 2]} x ^ 1 Structural analysis shows that chromium chloride 1 is dinuclear in the solid state, with a five-coordinate slightly distorted trigonal bipyramidal geometry around each chromium centre. Each trigonal bipyramid is defined by the amide nitrogen of the tridentate ligand and one of the 35 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Table 2.1 Selected bond lengths and angles in {[(Ph2PCH2SiMe2)2N]Cr}2(u-Cl)2 (1). Atom Atom Distance ( A ) Atom Atom Distance (A) Cr(l) P(D 2.529(4) Cr(l) Cr(l)* 3.64 Cr(l) P(2) 2.514(4) Cr(l) N 2.078(8) Cr(l) Cl(l) 2.397(3) N Si(l) 1.745(9) Cr(l) Cl(l)* 2.546(3) N Si(2) 1.629(9) Atom Atom Atom Angle (°) Atom Atom Atom Angle (°) P(l) Cr(l) P(2) 136.8(1) Cl(l) Cr(l) N(l) 179.2(3) Cl(l)* Cr(l) P(l) 107.4(1) Cr(l) Cl(l) Cr(l)* 94.6(1) Cl(l)* Cr(l) P(2) 115.4(1) Si(l) N Si(2) 123.0(5) Figure 2.1 Molecular structure (ORTEP) and numbering scheme for {[(Ph2PCH2SiMe2)2N]Cr>2(p-Cl)2 (1). Phenyl substituents on phosphorus have been removed for clarity. 36 References begin on page 100 Chapter 2: Cr(H) and Co(II) Halide and Alkyl Complexes bridging chloride ligands in the axial sites, and the two phosphine donors and the remaining bridging chloride in the equatorial positions. There is little axial distortion as is evident from the Cl(l)-Cr(l)-N(l) angle of 179.2(3)°. In the equatorial plane, however, the P(l)-Cr(l)-P(2), Cl(l)*-Cr(l)-P(l) and Cl(l)*-Cr(l)-P(2) angles of 136.8(1)°, 107.4(1)° and 115.4(1)°, respectively, do show substantial distortion from ideal trigonal planar geometry due to the large steric repulsion between PPh 2 donors. The Cr-Cr distance of 3.64 A is too large to invoke the existence of a metal-metal bond. 1 9 The Cr-Cl bond lengths of 2.397(3) and 2.546(3) A indicate that the Cr2(u.-Cl)2 bridging core is highly unsymmetrical, similar to that found for [(dippe)CrCl]2(|i-Cl)2 (dippe = 1,2-bis(diisopropylphosphino)ethane);20 the Cr-Cl-Cr angle in 1 is 94.6(1)°. Hence, the structure is best described as two loosely associated monomers. As a result it is not surprising to note that in the mass spectrum of 1 only a monomer peak at 615 mle is observed (as opposed to a dimer peak). Room temperature magnetic susceptibility measurements indicate a magnetic moment of 4.6 B.M. per chromium, consistent with an uncoupled high-spin d 4 system. As will be shown later, the two chromium centres are weakly antiferromagnetically coupled via chloride-mediated superexchange. The Cr-P bond lengths in 1 of 2.529(4) and 2.514(4) A are longer than the Cr-P bond lengths of 2.488(2) and 2.507(2) A reported for [(dippe)CrCl]2(u-Cl)2.20 The Cr-N bond length of 2.078(8) A is similar to that found in rraAW-Cr[N(SiMe3)2]2(THF)2 (2.089(12) A ) . 2 1 The N-Si bond lengths of 1.745(9) and 1.629(9) A indicate the degree of ligand distortion; the shorter of the two Si-N bond distances is the shortest reported, considerably shorter than in free disilazane (1.735(12) A ) 2 2 but comparable to that found in rrans-Cr[N(SiMe3)2]2(THF)2 (1.674(17) A ) . 2 1 37 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (ii) Reactivity of {[(Ph2PCH2SiMe2)2N]Cr}2(ji-Cl)2 (1) with donor ligands The extreme asymmetry of the metal-halide core in chromium chloride 1 suggested that the bridge could be easily cleaved by coordinating solvents or ligands. Consistent with this idea, 1 is only slightly soluble in non-coordinating solvents such as benzene or toluene, but is extremely soluble in THF, acetonitrile or pyridine. Addition of CO to a blue solution of 1 in toluene results in the rapid formation of a bright yellow adduct, which quickly reverts to its original blue upon removal of CO. A solution IR measurement of the yellow solution shows one assignable CO stretch at v c o = 1971 cnr 1, a value consistent with a simple, five coordinate CO adduct. Similarly, removal of THF or acetonitrile results in the recovery of blue, dimeric 1. [2.2] 38 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes In the case of pyridine, the resulting adduct is stable even with no excess pyridine present. Although crystals suitable for X-ray analysis could not be prepared, elemental analysis indicated that only one pyridine was bound, and therefore the formula CrCl(py)[N(SiMe2CH2PPh2)2] (l'py) is suggested. The formation of five-coordinate, labile monomers is consistent with both the solid-state structural data and the vastly increased solubility in coordinating solvents such as THF; this general reaction is typified in equation 2.2. Similar behaviour has been observed in the system [(dippe)CrCl]2(p-Cl)2 and the formation of the monomer Cr(NCCH3)Cl2(dippe) by reaction with acetonitrile.20 The adducts all have substantially different colors, depending on the binding ligand, and hence equilibrium constants for ligand binding could be determined by spectrophotometric titration. All UV-vis titrations showed single isosbestic points, implying that the five-coordinate species was a final product; no six-coordinate monomer was formed. Table 2.2 UV-vis spectral data, equilibrium constants (equation 2.2) and solution magnetic moments for {[(Ph2PCH2SiMe2)2N]Cr}2(p-Cl)2 (1) and its adducts with various ligands. Ligand Colour UV-Vis (nm) logK lieff (B.M.) - Purple 284, 342(sh), 4.8 522(w), 688(w) py Dark Green 290, 352(sh), 3 .1±0 .4 4.7 600(w, broad) CO Bright Yellow 294, 390 2.5 ± 0.3 2.7 THF Navy Blue 304,562(w) 1.0±0.1 -CH 3 CN Baby Blue 296 0.7 ± 0.05 4.7 39 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Table 2.2 shows the UV-vis data, equilibrium constants and adduct solution magnetic moments (measured by Evans' method) for a series of ligands. The equilibrium constant data shows that strong rj-donors such as pyridine have a much higher affinity (by two orders of magnitude) than the more weakly basic C H 3 C N and THF. Although there is a lack of comparable data in the literature, a more in-depth study of this equilibrium was not attempted. The magnetic moments observed follow what could be predicted from the ligand spectrochemical series; of the ligands tested, i.e CO is the strongest field ligand and thus it is able to force a low-spin configuration. (iii) Synthesis and structure of CoX[N(SiMe2CH2PPh2)2] (X = Cl, Br, I) Addition of one equiv of LiN(SiMe2CH2PPh2)2 to a THF suspension of CoX 2 (X = Cl, Br) results in the rapid formation of a dark blue solution. Removal of the THF and recrystallization from hexanes/toluene (95:5) yields blue crystals of CoX[N(SiMe2CH2PPh2)2] (X = Cl (2); X = Br (3)) in high yield (equation 2.3). P h 2 P THF, rt C o - ^ X [2.3] / X = CI 2 X = Br 3 P P h 2 40 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes A solution magnetic moment measurement (Evans' method)23,24 showed a value of u\eff = 4.2 B.M., consistent with a high-spin tetrahedral Co(II) complex with some orbital contribution via second-order spin-orbit coupling (d7 u , s o = 3.89 B.M.). 2 5 The dark, intense blue colour is also consistent with most high-spin tetrahedral Co(II) complexes.2'1,26,27 The visible spectrum of chloride 2 has three bands at 506, 602 and 786 nm; for the bromide these bands shift to 518, 616 and 792 nm (Table 2.3). Table 2.3 UV-vis spectral data for CoX[N(SiMe2CH2PPh2)2]. X UV-vis spectrum (nm, (e, M^cm-1)) Cl 506 (330), 602 (700), 786 (290) Br 518 (310), 616 (530), 792 (270) I 538 (230), 634 (560), 806 (290) Although paramagnetic, the lH NMR spectrum of chloride 2 shows two broad peaks at 15.0 and -5.4 ppm (the spectrum of bromide 3 is similar) with an integration ratio of 2:1. Obviously several peaks are too broad to be observed, and although the assignment of these two peaks as the meta and para protons on the phosphorus phenyl substituents may be correct (these protons are farthest from the metal centre and the integration is consistent) without a detailed labelling study any examination of the NMR spectrum for purposes other than fingerprinting the compound would be fairly speculative. Knowledge of the spectrum is, however, useful in compound identification. Other high-spin Co(II) complexes have been characterized by lH NMR spectroscopy.28 In order to obtain a representative crystal structure of the starting halide complexes, a solution of CoI[N(SiMe2CH2PPh2)2] (4) (preparation described in Chapter 4) in toluene was 41 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes allowed to evaporate to near-dryness, depositing crystals suitable for X-ray diffraction. The structure, shown in Figure 2.2, reveals the tetrahedral coordination sphere around cobalt; selected bond lengths and angles are in Table 2.4. The dihedral angle between the planes defined by P(l)-Co-N and P(2)-Co-I is 71.2° (90° is ideal), indicating the extent of the distortion from pure tetrahedral coordination. Examples of other tetrahedral high-spin Co(II) phosphine halide complexes (X = Cl, Br, I) include CoX 2(PR3)2 (R = Ph, C6Hn), 2 6 CoX2(dppe) and CoX2(dppp),27 and CoCl2(dippe),20 some of which have been structurally characterized. Reported magnetic moments range from 3.9-4.6 B.M. A typical high-spin Co(II)-P bond length is 2.384(1) A, in Co (PPh3 ) 2 X 2 ; 2 9 in trigonal Co[N(SiMe3) 2] 2PPh3 an unusually long Co-P bond length of 2.479(5) A is observed.30 These compare well with those observed in iodide 4, 2.348(3) and 2.373(3) A. Note that these Co-P bond distances are substantially shorter than the corresponding Cr-P distances in the chromium analogue 1; this is partly a reflection of the smaller size of Co(II) relative to Cr(II), although steric considerations (dimer vs. monomer) could play a part. Effective ionic radii for octahedral Cr(II) vs. Co(II) (high spin) centres are 0.94 vs 0.885 A . 3 1 - 3 2 Due to the different coordination number and geometry of these two systems a detailed comparison is probably not worthwhile. Cobalt(II) silylamides that have been structurally characterized include trigonal planar Co[N(SiMe3) 2] 2PPh 3, 3 0 [Co2(N(SiMe3)2)4]3 3-3 4 and two coordinate Co[N(SiMePh2)2]2 3 5. All are high-spin complexes and have Co-N bond lengths ranging from 1.898(3) and 1.904(3) A in Co[N(SiMePh2)2]2 3 5 to 1.931(14) and 1.924(13) A in Co[N(SiMe3)2]2PPh3.30 The Co-N bond length of 1.936(6) A in iodide 4 is comparable to these values; the N-Si bond lengths of 1.717(7) and 1.704(7) A are similar to those observed in the other systems mentioned. The Co-I distance of 2.571(1) A is unremarkable. 42 References begin on page 100 Chapter 2: Cr(II) and Co(H) Halide and Alkyl Complexes Table 2.4 Selected bond lengths and angles in CoI[N(SiMe2CH2PPh2)2] (4). Atom Atom Distance (A) Atom Atom Distance (A) Co P(l) 2.348(3) Co N 1.936(6) Co P(2) 2.373(3) N Si(l) 1.717(7) Co I 2.571(1) N Si(2) 1.704(7) Atom Atom Atom Angle (1 Atom Atom Atom Angle (") N Co P(2) 88.5(2) P(l) Co P(2) 120.26(9) N Co I 125.7(2) P(l) Co I 116.40(7) P(2) Co I 108.14(7) P(D Co N 95.6(2) Co N Si(l) 114.3(4) Si(l) N Si(2) 132.5(4) C(3) Figure 2.2 Molecular structure (ORTEP, 50% ellipsoids) and numbering scheme for CoI[N(SiMe 2CH 2PPh2)2] (4). 43 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Upon comparison of the precursor ligated metal halide complexes for Cr(II) and Co(II) some geometric differences can be observed. The Cr(II) halides prefer pentacoordination, obtained either by dimerization or base coordination, while the Co(II) system is tetrahedral; no evidence for dimerization is observed. The higher coordination number observed in the chromium(II) system is likely a reflection of the extreme electronic unsaturation of the chromium system relative to the cobalt system. Both types of complexes are high-spin, that is, have the maximum possible number of unpaired electrons. (iv) Reactivity of CoCl[N(SiMe2CH2PPh2)2] (2) with donor ligands The reaction of tetrahedral cobalt chloride 2 with donor ligands is much less interesting than that of the chromium analogue. Addition of pyridine, THF and acetonitrile to cobalt chloride 2 results in essentially no change in the UV-vis spectra, indicating little or no binding is occurring. Only in the case of CO addition does binding occur, but there is certainly no major colour change; there is no spin-state change like that observed in the chromium(II) halide system. A vco stretch at 1981 cnr 1 was observed in the solution IR spectrum. This binding is also reversible; if the CO atmosphere is removed the adduct slowly releases CO over a matter of hours to regenerate the original cobalt chloride complex 2. 2.3 Synthesis and Characterization of Chromium(II) and Cobalt(II) Alkyl Complexes (i) Synthesis and structure of CrMe[N(SiMe2CH2PPh2)2] (5) 3 6 The dimer {[(Ph2PCH2SiMe2)2N]Cr}2(p:-Cl)2 (1) reacts with MeMgBr or MeLi in THF at -78 °C to yield red-brown CrMe[N(SiMe2CH2PPh2)2] (5) in high yield, by alkali metal halide 44 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes salt metathesis (equation 2.4). Solid-state Gouy and solution Evans' method2 3 , 2 4 magnetic susceptibility measurements of the methyl complex 5 were consistent with a high-spin d 4 complex,25 indicating a monomelic compound. This derivative is soluble in aromatic solvents and THF but only slightly soluble in hexanes. X-ray quality crystals of 5 could be grown by cooling a saturated hexanes/toluene solution. P h 2 •P Cr C H 3 -P P h 2 5 The structural analysis confirms that 5 is monomeric and square planar in the solid state (Figure 2.3), this being the preferred geometry for four-coordinate Cr(II).2a Hence a geometry change from trigonal bipyramidal to square planar has occurred upon methylation, although the spin state remains unchanged. What is particularly unique about this twelve-electron species is that it contains the stoically undemanding methyl ligand; this is in stark contrast to other reported four coordinate neutral organochromium(II) complexes that contain bulky alkyl or aryl ligands such as CH2CMe3 or mesityl, respectively.21,37"41 Anionic, square planar, formally 14-electron Cr(II) alkyls are known. For example, the Cr(II) anion in the complex [Me4Cr][]Li(TMEDA)2] appears to be stabilized by the associated lithium counterions;42 a similar system with bulky neosilyl groups has also been structurally characterized 4 3 The few other Cr(II) alkyls are dimeric with "supershort" Cr-Cr bonds (e.g. Li2Cr2Me8»4THF, [(Me2PCH2)4]Cr2)44"46 and/or { [ (Ph 2 PCH 2 SiMe 2 ) 2 N]Cr} 2 (Li -CI ) 2 1 M'R M e 2 S i \ : N --78 °C / M e 2 S K M'R = MeLi, MeMgBr 4 5 References begin on page 100 Chapter 2: Cr(II) and Co(H) Halide and Alkyl Complexes bridging alkyl or aryl groups.4 7-4 8 Metal-metal bonding also provides the stability in the series of chromium alkyls [Cp*Cr(u,-R)]2.4 9 A n octahedral Cr(II) alkyl, CrMe 2 (dmpe) 2 . 5 0 , 5 1 , has also been reported. In the methyl complex 5 there is some distortion from pure square planar geometry (selected bond lengths and angles in Table 2.5); the P(l)-Cr-P(2) and N-Cr-C(31) angles of 154.77(4)° and 174.8(1)° respectively (180° is ideal) indicate the extent of the distortion. In addition, the dihedral angle between the planes defined by Cr-P(2)-C(31) and Cr-N-P(2) is 23.9°, and this represents the extent of the twist in what should ideally be a flat square planar core. The Cr-P distances of 2.485(1) and 2.451(1) A and the Cr-C distance of 2.151(3) A are both comparable with known square-planar neutral high-spin d 4 Cr(II) a lkyls . 2 0 , 3 7 Anionic Cr(II)-C bonds are approximately 0.1 A longer, as would be expected. 4 2 , 4 3 The Cr-N bond length of 2.117(3) A is extremely l o n g 2 1 , 4 0 and this could indicate that there is very little, if any, amide lone pair - metal prc-drc interaction. Figure 2.3 Molecular structure (ORTEP) and numbering scheme for CrMe[N(SiMe 2 CH 2 PPh 2)2] (5). 46 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes The electron density difference map revealed the Cr-methyl hydrogens to be one to one disordered, and in a normal geometry. In addition, there is no IR evidence for any Cr-C-H bond agostic interaction.52 This lack of such an interaction in neutral Cr(II) alkyls, despite the coordinative and electronic unsaturation, has been previously noted as being due to the unavailability of an appropriate empty orbital to interact with the C-H bond electron density,37 although a static agostic interaction was observed in [Cr2(CH2SiMe3)6]2~ by *H NMR spectroscopy.43 (ii) Synthesis and structure of CoMe[N(SiMe2CH2PPh2)2] (6) In a similar manner to the synthesis of CrMe[N(SiMe2C H2PPh2)2] (5), CoCl[N(SiMe2CH2PPh2)2] (2) reacts with MeLi or MeMgBr in THF or toluene to form the desired CoMe[N(SiMe2CH2PPh2)2] (6) complex, which is a bright yellow colour, in stark contrast to the intense blue precursor 2. Hence, although the methylation can be done in THF at low temperatures (-78 °C), it is easier to simply titrate 2 in toluene at room temperature with alkyl reagent until the blue colour has changed to yellow-orange (equation 2.5). This option is not available for the chromium system. CoMe[N(SiMe2CH2PPh2)2] (6) is fully soluble in aromatic solvents and moderately soluble in hexanes. P h 2 P h 2 Toluene C o — C I + M'R : N Co C H 3 [2.5] / rt P h 2 P h 2 Tetrahedral, high spin Square planar, low spin 2 M'R = MeLi, MeMgBr 6 47 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Table 2.5 Selected bond lengths and angles in MCH3[N(SiMe2CH2PPh2)2] (M= Cr, Co). Atom Atom Distance (A) M = Cr(5) M = Co (6) M P(l) 2.485(1) 2.191(2) M P(2) 2.451(1) 2.201(2) M N. 2.117(3) 1.954(6) M C(31) 2.151(3) 2.031(7) N Si(l) 1.697(3) 1.711(6) N Si(2) 1.699(3) 1.716(6) Atom Atom Atom Angle O M = Cr (5) M = Co (6) P(l) M P(2) 154.77(4) 169.64(8) N M C(31) 174.8(1) 178.4(3) P(l) M N 88.13(7) 91.4(2) N M P(2) 83.82(7) 92.2(2) P(2) M C(31) 95.1(1) 88.9(2) P(l) M C(31) 94.87(9) 87.7(2) Si(l) N Si(2) 121.9(2) 123.0(3) The colour change is an indication of the geometry and spin-state change that occurs upon halide for alkyl metathesis. The solution room temperature magnetic moment for 6 is 2.2 B.M., consistent with a low-spin d 7 centre with some second-order spin-orbit coupling.25 This indicates a geometry change from tetrahedral to square planar has occurred, in contrast to the chromium system where no spin state change was observed upon alkylation. 48 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes X-ray quality crystals of CoMe[N(SiMe2CH2PPh2)2] (6) were grown by slow evaporation of a hexanes solution. The result is shown in Figure 2.4; selected bond lengths and angles are in Table 2.5 beside the data for the analogous chromium system. The structural analysis confirms that 6 is monomeric and square planar; this complex is nearly isostructural to the chromium analogue already prepared. There is less distortion in the cobalt complex than in the chromium case. The Co(II) coordination sphere is reasonably flat; the dihedral angle between the planes defined by Co-P(2)-N and Co-P(l)-C(31) is only 9.8° in 6, a slight twist compared to the 23.9° observed in chromium methyl 5. Structurally characterized examples of Co(II) alkyls are uncommon.53 Five coordinate CoMe2(PMe3)3, [CoMe(OMe)(PMe3)2]2> and octahedral CoMe(acac)(PMe3)2 are known54 but no X-ray structures have been reported. Structurally characterized compounds with Co(II)-C bonds include the cobalt(II) aryl complexes Co(PEt2Ph)2(mesityl)255"57 and bis(p-mesityl)dimesityldicobalt(EI)58 which are low-spin square planar and high-spin trigonal respectively, with Co-C(sp2) distances of 1.994(3) and 1.988(7) A . A series of bulky cobalt alkyls stabilized by TMEDA have also been structurally characterized.59 These are high-spin and tetrahedral, with Co-C distances ranging from 2.025(7) - 2.151(8) A . In comparison, the Co-C bond length of 2.031(7) A in low-spin, square planar methyl 6 is shorter than in the high-spin alkyls but not as short as for the low-spin, stronger aryl bond. This difference is rationalized by the smaller size of a low-spin Co(II) centre vs. the high-spin case. For octahedral Co(II), effective ionic (crystal) radii of 0.79 and 0.885 A for low and high-spin Co(II) respectively are reported.31,32 Metal-phosphorus bond lengths are particularly sensitive to spin state; this is partly a reflection of the metal's size differences in different spin states. In the case of Co(II), a typical high-spin Co(II)-P bond length is 2.373(3) A in iodide 4 and 2.384(1) A in Co(PPh3)2X2;29. Low-spin Co(II)-P bond lengths are substantially shorter, as observed in Co(PEt2Ph)2(mesityl)2 (2.232(4) A ) 5 6 , 5 7 and CoMe[N(SiMe2CH2PPh2)2] (6) (2.191(2), 2.201(2) A ) . 49 References begin on page 100 Chapter 2: Cr(II) and Co(H) Halide and Alkyl Complexes C I 1 5 ) Figure 2.4 Molecular structure (ORTEP, 50% ellipsoids) and numbering scheme for CoMe[N(SiMe 2CH 2PPh 2)2] (6). Known Co(II) silylamide complexes are all high-spin complexes and have Co-N bond lengths ranging from 1.898(3) and 1.904(3) A in Co[N(SiMePh 2 ) 2 ] 2 3 5 to 1.931(14) and 1.924(13) A in Co[N(SiMe 3) 2] 2PPh3 3 0 In methyl 6 the Co-N bond length of 1.954(6) A is much shorter than the 2.117(3) A observed in chromium methyl 5. This could be due to the difference in size between high-spin Cr(II) (octahedral ionic radius 0.94 A) and low-spin Co(II) (octahedral ionic radius 0.79 A) centres (see Table 2.5) . 3 1 , 3 2 (iii) Paramagnetic lH NMR spectrum of CoMe[N(SiMe2CH2PPh2)2] (6) Despite the fact that Co(II) alkyls are paramagnetic, lH N M R spectra can be observed.60 In the absence of a detailed labelling study definitive assignments are challenging but integration ratios, peak shifts with variable temperature and comparison of a series of similar complexes allows for some assignments to be made.61 In addition, the A H N M R spectrum of a given 50 References begin on page 100 Chapter 2: Cr(H) and Co(II) Halide and Alkyl Complexes compound serves as an excellent fingerprint for its identification in reactivity studies. Note that 3 1 P NMR spectra are not observed for any paramagnetic complex in this thesis. Figure 2.5 *H NMR spectrum of CoMe[N(SiMe2CH2PPh2)2] (6) in C 6D 6(*). The iff NMR spectrum of CoMe[N(SiMe2CH2PPh2)2] (6), shown in Figure 2.5, has five peaks present; the C0-C//3 peak is not observed. The assignments are based on integration (60 °C spectrum, CgDg), the direction of shifting peaks upon increased temperature (peaks shift towards their diamagnetic values) and the inherent broadness of each peak (resonances due to protons closer to the metal are broader). The SiMe2 peak at -3.3 ppm is easily identified by its unique integration (12H). Although the resonances due to the phenyl ortho and meta protons have the same integration (8H), the ortho proton resonance (closer to the metal centre) at 9.8 ppm is very broad (340 Hz width at half-height), much broader than the meta proton resonance at 6.5 ppm (40 Hz width at half-height). Similarly the para proton and the backbone methylene resonances (both integrate to 4H) can be distinguished by the fact that the para proton peak at 8.4 ppm is sharp while the much closer methylene proton peak is observed at -6.3 ppm (260 Hz width at half-height). 51 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (iv) Synthesis and structure of M(CH2Ph)[N(SiMe2CH2PPh2)2] (M = Cr (7), Co (8)) Reaction of the chromium chloride dimer 1 with KCH2Ph or Mg(CH2Ph)2*2THF (Scheme 2.1) yields the purple compound Cr(CH2Ph)[N(SiMe2CH2PPh2)2l (7) which has a magnetic moment of 5.1 B .M. , 3 6 consistent with a monomeric, high-spin d 4 complex.25 The analogous cobalt(II) benzyl complex, Co(CH2Ph)[N(SiMe2CH2PPh2)2] (8) was prepared in a similar manner. Complex 8 is orange in colour, with a magnetic moment of 2.1 B.M., consistent with a low-spin Co(II) centre. Crystals of 7 and 8 suitable for X-ray analysis were grown by slow evaporation of saturated hexanes solutions. Scheme 2.1 M = Cr (1) Me 2 S i .pPh* 7 • I - - - \ ^ C H ; M C I [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] + M'R M'R= K C H 2 P h M e 2 S i ( Mg(dH 2 Ph) 2 .2THF M = C o (2) : \ M e 2 S ; / The X-ray structures reveal distorted square-planar, monomeric metal centres for both the chromium (7) and cobalt (8) benzyl compounds (Figure 2.6); this distortion from square planar is more pronounced than in the methyl complexes (Table 2.5 and 2.6). In particular, the chromium benzyl complex 7 appears highly distorted. The P(l)-Cr-P(2) and N-Cr-C(31) angles of 145.18(3) and 157.8(1)° show greater "bending-back" than in chromium methyl 5. The dihedral 52 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes angle between the Cr-P(l)-C(31) and Cr-P(2)-N planes is 38.7° (vs. 23.9° in 5), indicating again, the increased twist in chromium benzyl 7 from an ideal flat core compared to the chromium methyl complex 5. The Cr-P distances of 2.4758(9) and 2.4808(8) A are similar to 5 and to other known square-planar Cr(II) alkyls.37 The Cr-N bond of 2.067(2) A, shorter than in 5, coupled with comparatively slightly longer Si-N bonds of 1.707(2) and 1.706(2) A suggest that the lone-pair on nitrogen is interacting with the metal to a slightly greater extent than in the methyl complex 5. The cobalt benzyl complex 8, on the other hand, is much more similar to its methyl analogue (6) than in the chromium case. There is still more distortion from square planar in the cobalt benzyl vs. the cobalt methyl complex as shown by the P(l)-Co-P(2) and N-Co-C(31) angles of 162.93(3)° and 173.6(1)° (closer to ideal 180° in 8) and the dihedral angle between the planes defined by Co-P(2)-N and Co-P(l)-C(31) of 14.3° (9.8° in methyl 8). Figure 2.6 Molecular structure (ORTEP) and numbering scheme for M(CH2Ph)[N(SiMe2CH2PPh2)2] (M= Cr (7), Co (8)). 53 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Table 2.6 Selected bond lengths and angles in M(CH2Ph)[N(SiMe2CH2PPh2)2] (M= Cr, Co). Atom Atom Distance (A) M = Cr (7) M = Co (8) M P(l) 2.4758(9) 2.2127(8) M P(2) 2.4808(8) 2.2162(8) M N 2.067(2) 1.982(2) M C(31) 2.132(3) 2.009(3) N Si(D 1.707(2) 1.702(2) N Si(2) 1.706(2) 1.707(2) M C(32)ipSO 2.576(3) >3 M C(33)ortho 3.093(3), 3.364(3) -C(31) C(32)ipSO 1.456(4) 1.480(4) C(33) C(34) 1.380(5) 1.383(5) C(34) C(35) 1.366(6) 1.359(6) C(35) C(36) 1.359(5) 1.361(7) C(36) C(37) 1.379(4) 1.410(7) C(32)ipSO C(33)ortho 1.402(4) 1.378(5) C(32)iDSO C(37)orth0 1.394(4) 1.393(4) Atom Atom Atom Angle O M = Cr (5) M = Co (6) P(l) M P(2) 145.18(3) 162.93(3) N M C(31) 157.8(1) 173.6(1) P(l) M N 84.31(6) 82.83(6) P(2) M N 87.96(6) 86.98(6) P(2) M C(31) 99.42(1) 95.55(9) P(l) M C(31) 100.4(1) 96.01(9) Si(l) N Si(2) 122.3(1) 126.3(1) M C(31) C(32)jpSO 89.7(2) 100.1(2) M C(31) H(37) 109(1) 111.8 (idealized) M C(31) H(38) 116(1) 111.8 (idealized) 54 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes The increased distortion from ideal square planar geometry could be due to the steric bulk of the benzyl moiety as compared to the methyl ligand and this is likely the case for the cobalt benzyl complex. However, the large distortion in the chromium case has another source: the weak interaction of the benzyl ip so-carbon with the metal center. This type of interaction has been represented in the literature62 as an T]2-bound benzyl, and is fundamentally distinct from either an r^-a-bonded benzyl or an rj^ -benzyl complex. Several structural features have been highlighted as key in characterizing rj2-benzyls62,63 and these have, for the most part, been recognized in the chromium benzyl complex 7. Firstly, the Cr-QpSo distance of 2.576(3) A is short enough to consider the presence of a weak metal-carbon bond. On the other hand, the Cr-C0rtho distances of 3.093(3) and 3.364(3) A are clearly too long to consider any interaction as would exist in an r)3-bonding mode. The C r - C H 2 - Q p s o angle of 89.7(2)° indicates the extent of the pull of the ipso-carbon towards the metal, resulting in an angle much more acute than the expected tetrahedral 109°. The CH2 -Qp S 0 distance of 1.456(4) A is comparable to the values found in other r|2-benzyls62'64 and this length is shorter than in r^-benzyls. Four of the C-C bonds in the phenyl ring of the benzyl ligand are similar in length: 1.380(5), 1.366(6), 1.359(5) and 1.379(4) A , indicating a delocalized system consistent with rj2-bonding, as shown below: 55 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes An r)3-bound benzyl group normally forces localization of the bonding in the phenyl ring and this is not observed. Also consistent with r|2-coordination, the Qp S O-C o r tho bond lengths are elongated relative to the rest of the benzyl ring, being 1.402(4) and 1.394(4) A . In fact, one could almost consider this complex to be pentacoordinate. In terms of total electron count, chromium benzyl 7 can be considered as a fourteen electron system if one takes into account the ri2-benzyl formulation. The possibility of an agostic interaction between the benzyl CH2 and the chromium centre can be discounted. The methylene hydrogens were located and refined and the Cr-C(31)-H angles of 109(1)° and 116(1)° clearly show no distortion characteristic of agostic interactions.52 The Cr-C(31) bond length of 2.132(3) A is shorter than that found in methyl 5. This length compares well with other short Cr-C bonds recorded, for example, 2.131(2) A for Cr2(CH2SiMe3)4(PMe3)2 and 2.128(4) A for Cr(CH2SiMe3)2(dippe).37-65 A comparison of the structural features of the rj2-benzyl in 7 with other chromium benzyl complexes reveals the predominance of the r\1-binding mode. The formation of the Cr(II) bis (benzyl) complex Cr(CH2Ph)2(TMEDA) has been reported;38 the solid state structure clearly shows r^-benzyl type ligands with Cr-C bond lengths of 2.177(2) A and Cr-CH 2 -C i p s o bond angles of 111.9(1)°. The Cr(II) benzyl complex Cp*Cr(CH2Ph)bipy also shows simple T J 1 coordination (Cr-CH 2-Qp S 0 bond angle is 114.3(4)°). 6 6 Chromium(III) benzyls are certainly known as compounds such as Cr(CH2Ph)Cl2(THF), Cr(CH2Ph)Cl2(py)3 and [Cp*Cr(CH2Ph)Cl]2 illustrate;67'70 all contain r| l-benzyl groups. Especially well studied are the [Cr(CH2Ph)(OH2)5]2+ cations71 but the only crystal structure of such a cation is that of [p-BrC6H4CF£2CrL(H20)](C104)2 (L = [15]aneN4).72 Here the sterically enlarged Cr-CH 2 -Qp S 0 angle of 123°, slightly longer Cr-CH 2 bond length of 2.14(2) A and longer C H 2 - Q p s o bond length of 1.48(2) A all are indicative of a standard ri 1-benzyl. There is one example of a chromium-based T|2-benzyl complex, namely Cr(NR)2(CH2Ph)2, a Cr(VI) complex 7 3 This compound contains both an r| 2 and an t]1 benzyl ligand. The C r - C H 2 - C i p s o bond angles are 82.16(11) and 114.73(13)° for the T ] 2 and rj 1 benzyls 56 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes respectively. The r | 2 -benzyl is further structurally characterized by a C r - C i p s o bond length of 2.357(2) A, a CH 2 -C ipso bond distance of 1.443(3) A and elongated CipSO-Cortho distances of 1.409 A (vs. 1.395(3) A (average) in the rest of the ring). These data suggest an even greater interaction with chromium in this r | 2 -benzy l than that observed in 7; this is l ikely due to the extremely high-valent Cr(VI) centre in Cr(NR)2(CH2Ph)2 vs. the high spin Cr(II) centre in benzyl 7. Comparison with known Mo, W and Zr T] 2-benzyls shows that the structural features are similar, however, a meaningful, quantitative comparison may be difficult to m a k e . 6 2 ' 6 4 ' 7 4 Complex 7 appears to be the first example of a paramagnetic r\ 2 -benzyl complex. The cobalt benzyl complex 8, which contains the same ligand set and gross geometry as the chromium benzyl 7, does not show any evidence of an r j 2 -benzyl interaction. The electron deficient chromium complex (12 electron system) binds the benzyl group in an ri 2-fashion in an attempt to satisfy the metal's electronic unsaturation. O n the other hand, the more electron-rich 15-electron cobalt system does not require the extra electron density for stabilization and hence remains bound in an r^-fashion, as illustrated by the Co-C(31)-C(32) bond angle of 100.1(2)° and C o - Q p S O distance of over 3 A. Note that cobalt methyl 6 and benzyl 8 are the only examples of structurally characterized low-spin square planar Co(II) silylamide complexes. (v) Synthesis of M(CH2SiMe3)[N(SiMe2CH2PPh2)2] (M = Cr (9), Co (10)) The reaction of L i C H 2 S i M e 3 (neosilyl lithium) with chromium chloride 1 proceeds (equation 2.6) to generate the expected compound Cr(CH2SiMe3)[N(SiMe2CH2PPh2)2] (9), however, it was not isolated as a solid, but rather as a purple o i l and thus was not characterized fully. One feature of 9 that was noted in passing was the temperature dependence of its colour. A t room temperature 9 is purple in toluene but becomes brown at -78 °C; this is fully reversible. This compound was used in situ for a variety of redox reactions. 57 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Ph 2 P LiCH 2SiMe 3 M e 2 S i \ MCI[N(SiMe2CH2PPh2)2] M CH 2 SiMe 3 [ 2 . 6 ] -78 °C Me2Si / P Ph 2 M = Cr 9 M = Co 10 On the other hand the cobalt derivative, Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] ( 1 0 ) can be isolated by crystallization from hexanes. This cobalt(II) alkyl is also yellow and low-spin (u,eff = 2.0 B.M.) and is likely identical in structure to 6 and 8 , the other cobalt(II) alkyls. The lH NMR spectrum of 1 0 shows four peaks only; the backbone methylene and the neosilyl-methylene protons are probably too close to the metal centre to be observed. (vi) Synthesis of M(ri 5-C5H 5)[N(SiMe2CH 2PPh2)2] and structure of the Cr(II) derivative The addition of NaCp^DME to the starting chromium chloride 1 gives the cyclopentadienyl compound Cr(T|5-C5H5)[N(SiMe2CH2PPh2)2] ( 1 1 ) which has a magnetic moment of 2.7 B.M., consistent with a low-spin configuration (equation 2.7). The same compound can also be synthesized by reaction of the methyl complex 2 in toluene with freshly cracked cyclopentadiene, forming 1 1 with concomitant loss of CH4 (equation 2.8).36 58 References begin on page 100 Chapter 2: Cr(H) and Co(II) Halide and Alkyl Complexes r M e 2 S i MeoSi HIGH spin 6A 1 T H F , -78 °C 2 N a C p - D M E 2 P h 2 P h 2 P LOW spin d' 11 [2.7] The magnetic moment of 11 is comparable to the similar compound Cp*CrMe(dmpe), which has a reported magnetic moment of 2.79 B . M . 1 4 This is an example of a spin state change upon alkylation of the Cr(II) halide precursor. The coordination of the Cp ligand was deduced to be T ] 5 , as an r^-Cp would be expected to be high-spin, as found for complexes 5 and 7, the other Cr(II) alkyls. The analogous cobalt complex, Co(C5H5)[N(SiMe2CH2PPh2)2] (12) can be made in similar fashion and has been characterized by elemental analysis and NMR spectroscopy (see below). The hapticity of the Cp in this compound is unknown, although it is also low spin, withpeff = 1.9 B.M. 59 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes With regard to the chromium cyclopentadienyl compound 11, the preference for a low-spin environment can be rationalized by invoking a comparison with Cp2Cr, chromocene, which is also a 16-electron, low-spin d 4 system. The tridentate ancillary ligand, -N(SiMe2CH2PPh2)2, formally an anionic, six-electron donor, can qualitatively be considered as isoelectronic to a "C5H5 moiety. One could, therefore, envision the MO-diagram for 11 to be similar to chromocene, with the appropriate removals of degeneracy for the reduction in symmetry. Thus chromocene has an E2 g, doubly degenerate ground state with two unpaired electrons;75 Cr(r)5-C5H5)[N(SiMe2CH2PPh2)2] (11) would be expected to have a similar pair of SOMO's and from this its low-spin nature is derived. Dark red prismatic crystals of 11 were grown by slow evaporation of a saturated hexanes solution. The crystal structure (Figure 2.7), shows a pseudo-octahedral three-legged piano-stool-like monomeric Cr(II)-Cp complex with C s symmetry. Selected bond lengths and bond angles are given in Table 2.7. The Cr-N bond length of 2.066(7) A is considerably shorter (0.05 A) than in the methyl complex 5 but comparable to benzyl 7; this bond length is also very similar to that found in the chloride dimer 1 (2.078(8) A). The N-Si bond lengths of 1.687(8) and 1.689(8) A are comparable to those in Cr[N(SiMe3)2]2(THF)2 (1.674(17) A) . 2 1 The Cr-P distances in 11 are 2.353(3) and 2.366(3) A, much shorter (approx. AO.lA) than the analogous distances in methyl 5 or benzyl 7. This can easily be rationalized as being due to the low-spin nature of 11 as compared to the other, high-spin compounds. The sensitivity of the Cr-P distance to metal spin-state and environment has been previously noted both in the literature and in this thesis.37 The octahedral, low spin CrX2(dmpe)2 (X = Cl, Me, O 2 C C F 3 ) complexes have Cr-P distances ranging from 2.345-2.387 A; compound 11 is within this 60 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Table 2.7 Selected bond lengths and angles in Cr(Ti5-C5H5)[N(SiMe2CH2PPh2)2] (11). Atom Atom Distance (A) Atom Atom Distance (A) Cr P(D 2.353(3) Cr N 2.066(7) Cr P(2) 2.366(3) N Si(l) 1.687(8) Cr Cp 1.86 N Si(2) 1.689(8) Atom Atom Atom Angle (°) Atom Atom Atom Angle (°) N Cr P(2) 89.0(2) P(l) Cr P(2) 103.9(1) P(l) Cr Cp 120.5 Si(l) N Si(2) 132.7(4) P(2) Cr Cp 121.7 PCD Cr N 83.9(2) Figure 2.7 Molecular structure (ORTEP) and numbering scheme for Cr ( r f -C 5 H 5 ) [N(SiMe2CH2PPh2)2] (11)-61 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (vii) Paramagnetic A H NMR spectra of M(C5H5)[N(SiMe2CH2PPh2)2] complexes Unlike the other chromium(II) alkyls prepared, the Cr(II) cyclopentadienyl complex 11 possesses an observable A H NMR spectrum. However, only two peaks are visible and both are very broad. The two resonances, at 3.3 and 10.3 ppm, integrate in a 1:2 relative ratio and are highly temperature-dependent. These are likely the meta and para protons on the phenyl ring but in the absence of a detailed study, a definitive assignment is impossible. n * T - | i • i i | i i i i | • • i i i • • i • i • • • • i i i i i i 1 • 1 1 i 1 • ' • i • 1 1 1 i 1 1 1 ' l 1 ' 1 1 i ' • ' 1 i 1 1 1 1 i • 1 1 1 i 1 1 1 • i 1 1 1 1 ) 1 1 • 1 r 85 7.5 6.5 5.5 4.5 D D . . 3.5 2.5 1.5 0.5 Figure 2.8 A H N M R spectrum of Co(C5H 5 ) [N(S iMe2CH 2 PPh 2 )2 ] (12) in C6D 6 (* ) . On the other hand, the lH NMR spectrum of Co(C5H5)[N(SiMe2CH2PPh2)2] (12) can be tentatively assigned (Figure 2.8). Resonances for all six sets of protons are observed. Notably, the resonance due to the C0-C5//5 protons at -25 ppm is visible, but is very broad (1000 Hz width at half-height). The fact that there is only one resonance for the Cp group indicates either that the ring is T]5-bound or is highly fluxional. In the absence of a crystal structure, the hapticity of the Cp ring is difficult to ascertain. The SiMe2 resonanace at 1.0 ppm is uniquely identified by 62 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes its integration (12H). The meta and ortho protons are difficult to distinguish but integration identifies them as the resonances at 7.6 and 8.1 ppm. The ortho proton is likely further downfield (8.1) as it is quite broad (0.6 ppm) while the peak at 7.6 ppm is quite sharp. The sharp peak at 7.0 is the para-proton while the broad peak at -10.2 ppm is likely due to the backbone methylenes; both integrate to four protons. In general, then, low-spin cobalt(II) complexes provide reasonable paramagnetic *H NMR spectra but neither high or low-spin chromium(II) systems presented here possess assignable spectra. (viii) Electronic structure and magnetism of CrR[N(SiMe2CH2PPh2)2] High-spin, four-coordinate Cr(II) complexes have been shown to electronically prefer a square-planar environment as the four lowest energy d-orbitals are half-filled, similar to that found for the spin-paired d 8 systems such as Rh(I) and Pt(II).2e The chromium alkyls synthesized illustrate a wide variety of hapticity and total electron-donating ability. The Cp ligand binds in an T ) 5 fashion to enable donation of its full six electrons (using oxidation state formalism). As well, the benzyl ligand binds in an rj 2 fashion, formally acting as a four-electron donor as opposed to an r)1, two-electron donor. Steric constraints may restrict T | 3 or higher coordination of the benzyl ligand but this seems unlikely given the existence of the cyclopentadienyl complex 11 having the ancillary tridentate ligand bound in a facial manner. However, it could be that the high-spin state Cr(II) centre cannot electronically bind the benzyl in a higher hapticity mode without a change in spin state.1'19'*0 The fact that the benzyl derivative 7 remains a high-spin compound indicates that the T J 2 mode of coordination is insufficient to force spin-pairing. 63 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes A Z I N D O 8 1 , 8 2 restricted open shell Hartree Fock (ROHF) calculation study confirms the stability of the experimentally observed spin states. The ability of the Z I N D O method to successfully describe the ground states of simple transition metal complexes has been noted ea r l i e r . 8 1 - 8 3 Using a simplified coordination sphere as a calculation model (Figure 2.9), the total energy for each compound in a singlet, triplet and quintet ground state mult ipl ici ty was calculated (Table 2.8). The bond lengths of each model were taken from the relevant crystal structure and were fixed for each calculation; in this way a meaningful comparison based on spin state could be made. The assumption of a fixed structure, while obviously not ideal (spin changes cause geometry changes in many cases), nevertheless suffices. To test this, the Cr-P bond lengths in the model compound fran^-Cr(CH3)NH2(PH3)2 were fixed to both typical high and typical low spin values and the multiplicity calculations repeated. Although the absolute energy values were different, the stability of the quintet state relative to the triplet and singlet state did not change, nor did the approximate magnitude of the relative stability. In addition, the calculations were performed on a tetrahedral model of Cr(CFf3)NH2(PH3)2 without any change in the qualitative order of stability. Ideally, a geometry optimization with total energy should be performed, however this feature is not available with this relatively low-level calculation program. The graph in Figure 2.10 shows the energy differences between the three possible multiplicities for the models based on the crystal structures of the hydrocarbyl compounds 5, 7 and 11, where the most stable multiplicity is set at 0 kJ/mol and the energy difference between the most stable multiplicity and the others is plotted on the y-axis. The results show that for methyl 5 and benzyl 7, there is a substantial stabilization of the quintet state relative to the singlet of 309 and 276 kJ/mol respectively. The stabilization from the triplet state is 205 and 192 kJ/mol respectively. Calculations on the cyclopentadienyl derivative 11, however, show that the triplet 64 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Table 2.8 ZINDO relative total energy calculations for the simplified complexes CrR(NH2)(PH3)2 in the indicated geometries. All values given are in kJ/mol. Spin State C H 3 a CH 2Ph a C 5 H 5 a CH^b C H 3C 1 -114261 -212564 -181548 -114633 -114311 3 -114365 -212648 -181686 -114734 -114369 5 -114570 -212840 -181623 -114926 -114512 a. model based on the crystal structure b. crystal structure geometry with Cr-P and Cr-N bond lengths altered to 2.36 and 2.06 A respectively; typical low-spin Cr(II) values taken from the crystal structure of 11 c. mechanics minimized tetrahedral structure Figure 2.9 Simplified coordination sphere models for complexes 2, 5 and 11 (left to right) used in ZINDO ROHF MO-energy calculations. 65 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes 1 1 T 1 3 5 Multiplicity Figure 2.10 Graph of calculated relative energies vs. spin multiplicity. state is the most preferred state, with stabilizations over the singlet and quintet states of 138 and 63 kJ/mol, respectively. As expected, for a high-spin square planar d 4 system, the unpaired electrons in 5 and 7 occupy the lower four d-orbitals, leaving the d x 2-y 2 orbital empty. Calculations on the analogous model system using a cobalt centre instead of a chromium centre resulted in less convincing results. In this case, the energy difference between a spin doublet and a spin quartet was calculated. The vital difference is in geometrical possibilities; in the chromium case, a square planar ligand arrangement is generally observed, but in the cobalt case both tetrahedral and square planar coordination spheres are common. Since the level of theory used in these calculations is not sufficient to geometry-optimize a transition metal complex, assumptions about the cobalt-centre geometry become necessary. Given the correct 66 References begin on page 100 Chapter 2: Cr(H) and Co(Il) Halide and Alkyl Complexes geometry of a given system (and the appropriate bond lengths), the correct spin state was predicted. For example, a square planar model of the cobalt methyl complex 6 was calculated to prefer a spin doublet by 80 kJ/mol over the S = 3/2 state. However, upon changing the geometry to a tetrahedral system, the quartet became the preferred spin-state, although the absolute energy was substantially lower for the square-planar case. Essentially, in the situation where geometry becomes a factor, the simple calculation method described demands prior knowledge of the structure in order to make any sort of prediction. A higher level of theory would substantially improve this situation. In essence, the concept of using theoretical calculations to predict the spin state of a given complex is viable given an appropriate level of theory (which we lack here) and a geometry optimization routine which can deal with transition metal complexes. Such calculations have been reported for the molybenum(III) system Cp*MoCl(PH3)2 and successfully predict the stability of the triplet state over the singlet state. Calculations indicating that spin-state change and the associated geometry changes are important thermodynamic factors with regard to reactivity were also included. 7 9 2.4 Synthesis and Structure of Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13) Given the synthetic success and stability of paramagnetic metal-carbon bonds observed up to this point, it was of interest to examine the stability of the heavier congener, silicon bonds to paramagnetic metals, particularly in the electron-deficient chromium system. Metal-si l icon bonds have been thoroughly examined and examples util izing both early and late metals in high and low oxidation states exist, but there are relatively few paramagnetic examples. 8 4 There is a series of lanthanide-silyl complexes, [Cp2M(SiMe3)2][Li(DMF)3] ( M = Sm 8 5 , L u 8 6 structurally characterized, other M = D y , H o , Er, T m ) . 8 7 , 8 8 Some paramagnetic transition metal-silyl complexes include Cp 2Nb(SiMe3)(r| 2-C2H4), 8 9 Cp*Ta(SiMe3)(PMe3)Cl2,90 Cp 2 Ti(SiH 2 Ph)(PEt3) and C p 2 T i ( S i H P h 2 ) ( P M e 3 ) , 9 1 all of which are d 1 systems. 67 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Metal-silyl complexes containing multiple unpaired electrons are restricted to one reported series, namely {[Si(SiMe3)3]2MCl}_ (M = Cr, Mn, Fe), a trigonal planar anion by virtue of the high steric bulk of the silyl fragment.92 The reported magnetic moments indicate high-spin systems, as would be expected for such electronically and coordinatively unsaturated systems, however some of the magnetic data must be considered suspect. For example, the chromium(II) silyl analogue has a reported u.eff of 6.2 B.M., a seemingly improbable result. The spin-only magnetic moment (p.S-0) of a high-spin chromium(II) centre (d4) is 4.89 B.M. Deviations from | i S - 0 due to second-order spin-orbit coupling would reduce this moment, not raise it. The other reported magnetic data show similar discrepencies. In this paper, only the iron(II) anion was structurally characterized. Other metal-silyl complexes reported in this paper include neutral Fe[Si(SiMe3)3]2(solv) (solv = OEt2, DME (bidentate)) and Mn[Si(SiMe3)3]2(DME) as well as some other anionic species. This communication was not followed by a full paper, hence it is the only report of multiple unpaired electron transition metal-silyl complexes.92 Compounds with chromium bound to silicon have been examined in some detail due to the observed stability of the chromium(0)-silicon double bond (Cr=Si) (silylene or silanediyl) complex formation. Structurally characterized examples include (o-Me2NCH2C6H4)2Si=Cr(CO)5 and similar analogues,93-94 (2-Ph2PCH2C6H4)2Si=Cr(CO)5,95 (t-C 4H 90)2Si=Cr(CO)5«HMPA, 9 6 (CO) 5 Cr=Si(HMPA)X 2 (X = Me, 9 7 Cl , 9 7 'Bu 9 8 ) and 'Bu2Si=Cr(CO)5 ,Na(CF3S03)»2TF£F.98 All of these compounds are diamagnetic systems. Chromium-silyl complexes are less common; examples include (r| 6-arene)Cr(CO)2(SiCl3)2 (arene = mesitylene, 9 9 C6H5F100) and (rj 6 . C6Me6)Cr(CO)2(H)(SiHPh2),101 the latter of which shows Cr-H-Si three-centre two-electron bonding involving the hydride. The compounds (r|6-arene)Cr(CO)(H)2(SiCl3)2 (arene = mesitylene, C6H5F) have also been reported100 and interpreted as Cr(IV) dihydride species. Again, these compounds are all diamagnetic eighteen electron species. Hence, there seems to be 68 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes no structurally characterized example of a paramagnetic chromium silyl complex and only one reported complex characterized spectroscopically. Accordingly, the metathesis reaction of chromium chloride 1 with a bulky silyl group, LiSi(Mes)2H (Mes = 2,4,6-Me3C6H2) was attempted (equation 2.9). Upon addition of the lithium salt to 1 at -78 °C in THF, the blue colour changed to green, and upon warming changed to red/brown, the same colour as the chromium alkyls. Workup and recrystallization from toluene/hexanes yielded brown crystals which were ESR and NMR silent, in line with the chromium alkyls discussed above, and were identified as a high-spin d 4 complex by Evans' method23,24 with p^ ff = 5.1 B.M. Me 2 Si 0.5 {[(Ph2PCH2SiMe2)2N]Cr}2(p-CI)2 1 LiSi(Mes) 2H Me,Si [2.9] The X-ray crystal structure of Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13) is shown in Figure 2.11 (selected bond lengths and angles are in Table 2.9) and it contains the gross features expected, namely a Cr-Si bond and a roughly square planar environment, but there are some surprises as well. In particular, the highly electronically and coordinatively unsaturated chromium centre is stabilized by a 8-C-H agostic interaction from an ortho-methyl group of one mesityl group. This interaction occurs preferentially to a Si-H oc-agostic interaction. The preference against a Si-H cc-agostic interaction is somewhat surprising as, unlike in C-H agostic systems,52 a-agostic Si-H bonds are fairly common.102 The difference in this case could 69 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes possibly be because the competing 8-C-H agostic interaction effectively blocks the open face of the molecule. The 8-C-H interaction may involve less conformational strain and be less sterically hindered. This agostic interaction is not visible in the *H NMR and IR spectra but is not an artifact; the hydrogens of interest were all located and refined. The Cr-H(40) distance is 2.62(5) A and the C(37)-H(40) distance of 1.00(4) A is elongated relative to the other two C-H bonds on C(37), which are 0.89(4) and 0.93(4) A . The Cr-H(40)-C(37) angle is 114(2)°. The Cr-Si-H(l) angle of 108(1)° indicates a tetrahedral array around silicon, consistent with a lack of Si-H interaction with chromium. The Si-H(l) bond distance of 1.45(3) A is unremarkable. Due to this agostic interaction there is substantial distortion from square planar geometry, as the dihedral angle of 25.10° between the planes defined by Cr-P(l)-N and Cr-P(2)-Si(3) shows. The angles of P(l)-Cr-P(2) and Si(3)-Cr-N of 156.33(4)° and 167.16(9)° are distorted from their ideal 180° although the angles of P-Cr-N and P-Cr-Si(3) range only from 86.68(8)° to 99.04(4)° (90° is ideal). The Cr-P bond lengths of 2.487(1) and 2.490(1) A are typical of high-spin Cr(II)-P bonds, as for the chromium alkyls reported earlier (section 2.3). The Cr-N bond length of 2.072(3) A and N-Si bond lengths of 1.691(3) and 1.704(3) A are also comparable to the chromium alkyls. The Cr-Si(3) bond length of 2.577(1) A is quite long. Metal-silicon bond distances are influenced by several key factors, the most important one being the ability of the metal centre to back-bond. Hence, low valent, electron-rich metal-silyl bonds are quite short, while rarer high-valent electron poor ones are long.84 Of course, the bond order is important, i.e., metal-silicon double bonds are shorter than single bonds. Finally, the size of the metal centre, substitutents on the silicon and steric effects in the complex in question also play a role in determining the observed metal-silicon bond length. 70 References begin on page 100 Chapter 2: Cr(ll) and Co(H) Halide and Alkyl Complexes Table 2.9 Selected bond lengths and angles in Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13). Atom Atom Distance (A) Atom Atom Distance (A) Cr P(D 2.487(1) Cr N 2.072(3) Cr P(2) 2.490(1) Cr Si(3) 2.577(1) Cr H(40) 2.62(5) N Si(l) 1.704(3) C(37) H(40) 1.00(4) N Si(2) 1.691(3) C(37) H(41) 0.89(4) C(37) H(42) 0.93(4) C(32) C(37) 1.516(6) Si(3) H(l) 1.45(3) Atom Atom Atom Angle (°) Atom Atom Atom Angle O P C D Cr P(2) 156.33(4) Si(3) Cr N 167.16(9) Cr H(40) C(37) 114(2) Cr Si(3) H(l) 108.0(10) C25 C13 Figure 2.11 Molecular structure (ORTEP) and numbering scheme for Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13). Phenyl substituents on phosphorus have been removed for clarity. 71 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Due to the paucity of comparable complexes it is difficult to analyze the Cr-Si bond length observed in 13 in great detail. Some general comparisons can be made, however. Firstly, the Cr -Si distances in the structurally characterized Cr(II) s i l y l c o m p l e x " ( r | 6 -mesitylene)Cr(CO)2(SiCl3)2 of 2.382(3) and 2.371(3) A are substantially shorter that the 2.577(1) A observed in 13 due to the fact that the arene Cr(II) complex is a diamagnetic, 18-electron compound (more back-bonding and a smaller Cr(II) centre) and the s i ly l group contains very electron withdrawing substituents. A longer bond length of 2.456(1) A is observed in (n, 6-C6Me6)Cr(CO)2(H)(SiHPh2) 1 0 1 but this length is due to the fact that a three-centre two-electron bond involving a bridging hydride lengthens the bond. It is still shorter than that observed in electron-deficient, formally twelve electron 13. Chromium-silicon double bonds in Cr(0) silanediyl complexes range from 2.342(1) A in (CO)5Cr=Si(HMPA)Cl2 (electron-withdrawing groups on s i l i c o n ) 9 7 to 2.527(3) A in (CO)5Cr=Si(FLMPA)*Bu2 (electron-donating groups on silicon, and steric factors). 9 8 In these, the larger size of a Cr(0) centre is balanced with the shorter Cr=Si double bond, which is more sensitive to electronic factors than a simple s i ly l bond . 8 4 However, these bonds are still shorter than the 2.577(1) A observed in 13. The C r - S i bond length in 13 compares well those M - S i bond lengths found in high-valent (i.e., low back-bonding) metal-si lyl complexes. The longest metal-si lyl bonds belong to paramagnetic lanthanide-silyl complexes. The L u - S i bond length in [Cp2Lu(SiMe3)2]" is an astounding 2.888(2) A, a function of the extremely large lutetium centre and the total lack of back bonding from the d° centre. 8 6 In some titanium(IV)-silyl complexes recentiy reported, 1 0 3 T i - S i bond lengths of 2.594(7) A in (Me3CCH 2 )3T iS i (S iMe 3 )3 and 2.603(3) A in (Me3SiCH2)3TiSi(SiMe3)3 are quite long, despite the small metal centre. The d ° Ti ( IV) centre cannot back-bond, thus lengthening the Ti-Si bond, and also its small size, coupled with four bulky ligands, ensures long Ti-ligand bonds due to steric interactions. Other relevant metal-silyl bond distances include 2.669(1) A in C p 2 N b ( S i M e 3 ) ( r i 2 - C 2 H 4 ) 8 9 and 2.642(1) A in Cp*Ta(SiMe 3)(PMe3)Cl2. 9 0 The Ta-Si bond length of 2.669(4) A in d ° (i.e., diamagnetic) Cp*Ta(SiMe3)Cl 3 is slightly longer than that found in d 1 C p * T a ( S i M e 3 ) ( P M e 3 ) C l 2 9 0 72 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes In summary, the long Cr-Si bond length of 2.577(1) A in 13 could be due to a combination of factors, namely the apparently minimal back-bonding in this system (due to the electron-deficient chromium centre), steric crowding at the metal centre and the comparatively large size of high-spin Cr(II) compared to diamagnetic chromium systems. 2.5 Paramagnetic Hydrides of Chromium(II) and Cobalt(II) (i) Synthesis and structure of {[(Ph2PCH2SiMe2)2N]Cr}2(p-H)216 Reaction of the chromium methyl derivative 5 with H2 in toluene (equation 2.10) yielded a colour change from red to green with the formation of a dark olive-green precipitate having the empirical formula CrH[N(SiMe2CH2PPh2)2 (14). The deuteride analogue of 14 was also prepared by reaction of 5 with D2 and the IR (KBr) spectrum clearly showed the Cr-H(D) stretching frequency shift from 1368 to 988 cm"1. The theoretical shift from 1368 cm-1 on deuteration is calculated to be 976 cnr 1 , 1 0 4 in good agreement with the observed shift. This value is diagnostic of bridging hydrides. Terminal metal hydrides are characteristically in the 1700-2500 cm-1 range.2f An oligonuclear structure was also indicated by the room temperature magnetic moment, which was determined to be 1.6 B.M., evidently a coupled system (see variable temperature magnetic studies below). X-ray quality crystals of {[(Ph2PCH2SiMe2)2N]Cr}2(p-H)2 were grown from a toluene solution of 5 to which was added 1 atm of H2 and allowed to stand without stirring for one week. Note that other square planar Cr(II) alkyls of the formula CrR2(dippe) are reported to be unreactive to H2.37 In fact, the only other examples of paramagnetic chromium(H) hydrides are the Cp*4CT4H.4 cubane105 and the mixed-valent Cp*4Cr4lT7 cubane.106 To my knowledge chromium(III) hydrides are unknown. 73 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes [2.10] Single-crystal X-ray structural analysis (Figure 2.12) shows that 14 is dinuclear, with a distorted trigonal bipyramidal coordination around the each of the Cr(II) centres; selected bond lengths and angles are in Table 2.10. In comparison to the chloride bridged dimer 1, the presence of the smaller bridging hydrides in 14 results in larger distortions from pure trigonal bipyramidal geometry. For example, the axial amide nitrogen and hydride ligands are not quite linearly disposed as the N(l)-Cr(l)-H(l) angle is 173.2(8)°. The equatorial ligands defined by the two phosphine donors and the other bridging hydride subtend the following angles: P(l)-Cr(l)-P(2), 106.09(4)°; P(l)-Cr(l)-H(l)*, 114.0(9)°; and P(2)-Cr(l)-H(l)*, 139.4(9)°. In contrast to the structure of 1 the Cr2(p-H)2 bridging core is quite symmetrical. The Cr-H distances of 1.78(3) and 1.76(3) A and the H(l)-Cr(l)-H(l)* angle of 83(1)° describe the nearly symmetrical bridge. 74 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Table 2.10 Selected bond lengths and angles in {[(Ph2PCH2SiMe2)2N]Cr}2(u.-H)2 (14). Atom Atom Distance (A) Atom Atom Distance (A) Cr(l) P(l) 2.537(1) Cr(l) Cr(l)* 2.641(1) Cr(l) P(2) 2.513(1) Cr(l) N 2.076(2) Cr(l) H(l) 1.78(3) N Si(l) 1.694(3) Cr(l) H(l)* 1.76(3) N Si(2) 1.692(3) Atom Atom Atom Angle (°) Atom Atom Atom Angle (°) P(D Cr(l) P(2) 106.09(4) N(l) cm H(l) 173.2(8) P(D Cr(l) H(l)* 114.0(9) H(l) Cr(l) H(l)* 83(1) P(2) Cr(l) H(l)» 139.4(9) Si(l) N Si(2) 130.8(1) Figure 2.12 Molecular structure (ORTEP) and numbering scheme for {[(Ph2PCH2SiMe2)2N]Cr}2(u.-H)2 (14). Phenyl substituents on phosphorus have been removed for clarity. 75 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes The Cr-Cr distance of 2.641(1) A in 14 is 1 A shorter than in 1 and hence the potential existence of a Cr-Cr bond cannot be ruled out. The Cr-P distances of 2.513(1) and 2.537(1) A are comparable to those found in 1 and in other known high-spin Cr(II) phosphines.20,37,50 The Cr-N bond length of 2.076(2) A is similar to that found in Cr[N(SiMe3)2]2(THF)2 2 1 (ii) Variable temperature magnetic susceptibility of {[(Ph2PCH2SiMe2)2N]Cr}2(p>X)2 The pair of compounds {[(Ph2PCH2SiMe2)2N]Cr}20i-X)2 (X = Cl (1); H (14)), characterized by X-ray diffraction as being globally isostructural, provided an excellent opportunity to examine the effect of the bridging ligands on antiferromagnetic exchange. Communication between metal centres as manifested by magnetic interactions is an area of intense current interest.107 For example, the search for molecular ferromagnets108"110 and investigations on the nature of metal ion sites in proteins111 rely on an understanding of how electron spin information is transmitted between metal centres either through ligands or through direct metal-metal bonds. A simple approach to study the phenomenon of magnetic exchange is to examine dinuclear metal complexes connected by different types of bridging ligands.112"114 While a number of studies have addressed this approach, noticeably absent are studies that include hydrides as bridging ligands. This is partly due to the paucity of paramagnetic hydride dimers; only a few structurally characterized examples exist, including [[(triphos)RhJ2(p> H) 3][BPh 4] 1 1 5 and [(dippp)Co]2(H)(u.-H)3 u 6 The VT magnetic susceptibility of the Rh-dimer has not been reported and the cobalt dimer becomes diamagnetic below 280 K; the origin of the latter magnetic behaviour was attributed to several possible factors, including a monomer-dimer or a singlet-triplet state equilibrium. There is ample precedent in the literature to suggest that dinuclear Cr(II)-Cr(II) complexes should be diamagnetic with potentially multiple metal-metal bond character,23'5,6 but in this case paramagnetic chromium(II) dimers were isolated. In order to better understand the 76 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes magnetic behaviour of these systems, a solid state variable temperature magnetic susceptibility study was initiated. The magnetic susceptibility vs. temperature data for 1 (chloride dimer) were obtained from 4-119 K using a vibrating sample magnetometer. For 14 (hydride dimer) a SQUID magnetometer was used to collect the data from 2-340 K. The plot of the magnetic susceptibility of 1 (per mole of dimer), Xm» vs. temperature reveals a broad maximum in susceptibility around 66 K consistent with the presence of significant antiferromagnetic coupling in the compound (Figure. 2.13). The corresponding plot for 14 (Figure 2.14) indicates the approach to a maximum at temperatures above 300 K. Unfortunately this compound is not sufficiently thermally stable above this temperature to permit meaningful susceptibility measurements in this region. Nonetheless, the general features of the %m vs. T plot at temperatures below 300 K as well as the observation that the %m values for 14 are an order of magnitude smaller than those for 1 suggest even stronger antiferromagnetic exchange coupling in 14. The increase in susceptibility with decreasing temperature at the lowest temperatures studied as observed for both 1 and 14 likely arises from the presence of paramagnetic impurity. This is commonly seen in studies of antiferromagnetically coupled systems.117 The magnetic susceptibility data for 1 and 14 were analyzed employing the dimeric Ffeisenberg model for S = 2, where x = J/kT. 1 1 8 To account for the presence of paramagnetic impurity the expression was combined with the Curie law term, Xdimer — N A g 2 P 2 # 2e 2 x + 10e6x + 28e 1 2 x + 60e2 0 x kT * i + 3e2* + 5e6 x + 7e 1 2 x + 9e 2 0 x Xpara -NAg 2P 2S(S + 1) 3kT 77 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes o B » E I XI a, v v s u fi M eg S 100 120 Temperature (K) Figure 2.13 Magnetic susceptibility (per mole of dimer) vs. temperature plot for chloride 1. The line was generated using the Heisenberg dimer model with J = -12.4 cm"1, g = 1.99 and P = 0.007. o B s .a =i ft v w in 9 O S) C M cd S 200 300 Temperature (K) Figure 2.14 Magnetic susceptibility (per mole of dimer) vs. temperature plot for hydride 3. The line was generated using the Heisenberg dimer model with J = -139 cm"1, g = 1.98 and P = 0.0017. Experimental data were corrected for TIP = 300 x 10"6 cm3 mol"1. 78 References begin on page 100 Chapter 2: Cr(II) and Co(H) Halide and Alkyl Complexes according to, %m — [l~P]%dimer + PjCpara where P represents the fraction of paramagnetic impurity. Fits of the experimental data to the model were achieved using a nonlinear least-squares procedure with the following as the function minimized: F = j _ y o c i o b s - % i c a l c ) 2 n 1 = 1 eci o b s) 2 1/2 The F value provides a measure of the goodness of fit between experiment and theory. Experimental susceptibility vs. temperature data for the chloride dimer 1 are compared with the best fit results from theory with J = -12.4 cm"1, g = 1.99 and P = 0.007 (F = 0.00144). A g value slightly less than the free spin value is not unexpected for a d 4 system.119 The Cr-Cr separation of 3.64 A would preclude direct metal orbital overlap as the mechanism for exchange in this compound requiring, therefore, a superexchange mechanism via the bridging chloro-ligands. The exchange coupling constant, J, is comparable in magnitude to values reported for some extended chain and sheet chloro-bridged Cr(II) systems.120,121 The compounds CsfCrCbj], [(CH3)4N][CrCl3] and [CsHsNH^tCrCL;] are reported to be antiferromagnetic with J values of approximately -25, -12.5 and -6.5 cm - 1, respectively. The complex Cs[CrCl3(OH2)2] which is thought to have a dimeric, double chloro-bridged structure similar to that of 1, is antiferromagnetic with J = -3 cm - 1 . 1 2 1 It is somewhat surprising that the magnitude of J is so large for 1 considering the asymmetric nature of the chloro bridges. Unfortunately there seems 79 References begin on page 100 Chapter 2: Cr(II) and Co(IJ) Halide and Alkyl Complexes to be no other report on a chloro-bridged dimeric Cr(II) system where both detailed magnetic and single crystal X-ray diffraction studies have been reported. A single crystal X-ray diffraction study of [CrCl2(dippe)]2 (dippe = l,2-bis(diisopropylphosphino)ethane) revealed a dimeric structure with unsymmetrical bridging chloro-ligands (Cr-Cl distances of 2.380 and 2.606 A) much like the situation in l . 2 0 Unfortunately, the only magnetic study reported on this compound is a room-temperature solution measurement in acetonitrile. Under these conditions a mononuclear acetonitrile adduct is formed and the magnetic study merely confirms the presence of a high-spin d 4 metal center. In an unusual example, chloro-bridged Cr(II) systems also provide examples of ferromagnetism, a relatively rare phenomenon in metal complex chemistry. The compounds M2[CrCl4] (M = K, Rb, Cs) have structures involving two-dimensional networks of Cr and Cl atoms, each Cr being surrounded by four others, connected by chloride bridges with 180° Cr-Cl-Cr angles. Magnetic studies on these systems have revealed ferromagnetic exchange with J values estimated to lie in the range 6.5 to 8.5 cm - 1 . 1 2 2 The structure of the hydride dimer 14 reveals a dimetallic system with the metals separated by a symmetrical double hydride-bridge and a Cr-Cr separation of only 2.641(1) A. It is not surprising therefore that the metals are more strongly coupled antiferromagnetically than in chloride 1. The magnetic susceptibilities were analyzed as for 1 with the exception that the experimental data were first corrected for temperature independent paramagnetism (TIP).119 The best fit between experiment and theory, employing a TIP correction of 300 x 10"6 cm3moH (per dimer unit) yielded J = -139 cm"1, g = 1.98 and P = 0.0017 with F = 0.0989. Employing a larger TIP correction of 400 x 10"6 cm3mol"1 yields a better fit (F = 0.0521) with J = -150 cm"1, g = 2.069 and P = 0.0014. However, the former fit is favoured in the light of the more realistic g value it generates. In either case, it is clear that the absolute value of J calculated for the bridging hydride complex is at least of the order of 140 cm"1 and is considerably greater than that observed for the bridging chloride. The only other magnetic investigation of a paramagnetic chromium(II) hydride was marred with errors; the initial communication of the cubane Cp*4G"4H4 was later determined to be a mixture of two complexes.105 While this did not 80 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes significantly affect the structural results, the magnetic study in this publication and its conclusions are completely erroneous. In two later papers the chromium(II) cubane49 and the impurity, Cp*4Cr4H7 were separated and characterized; although the magnetic data for Cp*4Cr4H4 indicated some level of antiferromagnetic coupling, the data could not be fit to a model and no J-values were reported.49 The mixed-valent cluster Cp*4&4H7 shows Curie-Weiss behaviour consistent with an S = 7/2 ground state arising from a ferrimagnetic arrangement of three S = 3/2 Cr(III) and one S = 1 Cr(II) centre.106 It is impossible to distinguish between direct metal orbital overlap and ligand mediated superexchange as potential mechanisms for the interaction from the magnetic analysis given above. Comparisons with related systems could prove useful in this regard, however, the lack of relevant structural and magnetic studies, particularly on hydride-bridged complexes, precludes this possibility. Magnetic studies on dimetallic Cr(II) species involving Cr2(p-OR)2 cores (where OR is aryloxide or alkyloxide) have been analyzed qualitatively and the results reveal dependence of the strength of antiferromagnetic coupling on the Cr-Cr distance and the nature of the bridging group, however no definite magneto-structural correlations have been illustrated in this work. 7 , 1 2 3 An amide-bridged dimer containing a Cr2(ji-NR2)2 core has also been studied magnetically but only at high temperatures.124 The compound exhibits a reduced magnetic moment indicative of antiferromagnetic exchange between metal centres. To summarize, it has been shown that dinuclear chromium(II) complexes containing two single atom bridges can participate in magnetic exchange, however, the extent of the magnetic coupling is highly dependent on the nature of the bridge. For complex 1 having chloride bridges, the coupling is rather small and comparable to other halide-bridged complexes of Cr(II). However, the magnetic exchange coupling for the dinuclear hydride complex 14 is an order of magnitude larger than that of 1 and indicates that hydride ligands are extremely efficient mediators of magnetic information. 81 References begin on page 100 Chapter 2: Cr(II) and Cb(II) Halide and Alkyl Complexes (iii) Synthesis and decomposition of Cr(CH2CH3)[N(SiMe2CH2PPh2)2] (15) The stability of M-CH3 systems have long been understood in terms of the lack of a (3-hydrogen abstraction-decomposition pathway.125 In order to examine the stability of Cr(II) paramagnetic organometallics containing p-hydrogens, the chloride dimer 1 was reacted with EtLi in an attempt to generate Cr(CH2CH3)[N(SiMe2CH2PPh2)2] (15). Unfortunately, the ensuing reaction is not straightforward; although it appears that the ethyl compound is generated, it readily loses ethylene to form what is identified as the final product: compound 14, the hydride dimer (Scheme 2.2). The formation of other biproducts and the low yield of hydride complicates analysis of this reaction. Scheme 2.2 0.5 v N P h . CH 3CH 2Li Me 2 Si^ ^ c . r - c i c l ~ C r % I Me'si Me M e ' \ l e 0.5 : N _ C r — C H 2 C H 3 M e 2 S I 0 ,. P h , M e N Me M V S i M e w N H 2C :CH 0 Mesi^Me M e ' \ l e 82 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes The presence of a p-hydrogen abstraction pathway operating in the chromium ethyl derivative implied that the reverse reaction, ethylene insertion into a hydride, might also occur. Accordingly, reaction of the hydride dimer 14 with one atmosphere of ethylene generated a red/brown solution, consistent with a high-spin Cr(II) alkyl. No polyethylene was detected. Upon removal of the ethylene, the hydride was regenerated (Scheme 2.2). Hence, high-spin Cr(II) alkyls are susceptible to P-hydride abstraction-decomposition pathways. This likely hinders any ethylene polymerization activity of Cr(II) alkyls (section 2.6(i)). The hydride 14 was also examined for evidence of H/D exchange by solvent C-H(D) activation, as has been observed for Cp*4Cr4H4.126 Unfortunately the thermal instability over 50 °C of 14 precluded this investigation. (iv) Attempted synthesis of a cobalt(II) hydride complex In contrast to the clean reaction of CrMe[N(SiMe2CH2PPh2)2] with dihydrogen, reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with four atmospheres of H2 resulted in a slow darkening of the yellow solution to an eventual brown colour. Workup of this solution yielded only intractable products; solution and solid (KBr) IR spectroscopy showed no evidence of a Co-H vibration band. The same intractable brown product was also obtained on metathesis of the chloride of CoCl[N(SiMe2CH2PPh2)2] (2) with KBEt3H, hence it seems that the four-coordinate cobalt hydride, or a comparable bridging hydride, is inaccessible if the chromium is replaced by cobalt. Reaction of 2 with IJBH4 in an attempt to make a stabilized bridging borohydride resulted in total decomposition. Five-coordinate phosphine-stabilized cobalt(II) hydrides are known 1 2 7 , 1 2 8 so the reaction of 6 with H2 in the presence of PEt3 was attempted (equation 2.11). In this case, a bright orange toluene solution resulted. Removal of the toluene and slow evaporation of an orange hexamethyldisiloxane solution gave orange crystals which analyzed for CoH(PEt3)[N(SiMe2CH2PPh2)2] (16). 83 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes P h 2 P P h 2 M e 2 S i PEta Me 2 Si PEt 3 : N — C o — C H 3 [2.11] Me 2 Si Me 2 Si H P P h 2 P P h 2 6 16 The solution magnetic moment was 2.1 B.M., consistent with a low-spin Co(II) complex, as opposed to a Co(I) reduction product (i.e., Co(PEt3)[N(SiMe2CH2PPh2)2], which does not form). However, deuteration and examination of the IR spectra did not allow for a conclusive Co-H stretching frequency to be identified and no suitable crystals for X-ray diffraction could be obtained. 2.6 Reactivity of Chromium(II) and Cobalt(II) Methyl Complexes with Small Molecules (i) Ethylene Chromium(II) and cobalt(II) alkyl complexes were both unreactive to ethylene. As has been shown by others, Cr(II) does not seem to be the active species in ethylene polymerization.14-37'129 In the case of chromium, this inertness is attributed to the presence of four unpaired electrons, which means that four of the five d-orbitals are half-filled and hence may not be available to bind ethylene. In the cobalt complex, the empty d x 2 . y 2 orbital is of the wrong symmetry to interact with ethylene. It is possible that low molecular weight oligomers were produced in these systems but this was not examined in detail. 84 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes Scheme 2.3 H O C H M e j M e f N 14 PhoP ' \ \ ^ P P h j H ~ C r r polyacetylene N / / M © S i M e M e ' \ l e Me 2 Si P h 2 P \ no reaction Me 2Si<f : N — C r — C P -P P h 2 N C H 3 18 other products (ii) Carbon Monoxide Addition of one atmosphere of CO to a yellow toluene solution of the cobalt methyl complex 6 causes an immediate colour change to orange. The solution IR spectrum of this orange product showed two bands in the vCO region, at 1911 cm - 1 and 1829 cm - 1, which are assignable to a terminal Co-CO ligand and an acyl C0-COCH3 ligand; the product is thus tentatively formulated as Co(CO)(COCH3)[N(SiMe2CH2PPh2)2] (17). The reaction of the chromium methyl complex 5 with CO is much less straightforward, perhaps due to the higher electronic deficiency of the Cr(II) alkyl centre relative to cobalt. Upon addition of one atmosphere of CO to a solution of 5, the appearance of a large number of IR bands in the region from 2058 to 1585 cm - 1 indicate multiple product formation. Possible 85 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes structures include simple adducts, the acyl insertion product, rearranged reduced Cr(0) N-acylimidate species (as is observed in the NiMe[N(SiMe2CH2PPh2)2] system)130 or a combination of all of these. Separation and identification of the products were attempted without success. However, contrary to previously examined Cr(II) alkyl reactivity with C O , 3 7 the Cr[N(SiMe2CH2PPh2)2] fragment remains intact, as observed by the MS of product mixtures (rule 580). Treatment of CrR.2(dippe) with CO resulted in Cr(CO)4(dippe) and ketone products by reductive elimination. Reaction of methyl 5 with one equivalent of CO yields a difficult-to-purify compound with a single v c o = 1842 cm - 1 ; likely this is the acylated product, Cr(COCH3)[N(SiMe2CH2PPh2)2] (18). As well, solid 5 reacts with CO to yield a bright orange product with multiple CO stretches in the IR, but crystallinity was lost upon reaction so that even starting with X-ray quality crystals of 5 only orange powders were produced. (iii) Other substrates Upon addition of acetylene to the chromium methyl complex 5, the rapid production of polyacetylene was observed. The mechanism is likely organic-radical based, with the paramagnetic metal-based radical 5 merely initiating the polymerization. Even trace amounts of 5 can initiate polymerization. Using larger amounts of 5 resulted in the almost total recovery of unreacted methyl complex, strongly supportive of the notion that the compound itself is not directly interacting in the polymerization cycle. The chromium methyl complex 5 reacted with phenylacetylene quickly and diphenylacetylene more slowly but the products in both cases could not be identified. The cobalt methyl complex 6 reacted with excess phenylacetylene rapidly to give a dark red/brown solution from which oligomers of phenylacetylene (from two to nine units) were observed by mass spectroscopy and also in the *H NMR spectrum (alkenyl resonances between 4 and 6.5 ppm). 86 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes This reaction was not examined further, although the use of cobalt complexes in alkyne oligomerization is well known. 1 5 , 1 3 1 Scheme 2.4 2.7 Summary and Conclusions Using the monoanionic ancillary amidodiphosphine ligand ~N(SiMe2CH2PPh2)2, chromium(II) and cobalt(II) halide complexes were synthesized and the chloride derivatives used to synthesize a variety of organometallic compounds. The halide complexes are all high-spin systems, the chromium complex being a five-coordinate halide-bridged dimer and the cobalt complexes tetrahedral monomers. Alkylation of the chromium halide yielded high-spin square-planar alkyls, with the exception of the cyclopentadienyl complex, which is low-spin. On the other hand, all of the cobalt(II) alkyls became low-spin complexes upon metathesis of the 87 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes chloride ligand. The electron-deficiency of the chromium(II) centre was illustrated by the formation of an r]2-benzyl complex, which did not form in the analogous cobalt(II) case. Reactivity of these alkyl complexes with industrially important small molecules was generally disappointing, although a dimeric chromium(II) hydride was formed by reaction of the alkyls with hydrogen. Magnetostructural correlations were made between the isostructural chromium(II) chloride and hydride and showed that the hydride ligand is very efficient at mediating antiferromagnetic exchange interactions. A comparison of the d 4 chromium(II) alkyl complexes with the analogous diamagnetic d 8 nickel(II) complexes NiR[N(SiMe2CH2PPh2)2] reveals similar structures36,132 (both are square planar) and both react in a complex fashion with C O . 1 3 0 The nickel chloride complex NiCl[N(SiMe2CH2PPh2)2], on the other hand, is square planar; the chromium analogue is dinuclear and pentacoordinate.16,132 With the metal(II) alkyls and halides firmly in hand, the investigation of the redox chemistry of these systems will be described in the next two chapters, in particular, the one-electron redox reaction with alkyl halides. 2.8 Experimental 2.8.1 General Procedures Unless otherwise stated all manipulations were performed under an atmosphere of dry, oxygen-free dinitrogen or argon by means of standard Schlenk or glovebox techniques. The glovebox used was a Vacuum Atmospheres HE-553-2 workstation equipped with a MO-40-2H purification system and a -40 °C freezer. *H NMR spectroscopy was performed on a Varian XL-300 or a Bruker AC-200 instrument operating at 300 and 200 MHz respectively, and were 88 References begin on page 100 Chapter 2: Cr(II)and Co(II) Halide and Alkyl Complexes referenced to internal C6D5H (7.15 ppm). Magnetic moments were measured by a modification of Evans' method23 ,24 (C6D5H or Cp2Fe as a reference peak) on the NMR spectrometers listed above and were also measured in the solid state (compounds 1 and 14) using a Johnson-Matthey MSB-1 Gouy balance at room temperature. Infrared spectra were recorded on a BOMEM MB-100 spectrometer. UV-vis spectra were recorded on a HP-8452A or HP-8453A diode array spectrophotometer. Mass spectra were measured using a Kratos MS-50 EI instrument operating at 70 eV. Microanalyses (C, H, N) were performed by Mr. P. Borda of this department. 2.8.2 Materials The lithium salt LiN(SiMe2CH2PPh2)218 was prepared by the method described in the literature. CrCl2*THF was prepared by Soxhlet extraction of commerical CrCl2 (Strem) in T H F . 1 7 C0CI2 and CoBr2 were either used as received or dried in refluxing MesSiCl. NaCp»DME was prepared by the reaction of Na with CpH in dry DME. KCH2Ph133 and EtLi 1 3 4 were prepared by literature procedures. All other reagents were obtained from commercial sources and used as received. Hexanes, toluene and THF were heated to reflux over CaH2 prior to a final distillation from either sodium metal or sodium benzophenone ketyl under an Ar atmosphere. Deuterated solvents were dried by activated 3-A molecular sieves; oxygen was removed by trap-to-trap distillation and three freeze-pump-thaw cycles. 2.8.3 Molecular Orbital Calculations All molecular orbital calculations were performed on the CAChe Worksystem, a product developed by Tektronix. The parameters (INDO/1) used in the ZINDO semiempirical molecular orbital calculations on all model compounds were taken from the literature.82,83 The bond lengths for the models mws-CrR(NH2)(PH3)2 (R = CH3, ri2-CH2Ph, r^-CsHs) were taken from the X-ray crystal structure analyses of 5, 7 and 11, respectively and the PH3 groups were eclipsed to effect maximum symmetry. Similarly, for the cobalt methyl, compound 6, bond lengths were 89 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes taken from the X-ray crystal structure. The Cartesian coordinates of the models can be found in the appendix. For all models the following standard bond lengths were used: P-H, 1.380 A; C-H, 1.090 A; N-H, 1.070 A. 2.8.4 Synthesis and Reactivity of Complexes (i) {[(Ph2PCH2SiMe2)2N]Cr}2(p-Cl)2 (1) To a stirred suspension of CrCl2»TF£F (2.55 g, 11.1 mmol) in 30 mL TFTF was added dropwise a 20 mL THF solution of LiN(SiMe2CH2PPh2)2 (5.44 g, 10.1 mmol) over 5 minutes, at room temperature. The relatively insoluble light blue CrCl2»THF rapidly dissolved to generate the dark blue to indigo initial product Cr(THF)Cl[N(SiMe2CH2PPh2)2]. After 10 min, the THF was removed in vacuo to yield a light blue-green solid, which was extracted with toluene and filtered through Celite. The resulting blue solution was recrystallized from minimum toluene/THF (95:5) to give [{[(Ph2PCH2SiMe2)2N]Cr}2(u-CI)2 (1) as turquoise prisms. Yield 5.50 g (88%). Anal. Calcd. for C3oH36ClCrNP2Si2: C, 58.48; H, 5.89; N, 2.27. Found: C, 58.64; H, 5.81; N, 2.34. ' H NMR (C 7D 8): Too insoluble. 31p{lH} NMR (dg-THF): 6 -22.1 (100 Hz linewidth) MS: mle 615 (M+). UV-vis (C 7H 8): 284 (e = 3260 M-lcrrr1), 342 (shoulder, e = 450 M-lcirr1), 522 (e = 140 M^cnr 1), 688 (e = 160 M^cnr 1) nm. p e f f = 4.6 B.M. (ii) Cr(py)Cl[N(SiMe2CH2PPh2)2] (l.py) To a stirred blue slurry of {[(Ph2PCH2SiMe2)2N]Cr}2(p-Cl)2 (0.15 g, 0.12 mmol) in 10 mL toluene was added pyridine (0.30 mL, 3.7 mmol) to yield a dark green solution. Removal of the solvent in vacuo yielded a green oil. Addition of hexanes and filtration through Celite yielded a green solution, which upon being cooled formed purple crystals after two days. Yield 0.14 g (83%). Anal. Calcd. for C35H4iClCrN2P2Si2: C, 60.46; H, 5.94; N, 4.03. Found: C, 60.19; H,5.89; N, 3.90. 90 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (iii) CoCl[N(SiMe 2CH 2PPh 2)2] (2) To a 15 mL THF suspension of C0CI2 (0.35 g, 2.7 mmol) was added dropwise a 10 mL THF solution of LiN(SiMe2CH2PPh2)2 (1-40 g, 2.6 mmol). The light baby blue suspension formed a dark blue solution over 30 minutes. After overnight stirring the THF was removed in vacuo, the residue extracted with toluene, filtered through Celite and the toluene evaporated to near dryness. The resulting oil was layered with hexanes and allowed to stand. Two days later crystals of CoCl[N(SiMe2CH2PPh2)2] (2) were collected and dried. Yield: 1.34 g (82%). Anal. Calcd. for C3oH36ClCoNP2Si2: C, 57.83; H, 5.82; N, 2.25. Found: C, 57.74; H, 5.87; N, 2.29. !H NMR (C 6D 6): 5 15.0 (v br, 8H), -5.4 (br, 4H). UV-vis (C 7H 8): 506 (e = 330 M^cm'1), 602 (e = 700 M-icm-1), 786 (e = 290 M - W 1 ) nm. MS: mle 622 (M+). u.eff = 4.2 B.M. (iv) CoBr[N(SiMe2CH2PPh2)2] (3) To a 15 mL THF suspension of CoBr 2 (0.35 g, 1.60 mmol) was added dropwise a 10 mL THF solution of LiN(SiMe2CH2PPh2)2 (0.86 g, 1.61 mmol). From the light green suspension formed a dark blue-green solution over 30 minutes. After being stirred overnight the THF was removed in vacuo, the residue extracted with toluene, filtered through Celite and the toluene evaporated to near dryness. Crystals of CoBr[N(SiMe2CH2PPh2)2] (3) slowly formed from minimum toluene overnight and then were washed with hexanes, collected and dried. Yield: 0.76 g (71%). IH NMR (C 6D 6): 5 15.2 (v br, 8H), -5.2 (br, 4H). UV-vis (C/Hg): 518 (e = 310 M-icm-1), 616 (e = 530 M-icm-l), 792 (e = 270 M-icm-^nm. MS: mle 668 (M+). u.eff = 4.1 B.M. (v) CrMe[N(SiMe2CH2PPh2)2] (5) To a stirred solution of {[(Ph2PCH2SiMe2)2N]Cr}2(u.-Cl)2 (1.40 g, 1.13 mmol) in 20 mL THF was added dropwise MeLi (8.2 mL, 0.28 M in ether, 2.30 mmol) or MeMgBr (stock o solution in ether) at -78 C. The dark blue-indigo solution quickly turned dark brown. The reaction was allowed to warm to room temperature and then the THF was removed in vacuo over 91 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes 1 h to yield a brown tar. Extraction with hexanes and filtering through Celite yielded brown parallelogram prisms of CrMe[N(SiMe2CH2PPh2)2] (5) after 5 minutes, which were washed with cold hexanes. Yield: 1.36 g (90%). MS: mle 580 (M+-H.) Anal. Calcd. for C3iH39CrNP2Si2: C, 62.50; H, 6.60; N, 2.35. Found: C, 62.44; H, 6.62; N, 2.47. peff = 4.7 B.M. (vi) CoMe[N(SiMe2CH2PPh2)2] (6) A 10 mL THF solution of CoCl[N(SiMe2CH2PPh2)2] (2) (0.25 g, 0.4 mmol) was cooled to -78 °C and MeLi (0.3 mL, 1.4 M in ether, 0.4 mmol) was added dropwise until the blue colour changed to yellow-orange. This was warmed to room temperature, stirred for 30 minutes and the THF removed in vacuo. The residue was extracted with hexanes, filtered through Celite and the yellow hexanes solution allowed to slowly evaporate. Yellow crystals of 6 were deposited overnight. Alternately, a 10 mL toluene solution of 1 was titrated with MeLi until the blue colour dissipated to yellow-orange. Removal of the toluene and extraction with hexanes as before yielded CoMe[N(SiMe2CH2PPh2)2] (6). Yield: 0.23 g (95%). Anal. Calcd. for C31H39C0NP2S12: C, 61.78; H, 6.52; N, 2.32. Found: C, 61.64; H, 6.63; N, 2.40. *H NMR (C 6D 6): 5 9.8 (v br, 8H, ortho-Ph), 8.4 (s, 4H, para-Ph), 6.5 (s, 8H, meta-Ph), -3.3 (br, 12H, SiMe2), -6.3 (v br, 4H, CH2). MS: mle 602 (M+), 587 (M+-CH3). peff = 2.2 B.M. (vii) Cr(7i2.CH2Ph)[N(SiMe2CH2PPh2)2] (7) To a 15 mL THF solution of {[(Ph2PCH2SiMe2)2N]Cr}2(u-Cl)2 (0.15 g, 0.12 mmol) was added a solution of KCH2PI1 (0.032 g, 0.25 mmol) in 5 mL THF dropwise at -78 °C. The colour turned from navy blue to dark purple. The solution was warmed to room temperature and the THF removed in vacuo. The residue was extracted in hexanes and filtered through Celite, yielding a purple solution from which large cubes of Cr(ri2-CH2Ph)[N(SiMe2CH2PPh2)2] (7) formed after 5 hours. Yield: 0.13 g (80%). Anal. Calcd. for C37H43CrNP2Si2: C, 66.14; H, 92 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes 6.45; N, 2.08. Found: C, 65.93; H, 6.46; N, 2.15. |ieff = 5.1 B.M. IR: 1582 cm"1 (CH2Ph, C=C). (viii) Co(CH2Ph)[N(SiMe2CH2PPh2)2] (8) A 10 mL THF solution of CoCl[N(SiMe2CH2PPh2)2] (2) (0.17 g, 0.27 mmol) was cooled to -78 °C and a 5 mL THF solution of KCH 2Ph (0.035 g, 0.27 mmol) was added dropwise. The blue solution rapidly turned dark orange/brown. The reaction was brought to room temperature and after being stirred overnight the THF was removed in vacuo, the residue extracted with ether, filtered through Celite and hexanes (1:1) added. Slow evaporation yielded red prisms of Co(CH2Ph)[N(SiMe2CH2PPh2)2] (8). Yield: 0.10 g (55%). Anal. Calcd. for C37H43C0NP2S12: C, 65.47; H, 6.38; N, 2.06. Found: C, 65.15; H, 6.33; N, 2.00. *H NMR (C 6D 6): 5 12.5 (br, 2H, CH2Ph), 8.1 (br, 8H, meta-Ph), 7.3 (br, overlap), 7.1 (s, overlap), -1.8 (br, 2H, Ct\2Ph), -4.0 (br, 12H, SiMe2). MS: mle 678 (M+), 587 (M+- C7H7). u. e f f = 2.1 B.M. (ix) Cr(CH2SiMe3)[N(SiMe2CH2PPh2)2] (9) To a 20 mL THF solution of {[(Ph2PCH2SiMe2)2N]Cr}2(p>Cl)2 (1) (0.60 g, 0.49 mmol) was added a solution of LiCH2SiMe3 (0.091 g, 0.98 mmol) in 5 mL of toluene dropwise at -78 "C. The colour turned from navy blue to dark brown. The solution was warmed to room temperature and the colour changed to dark purple after ten minutes. After stirring for one hour, the solvents were removed in vacuo, the residue was extracted in hexanes and filtered through Celite yielding a purple solution of Cr(CH2SiMe3)[N(SiMe2CH2PPh2)2] (9). This solution only yielded purple oils, from which no solids could be obtained. Consequently, elemental analysis was not possible, and characterization is based on further reactivity of this complex with benzyl chloride to give a chromium(III) product (Chapter 3). Compound 9 was used in situ for a variety of reactions. 93 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (x) Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] (10) A 10 mL THF solution of CoCl[N(SiMe2CH2PPh2)2] (2) (0.16 g, 0.26 mmol) was cooled to -78 °C and a 5 mL toluene solution of LiCH2SiMe3 (0.024 g, 0.26 mmol) was added dropwise. The flask was brought to room temperature over 10 minutes during which the blue/green solution turned orange. After being stirred overnight the THF was removed in vacuo, the residue extracted with hexanes, filtered through Celite and reduced to minimum hexanes. The solution deposited orange crystals in the freezer overnight. Alternately, a 10 mL toluene solution of 2 was titrated with LiCH2SiMe3 until the blue colour changed to orange. Removal of the toluene and extraction with hexanes as before gave Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] (10). Yield: 0.090 g (52%). Anal. Calcd. for C37H4 7CoNP2Si3: C, 60.51; H, 7.02; N, 2.08. Found: C, 60.30; H, 7.34; N, 2.31. lH NMR (C 6D 6): 8 8.5 (s, 4H, para-Ph), 6.3 (br, 8H, meta-Ph), -4.0 (v br, 12H, SiMe2), -9.5 (v br, 9H, SiMe3). MS: mle 61A (M+), 587 (M+ - CH 2SiMe 3). peff = 2.1 B.M. (xi) Cr(ri5-C5H5)[N(SiMe2CH2PPh2)2] (11) Method 1. To a 15 mL THF solution of {[(Ph2PCH2SiMe2)2N]Cr}2(p-Cl)2 (1) (0.20 g, 0.16 mmol) at -78 °C was added dropwise a 5 mL THF solution of NaCp»DME (0.058 g, 0.32 mmol). The blue solution rapidly turned dark red and was warmed to room temperature. After being stirred for one hour, the THF was removed in vacuo and the residue extracted in hexanes. Red crystalline bars were obtained from the slow evaporation of a hexanes solution. Yield: 0.18 g (86%). Method 2. To a stirred solution of CrMe[N(SiMe2CH2PPfi2)2] (5) (0.11 g, 0.19 mmol) in 10 mL toluene was added a 10-fold excess of C5H6 by vacuum transfer. The light red solution turned a much darker red as soon as the transfer was complete. The reaction was stirred overnight and then pumped to dryness. Extraction with minimum hexanes yielded a dark red solution which yielded the dark red solid Cr(r|5-C5H5)[N(SiMe2CH2PPh2)2] (11) after 2 days of 94 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes slow evaporation. Yield: 0.10 g (85%). Anal. Calcd. for C35H4iCrNP2Si2: C, 65.09; H, 6.40; N, 2.17. Found: C, 64.89; H, 6.54; N, 2.18. (ieff = 2-7 B.M. MS: mle 645 (M+), 580 (M+-Cp) (xii) Co(C5H5)[N(SiMe2CH2PPh2)2] (12) CoCl[N(SiMe2CH2PPh2)2] (2) (0.10 g, 0.16 mmol) was dissolved in 10 mL toluene. To this was added NaCp»DME (0.028 g, 0.16 mmol) in 10 mL toluene dropwise to very quickly yield first an orange and then a red/brown solution upon completion of addition. This was stirred overnight then filtered through Celite and pumped to dryness to give a brown/red solid, which was recrystallized from a toluene/hexanes mixture (1:1) standing for one week to give brown crystals of Co(C5H5)[N(SiMe2CH2PPh2)2] (12). Yield: 0.085 g (82%). Anal. Calcd. for C35H41C0NP2S12: C, 64.40; H, 6.33; N, 2.15. Found: C, 64.70; H, 6.42; N, 2.01. *H NMR (C 6D 6): 8 8.1 (br, 8H, ortho-Ph), 7.6 (br, 8H, meta-Ph), 7.05 (s, 4H, para-Ph), 1.0 (br, 12H, SiMe2), -11 (v br, 4H, CH2), -25 (v br, 5H, Cp). MS: mle 652 (M+), 587 (M+ - C5H5). u.eff = 1.9 B.M. (xiii) Cr(SiMes2H)[N(SiMe2CH2PPh2)2] (13) To a 15 mL THF solution of {[(Ph2PCH2SiMe2)2N]Cr}2(u.-Cl)2 (1) (0.20 g, 0.16 mmol) at -78 °C was added dropwise a 5 mL THF solution of LiSiMes2H(THF)2 (0.15 g, 0.36 mmol). The blue solution rapidly turned dark green and was warmed to room temperature to yield a red/brown solution. After being stirred for one hour, the THF was removed in vacuo and the residue extracted in hexanes and filtered through Celite to give a red/brown solution. X-ray quality crystals of CrSiMes2H[N(SiMe2CH2PPh2)2] (13) were obtained from a slow evaporation of a hexanes solution. Yield 0.25 g (91%). Anal. Calcd. for C48H59CrNP2Si3: C, 67.97; H, 7.01; N, 1.65. Found: C, 67.22; H, 7.10; N, 1.62. ^ = 5 . 1 B.M. 95 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (xiv) {[(Ph2PCH2SiMe2)2N]Cr}2(p-H)2 (14) A solution of CrMe[N(SiMe2CH2PPh2)2] (5) (0.33 g, 0.57 mmol) in toluene (30 mL) was placed in a thick-walled bomb and placed under 4 atm of H2 (or D2). The brown-red solution turned olive green over one hour and a precipitate began to form. After 12 hours the H2 was removed, the green-black precipitate of {[(Ph2PCH2SiMe2)2N]Cr}2(p-H)2 (14) was collected on a fine frit and washed with cold THF. Yield: 0.25 g (78%). Anal. Calcd. for C3oH37CrNP2Si2: C, 61.94; H, 6.41; N, 2.41. Found: C, 62.09; H, 6.42; N, 2.54. IR: 1368 (v C r-H), (988, v C r -D)-MS: mle 580 (M+-H). peff = 1.6 B.M. In order to obtain crystals for suitable for X-ray diffraction, 0.08 g of 5 was dissolved in 10 mL toluene in a large (125 mL) Erlenmeyer flask and was carefully placed under 1 atm of H2 and left for 4 days. Large cubic X-ray quality crystals slowly deposited and were submitted for structural analysis. (xv) Attempted Synthesis of Cr(CH 2CH 3)[N(SiMe 2CH 2PPh 2) 2] (15) To a 10 mL THF solution of {[(Ph2PCH2SiMe2)2N]Cr}2(u-Cl)2 (1) (0.20 g, 0.32 mmol) was added EtLi (0.012 g, 0.32 mmol) in 2 mL ether dropwise at -78 °C. The dark blue solution turned brown rapidly. After being warmed to room temperature the THF was removed in vacuo to yield a dark brown/green solid which was extracted with 2 mL hexanes to give a brown solution. Brown crystals, along with some green crystals, formed overnight. The brown product was identified by IR spectroscopy as the {[(Ph2PCH2SiMe2)2N]Cr)2(p-H)2 dimer (14). The green product was unidentifiable. (xvi) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with H 2 (D2)/PEt3 To a 10 mL toluene solution of yellow CoMe[N(SiMe2CH2PPh2)2] (6) (0.10 g, 0.17 mmol) in a bomb was added an excess of neat PEt3 (0.1 mL), causing a slight darkening. One atmosphere of H2 (D2) was introduced to the bomb and was stirred overnight to yield an intense orange/brown solution, which was pumped to dryness, extracted with minimum 96 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes hexamethyldisiloxane and filtered through Celite. Orange crystals of CoH(PEt3)[N(SiMe2CH2PPh2)2] (16) formed overnight in the freezer. Yield: 0.080 g (68%). Anal. Calcd. for C36H52C0NP3S12: C, 61.17; H, 7.41; N, 1.98. Found: C, 61.18; H, 7.46; N, 1.82. IH NMR (C 6D 6): 8 7.9 (br, 4H), 5.8 (br, 8H), -0.5 (v br, overlap), -1.7 (v br, overlap). IR: 1784 (1248) cm-1 (vCo-H(D)). u.eff = 2.1 B.M. (xvii) Reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with CO CrMe[N(SiMe2CFf2PPh2)2] (5) (0.050 g) was dissolved in 15 mL of toluene in a bomb to give an orange/red solution, and one atmosphere of CO was added. Immediately the solution turned a darker orange, and over three days of being stirred, a dark green solution formed. The IR spectrum of this solution had multiple Vco bands from 2058 to 1585 cm - 1, indicating formation of multiple products, which could not be separated. Reaction of 5 with one equivalent of CO (gas addition cell, known volume) at -78 °C, warming to room temperature, and removal of solvents and CO after 30 minutes gave an orange compound with a Vco =1841 cm - 1, but no pure solid could be isolated. (xviii) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with CO CoMe[N(SiMe2CH2PPh2)2] (6) (0.10 g, 0.17 mmol) was dissolved in toluene to give a yellow solution. CO (1 atm.) was added and the solution immediately turned orange. After being stirred for two hours the toluene was removed in vacuo to yield an orange solid. IR: 1911, 1829 cm-1 (vco)-(xix) Conditions of reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with ethylene CrMe[N(SiMe2CH2PPh2)2] (5) (0.050 g) was dissolved in 15 mL of toluene in a bomb and charged with one atmosphere of ethylene. No immediate colour change was observed and, upon heating to 80 °C and stirring for 3 weeks, no reaction occurred and starting material was recovered. 97 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (xx) Conditions of reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with ethylene A solution of CoMe[N(SiMe2CH2PPh2)2] (6) (0.02 g, 0.03 mmol) in 10 mL toluene was placed under an atmosphere of ethylene. No reaction was observed over one week at room temperature or at 80 °C. Starting material was recovered. (xxi) Reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with acetylene A solution of 0.02 g CrMe[N(SiMe2CH2PPh2)2] (5) in 20 mL toluene was placed under an atmosphere of acetylene. Rapid formation of a purple solid (presumably polyacetylene) occurred and the starting material was mostly recovered. (xxii) Reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with phenylacetylene To a solution of 0.20 g CrMe[N(SiMe2CH2PPh2)2l (5) in 10 mL toluene was added 0.034 g phenylacetylene. The red/orange solution immediately turned a darker orange. After 30 minutes the solvents were removed, the remaining solid extracted with toluene, filtered and concentrated. Only grainy brown intractable solids were isolated, and no chromium-containing species could be identified by MS. Attempts to form a chromium(II) acetylide by metathesis of the chloride of 1 with LiCCPh formed similar intractable solids. (xxiii) Reaction of CoMe[N(SiMe2CH2PPh2)2l (6) with phenylacetylene To a solution of 0.03 g CoMe[N(SiMe2CH2PPh2)2] (6) in 10 mL toluene was added 0.05 g phenylacetylene (excess). The yellow solution turned a darker orange over fifteen minutes. After being stirred overnight the toluene was concentrated and hexanes added to give a brown/orange precipitate which could not be identified but had a paramagnetic NMR spectrum. The lH NMR spectrum of the total residue indicated oligomer formation due to the large number of peaks from 3.5 to 6.5 ppm, indicative of alkenyl protons. The MS also showed oligomer peaks at 204, 306,408, 510, 612,714 and 816 mle (M+(PhCCH) = 102 mle) 98 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (xxiv) Titration of {[(Ph2PCH2SiMe2)2N]Cr}2(p:-Cl)2 (1) with donor ligands. Toluene solutions of approximately 10 -3 M of 1 in an airtight UV-vis quartz cell were titrated with L = py, CH3CN and THF and CO. The pyridine titration used appropriately concentrated toluene stock solutions of the ligand while the weak donors CH3CN and THF were used neat. The CO titration was accomplished by injecting CO using a gas-tight syringe and calculating the [CO] in solution given the CO solubility in toluene and the solution and head space volumes. Concentration data were collected at the following wavelengths with the given AOD values (change in optical density): py, 400 nm (AOD = 0.97); CH3CN, 320 nm (AOD = 0.57); THF, 336 nm (AOD = 0.45); CO, 392 nm (AOD = 1.40). Equilibrium constants were calculated by the appropriate log ([CrL]/[Cr]) vs. log [L] plot, where the intercept is log K. Alternately, the mid-point of the titration yields K as K=1/[L]. 2.8.5. Variable temperature magnetic susceptibility analyses. This work was done in collaboration with Prof. R.C. Thompson. Magnetic susceptibility measurements on a powdered sample of chromium chloride 1 were made at an applied field of 9225 G over the temperature range 4.6 to 119 K using a PAR Model 155 vibrating-sample magnetometer as previously described.135 For chromium hydride 14, magnetic susceptibilities on a powdered sample over the range 2-340 K and at a field of 10000 G were measured using a Quantum Design (MPMS) SQUID magnetometer. The sample holder, made from PVC was specially designed to possess a constant cross-sectional area. Magnetic susceptibilities were corrected for the diamagnetism of all atoms. The rather low values of susceptibility measured for 14 required that a correction be made for temperature independent paramagnetism in this case (section 2.5(ii)). 99 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes 2.9 References (1) Poli, R. Chem. Rev. 1996, 96,2135. (2) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry; 5th ed.; John Wiley & Sons: New York, 1988. (a) Ch. 18-C; (b) Ch. 28; (c) Ch. 2-1; (d) Ch. 18-F; (e) Ch. 19-G, H; (f) Ch. 24. (3) Davis, R.; Kane-Maguire, L. A. P. In Comprehensive Organometallic Chemistry; G. Wilkinson, F. G. A. Stone and E. W. Abel, Ed.; Pergamon Press: New York, 1982; Vol. 3, Chapter 26.2. (4) Kirtley, S. W. In Comprehensive Organometallic Chemistry; G. Wilkinson, F. G. A. Stone and E. W. Abel, Eds.; Pergammon Press: New York, 1982; Vol. 3, Chapter 26.1. (5) Cotton, F. A.; Eglin, J. L.; Hong, B.; James, C. A. / . Am. Chem. Soc. 1992,114, 4915. (6) Cotton, F. A.; Chen, H.; Daniels, L. M.; Feng, X. / . Am. Chem. Soc. 1992,114, 8980. (7) Edema, J. J. H.; Gambarotta, S. Comments Inorg. Chem. 1991,11, 195. (8) Legzdins, P.; McNeil, W. S.; Shaw, M. J. Organometallics 1993,13, 562. (9) Legzdins, P.; McNeil, W. S.; Batchelor, R. J.; Einstein, F. W. B. / . Am. Chem. Soc. 1994, 116, 6021. (10) Clark, A. Catal. Rev. 1969,3, 145. (11) Karol, F. J.; Karapinka, G. L.; Wu, C ; Dow, A. W.; Johnson, R. N.; Carrick, W. L. J. Polym. Sci., Polym. Chem. Ed. 1973,10, 2621. (12) Karol, F. J.; Brown, G. L.; Davison, J. M. / . Polym. Sci., Polym. Chem. Ed. 1973,11, 413. (13) Theopold, K. H. Acc. Chem. Res. 1990,23,263. (14) Thomas, B. J.; Noh, S. K ; Schulte, G. K ; Sendlinger, S. C ; Theopold, K. H. / . Am. Chem. Soc. 1991,113, 893. (15) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles and Applications of Organotransition Metal Chemistry; University Science Books: Mill Valley, CA, 1987. 100 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (16) Fryzuk, M. D.; Leznoff, D. B.; Rettig, S. J.; Thompson, R. C. Inorg. Chem. 1994,33, 5528. (17) Kern, R. J. / . Inorg. Nucl. Chem. 1962,24, 1105. (18) Fryzuk, M. D.; MacNeil, P. A.; Rettig, S. J.; Secco, A. S.; Trotter, J. Organometallics 1982,7,918. (19) Yahrenkamp,H.Angew. Chem., Int. Ed. Engl. 1978, 77,379. (20) Hermes, A. R.; Girolami, G. S. Inorg. Chem. 1988,27, 1775. (21) Bradley, D. C.; Hursthouse, M. B.; Newing, C. W.; Welch, A. J. / . Chem. Soc, Chem. Commun. 1972, 567. (22) Beagley, B.; Cruickshank, D. W. J.; Monaghan, J. J.; Aylett, B. J.; Ellis, I. A. / . Chem. Soc, Chem. Commun. 1968, 909. (23) Evans, D. F. / . Chem. Soc. 1959, 2003. (24) Sur, S. K. / . Magn. Res. 1989, 82, 169. (25) Carlin, R. L. Magnetochemistry; Springer-Verlag: Heidelberg, 1986. (26) Cotton, F. A.; Faut, O. D.; Goodgame, D. M. L.; Holm, R. H. / . Am. Chem. Soc. 1961, 83, 1780. (27) Horrocks, W. D.; Hecke, G. R. V.; Hall, D. D. Inorg. Chem. 1967, 6, 694. (28) Reinaud, O. M.; Rheingold, A. L.; Theopold, K. H. Inorg. Chem. 1994,33, 2306. (29) Carlin, R. L.; Chirico, R. D.; Sinn, E.; Mennenga, G.; de Jongh, L. J. Inorg. Chem. 1982, 27,2218. (30) Bradley, D. C ; Hursthouse, M. B.; Smallwood, R. J.; Welch, A. J. / . Chem. Soc, Chem. Commun. 1972, 872. (31) Shannon, R. D.; Prewitt, C. T. Acta. Cryst. B 1969, B25, 925. (32) Shannon, R. D. Acta. Cryst. B 1976, A32, 751. (33) Burger, H.; Wannagat, U. Monatsch. Chem. 1963, 94, 1007. (34) Murray, B. D.; Power, P. P. Inorg. Chem. 1984,23, 4584. (35) Baitlett, R. A.; Power, P. P. J. Am. Chem. Soc. 1987, 709, 7563. 101 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (36) Fryzuk, M. D.; Leznoff, D. B.; Rettig, S. J. Organometallics 1995,14, 5193. (37) Hermes, A. R.; Morris, R. J.; Girolami, G. S. Organometallics 1988, 7, 2372. (38) Hao, S.; Song, J.; Berno, P.; Gambarotta, S. Organometallics 1994,13, 1326. (39) Hao, S.; Song, J.; Aghabozorg, H.; Gambarotta, S. / . Chem. Soc, Chem. Commun. 1994, 157. (40) Edema, J. J. H.; Gambarotta, S.; Meetsma, A.; Spek, A. L.; Smeets, W. J. J.; Chiang, M. Y. / . Chem. Soc, Dalton Trans. 1993, 789. (41) Edema, J. J. H.; Gambarotta, S.; van Bolhuis, F.; Spek, A. L. / . Am. Chem. Soc. 1989, 111, 2142. (42) Hao, S.; Gambarotta, S.; Bensimon, C. / . Am. Chem. Soc. 1992,114, 3556. (43) Morse, P. M.; Spencer, M. D.; Wilson, S. R.; Girolami, G. S. Organometallics 1994,13, 1646. (44) Cotton, F. A.; Hanson, B. E.; Ilsley, W. H.; Rice, G. W. Inorg. Chem. 1979,18, 2713. (45) Krausse, J.; Marx, G.; Schodl, G. / . Organomet. Chem. 1970,21, 159. (46) Krausse, J.; Schodl, G. / . Organomet. Chem. 1971,27, 59. (47) Hursthouse, M. B.; Malik, K. M. A.; Sales, K. D. / . Chem. Soc, Dalton Trans. 1978, 1314. (48) Koschmieder, S. U.; McGilligan, B. S.; McDermott, G.; Arnold, J.; Wilkinson, G.; Hussain-Bates, B.; Hursthouse, M. B. / . Chem. Soc, Dalton Trans. 1990, 3427. (49) Heintz, R. A.; Ostrander, R. L.; Rheingold, A. L.; Theopold, K. H. / . Am. Chem. Soc. 1994,116,11387. (50) Girolami, G. S.; Salt, J. E.; Wilkinson, G. / . Am. Chem. Soc. 1983,105, 5954. (51) Girolami, G. S.; Wilkinson, G.; Galas, A. M. R.; Thornton-Pett, M.; Hursthouse, M. B. / . Chem. Soc, Dalton Trans. 1985, 1339. (52) Brookhart, M.; Green, M. L. H.; Wong, L. L. Prog. Inorg. Chem. 1988,36, 1. (53) Allen, F. H.; Kennard, O.; Taylor, R. Acc. Chem. Res. 1983,16, 146. (54) Klein, H.-R; Karsch, H. H. Chem. Ber. 1976,109, 1453. 102 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (55) Chart, J.; Shaw, B. L. / . Chem. Soc. 1961, 285. (56) Falvello, L.; Gerloch, M. Acta. Cryst. B 1979, B35, 2547. (57) Owston, P. G.; Rowe, J. M. J. Chem. Soc. 1963, 3411. (58) Theopold, K. H.; Silverstre, J.; Byrne, E. K.; Richeson, D. S. Organometallics 1989, 8, 2001. (59) Hay-Motherwell, R. S.; Wilkinson, G.; Hussain, B.; Hursthouse, M. B. Polyhedron 1990, 9, 931. (60) NMR of Paramagnetic Molecules; Lamar, G. N.; Horrocks, W. D.; Holm, R. H., Eds.; Academic Press, Inc.: New York, 1973. (61) Miiller, G.; Neugebauer, D.; Geike, W.; Kohler, F. H.; Pebler, J.; Schmidbaur, H. Organometallics 1983,2, 257. (62) Legzdins, P.; Jones, R. H.; Phillips, E. C ; Yee, V. C ; Trotter, J.; Einstein, F. W. B. Organometallics 1991,10, 986. (63) Dryden, N. H.; Legzdins, P.; Phillips, E. C ; Trotter, J.; Yee, V. C. Organometallics 1990, 9, 882. (64) Jordan, R. F.; LaPointe, R. E.; Baenziger, N.; Hinch, G. D. Organometallics 1990, 9, 1539. (65) Hursthouse, M. B.; Malik, K. M. A.; Sales, K. D. / . Chem. Soc, Dalton Trans. 1978, 1314. (66) Bhandari, G.; Kim, Y.; McFarland, J. M.; Rheingold, A. L.; Theopold, K. H. Organometallics 1995,14,738. (67) Glocking, F.; Sneeden, R. P. A.; Zeiss, H. H. / . Organomet. Chem. 1966,2, 109. (68) Richeson, D. S.; Mitchell, J. F.; Theopold, K. H. Organometallics 1989, 8, 2570. (69) Sneeden, R. P. A.; Throndsen, H. P. / . Chem. Soc, Chem. Commun. 1965, 509. (70) Sneeden, R. P. A.; Throndsen, H. P. / . Organomet. Chem. 1966, 6, 542. (71) Espenson, J. H. Acc. Chem. Res. 1992,25, 222. (72) Shi, S.; Espenson, J. H.; Bakac, A. / . Am. Chem. Soc. 1990,112, 1841. 103 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (73) Coles, M. P.; Dalby, C. I.; Gibson, V. C ; Clegg, W.; Elsegood, M. R. J. / . Chem. Soc, Chem. Commun. 1995, 1709. (74) Davies, G. R.; Jarvis, J. A. J.; Kilbourn, B. T. / . Chem. Soc, Chem. Commun. 1971, 1511. (75) Gordon, K. R.; Warren, K. D. Inorg. Chem. 1978,17, 987. (76) Barron, A. R.; Salt, J. E.; Wilkinson, G.; Motevalli, M.; Hursthouse, M. B. / . Chem. Soc, Dalton Trans. 1987, 2947. (77) Salt, J. E.; Girolami, G. S.; Wilkinson, G.; Motevalli, M.; Thornton-Pett, M.; Hursthouse, M. B. / . Chem. Soc, Dalton Trans. 1985, 685. (78) Salt, J. E.; Wilkinson, G.; Motevalli, M.; Hursthouse, M. B. / . Chem. Soc, Dalton Trans. 1986, 1141. (79) Keogh, D. W.; Poli, R. / . Am. Chem. Soc. 1997,119, 2516. (80) Fettinger, J. C ; Mattamana, S. P.; Poli, R.; Rogers, R. D. Organometallics 1996,15, 4211. (81) Zerner, M. C ; Loew, G. H.; Kirchner, R. R ; Mueller-Westerhoff, U. T. / . Am. Chem. Soc. 1980,102, 589. (82) Anderson, W. P.; Edwards, W. D.; Zerner, M. C. Inorg. Chem. 1986,25, 2728. (83) Anderson, W. P.; Cundari, T. P.; Drago, R. S.; Zerner, M. C. Inorg. Chem. 1990,29, 1. (84) Tilley, T. D. In The Silicon-Heteroatom Bond; S. Patai and Z. Rappoport, Ed.; John Wiley & Sons: New York, 1991; pp 245. (85) Schumann, H.; Nickel, S.; Hahn, E.; Heeg, M. J. Organometallics 1985,4, 800. (86) Schumann, H.; Meese-Marktscheffel, J. A.; Hahn, F. E. / . Organomet. Chem. 1990, 390, 301. (87) Schumann, H.; Nickel, S.; Loebel, J.; Pickardt, J. Organometallics 1988, 7, 2004. (88) Schumann, H.; Albrecht, I.; Gallagher, M.; Hanh, E.; Janiak, C ; Kolax, C ; Loebel, J.; Nickel, S.; Palamidis, E. Polyhedron 1988, 7,2307. (89) Arnold, J.; Tilley, T. D.; Rheingold, A. L.; Geib, S. J. Organometallics 1987, 6, 473. 104 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (90) Arnold, J.; Shina, D. N.; Tilley, T. D.; Arif, A. M. Organometallics 1986,5, 2037. (91) Samuel, E.; Mu, Y.; Harrod, J. F.; Dromzee, Y.; Jeannin, Y. / . Am. Chem. Soc. 1990, 112, 3435. (92) Roddick, D. M.; Tilley, T. D.; Rheingold, A. L.; Geib, S. J. / . Am. Chem. Soc. 1987,109, 945. (93) Probst, R.; Leis, C ; Gamper, S.; Herdtweck, E.; Zybill, C ; Auner, N. Angew. Chem., Int. Ed. Engl. 1991,50,1132. (94) Handwerker, F£.; Leis, C.; Probst, R.; Bissinger, P.; Grohmann, A.; Kiprof, P.; Herdtweck, E.; Bliimel, J.; Auner, N.; Zybill, C. Organometallics 1993,12, 2162. (95) Handwerker, H.; Paul, M.; Bliimel, J.; Zybill, C. Angew. Chem., Int. Ed. Engl. 1993, 32, 1313. (96) Zybill, C.; Midler, G. Organometallics 1988, 7, 1368. (97) Leis, C.; Wilkinson, D. L.; Handwerker, H.; Zybill, C.; Miiller, G. Organometallics 1992, 77,514. (98) Handwerker, H.; Paul, M.; Riede, J.; Zybill, C. J. Organomet. Chem. 1993,459, 151. (99) Glavee, G. N.; Jagirdar, B. R.; Schneider, J. J.; Klabunde, K. J.; Radonovich, L. J.; Dodd, K. Organometallics 1992,11, 1043. (100) Jagirdar, B. R.; Palmer, R.; Klabunde, K. J.; Radonovich, L. J. Inorg. Chem. 1995,34, 278. (101) Schubert, U.; Miiller, J.; Alt, H. G. Organometallics 1987, 6, 469. (102) Schubert, U. Adv. Organomet. Chem. 1990,30, 151. (103) McAlexander, L. H.; Hung, M.; Li, L.; Diminnie, J. B.; Xue, Z.; Yap, G. P. A.; Rheingold, A. L. Organometallics 1996, 75, 5231. (104) Ebsworth, E. A. V.; Rankin, D. W. H.; Cradock, S. Structural Methods in Inorganic Chemistry; 2nd ed.; Blackwell Scientific Publications: Oxford, 1991. (105) Heintz, R. A.; Haggerty, B. S.; Wan, H.; Rheingold, A. L.; Theopold, K. H. Angew. Chem., Int. Ed. Engl. 1992,31, 1077. 105 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (106) Heintz, R. A.; Koetzle, T. F.; Ostrander, R. L.; Rheingold, A. L.; Theopold, K. H.; Wu, P. Nature 1995,378, 359. (107) Willet, R. D.; Gatteschi, D.; Kahn, O. Magneto-Structural Correlations in Exchange Coupled Systems; Reidel: Dordrecht, The Netherlands, 1985. (108) Miller, J. S.; Epstein, A. J. Angew. Chem., Int. Ed. Engl. 1994, 33, 385. (109) Miller, J. S.; Dougherty, D. A. Mol. Cryst. Liq. Cryst. 1989,176, 1. (110) Kahn, O. Molecular Magnetism; VCH: New York, 1993. (111) Solomon, E. I. In Metal Clusters in Proteins; J. L. Que, Ed.; American Chemical Society: Washington DC, 1988. (112) Moron, M. C ; Palacio, F.; Pons, J.; Casabo, J.; Solans, X.; Merabet, K. E.; Huang, D.; Shi, X.; Teo, B. K.; Carlin, R. L. Inorg. Chem. 1994, 33,746. (113) Kahn, O. Angew. Chem., Int. Ed. Engl. 1985, 24, 834. (114) Kato, M.; Muto, Y. Coord. Chem. Rev. 1988, 92, 45. (115) Bianchini, C ; Laschi, F.; Masi, D.; Mealli, C ; Meli, A.; Ottaviani, F. M.; Proserpio, D. M.; Sabat, M.; Zanello, P. Inorg. Chem. 1989, 28, 2552. (116) Fryzuk, M. D.; Ng, J. B.; Rettig, S. J.; Huffman, J. C ; Jonas, K. Inorg. Chem. 1991, 30, 2437. (117) Otieno, T.; Rettig, S. J.; Thompson, R. C ; Trotter, J. Inorg. Chem. 1993, 32, 4384. (118) O'Connor, C. J. Prog. Inorg. Chem. 1982,29, 203. (119) Mabbs, F. E.; Machin, D. S. Magnetism and Transition Metal Complexes; Chapman and Hall: London, 1973. (120) Leed, D. H.; Machin, D. J. / . Chem. Soc, Chem. Commun. 1974, 866. (121) Larkworthy, L. F.; Trigg, J. K ; Yavari, A. / . Chem. Soc, Dalton Trans. 1975, 1879. (122) Gregson, A. K.; Day, P.; Leech, D. H.; Fair, M. J.; Gardner, W. E. / . Chem. Soc, Dalton Trans. 1975,1306. (123) Edema, J. J. H.; Gambarotta, S.; Smeets, W. J. J.; Spek, A. L. Inorg. Chem. 1991, 30, 1380. 106 References begin on page 100 Chapter 2: Cr(II) and Co(II) Halide and Alkyl Complexes (124) Edema, J. J. H.; Gambarotta, S.; Spek, A. L. Inorg. Chem. 1989,28, 812. (125) Davidson, P. J.; Lappert, M. F.; Pearce, R. Chem. Rev. 1976, 76, 219. (126) Heintz, R. A.; Neiss, T.-G.; Theopold, K. H. Angew. Chem., Int. Ed. Engl. 1994,33, 2326. (127) Sanders, J. R. / . Chem. Soc, Dalton Trans. 1973, 748. (128) Albertin, G.; Amendola, P.; Antoniutti, S.; Bordignon, E. / . Chem. Soc, Dalton Trans. 1990, 2979. (129) Heintz, R. A.; Ostrander, R. L.; Rheingold, A. L.; Theopold, K. H. / . Am. Chem. Soc. 1994,116, 11387. (130) Fryzuk, M. D.; MacNeil, P. A. / . Am. Chem. Soc. 1984,106, 6993. (131) Schut, D. M.; Keana, K. J.; Tyler, D. R.; Rieger, P. H. / . Am. Chem. Soc. 1995,117, 8939. (132) Fryzuk, M. D.; MacNeil, P. A.; Rettig, S. J.; Secco, A. S.; Trotter, J. Organometallics 1982,7,918. (133) Schlosser, M.; Ladenberger, V. / . Organomet. Chem. 1967, 8, 193. (134) Bryce-Smith, D.; Turner, E. E. / . Chem. Soc. 1953, 861. (135) Haynes, J. S.; Oliver, K. W.; Rettig, S. J.; Thompson, R. C ; Trotter, J. Can. J. Chem. 1984, 62, 891. 107 References begin on page 100 Chapter Three One-Electron Oxidation Reactions of Cr(II) Complexes 3.1 Introduction Oxidative addition of a substrate to a metal complex is an important process in organometallic chemistry.1 , 2 The most commonly studied oxidative addition reactions involve formal two-electron redox processes in which the oxidation state of the metal centre is increased by two (equation 3.1). Typical examples include the oxidation of Pdi(I)/Ir(I) to M(III), Ni/Pd/Pt(0) to M(II) or Pd/Pt(II) to M(IV). 1 - 2 In particular the synergy between 16- and 18-electron systems in catalytic cycles has been a driving force for the emphasis of two-electron redox processes.1,3a L n M x + X 2 LnM (X)2 [3.1] 108 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(IJ) Complexes One-electron processes, on the other hand, have been less studied in an organometallic context.1 However, divalent first-row transition metals generally undergo one-electron redox processes to yield M(III) systems, and not M(IV) complexes (equation 3.2).1 ,3b Classical electron-transfer studies involved, in particular, the examination of Cr(II/III), Fe(II/III) and Co(II/ni) redox couples.4 Hence, the complexes prepared in Chapter 2 are ideal candidates for an exploration of one-electron redox reactions. This chapter will focus on one-electron redox reactions with chromium(II) and Chapter 4 will deal with cobalt(II) systems. 2 L n M x + X 2 ^ 2 L n M x + 1 X [3.2] 3.2 Reaction of Cr(II) Chloride Complex 1 with AgBPh4 (i) Synthesis and structure of CrCl2(THF)[N(SiMe2CH2Ph2)2] (19) Silver salts are commonly used as halide abstraction agents due to the insolubility of the AgX produced. However, silver salts are also known as oxidizing agents for metal complexes. Hence the addition of silver(I) to the chromium(II) chloride complex 1 could yield either the chromium(III) cation {CrCl(solvent)n[N(SiMe2CH2Ph2)2])+ (A) via oxidation of the chromium(II) centre or the chromium(II) cation {Cr(solvent)n[N(SiMe2CH2Ph2)2]}+ (B) by halide abstraction. Addition of one equivalent of AgBPh4 (BPh4 is a non-coordinating, bulky anion) to a THF solution of {[(Ph2PCH2SiMe2)2N]2Cr}2(|i-Cl)2 (1) gave a slow change to a purple solution and the concomitant observation of silver metal precipitate. The presence of Ag° suggested that a redox product had formed, but workup yielded a product soluble in toluene; a cationic product such as A or B would not be so soluble. The final purple product, isolated in 40% yield, was identified as the non-cationic chromium(III) product CrCl2(THF)[N(SiMe2CH2Ph2)2] (19) (equation 3.3). 109 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes Ph 2 Ph 2 •p Me2Si C I Me2Si THF : N — C r THF +Ag o [3.3] Me2Si THF A 9 B P h 4 M e 2 S i ^ c i •P Ph 2 1 19 The reaction is not particularly clean; although the amounts and ratios of products formed are unknown, the mass spectrum of the product mixture includes peaks attributable to 19, Cr(Ph)Cl[N(SiMe2CH2Ph2)2] (20) (mle 696), CrPh[N(SiMe2CH2Ph2)2] (21) (mle 657) and to BPh3, which is also easily observable as a white solid. It is proposed that the formation of 19 occurs via the chromium(III) chloride cation {CrCl(solvent)n[N(SiMe2CH2Ph2)2]}+{(BPh4)}- (A), which is apparently unstable under these reaction conditions. Two concurrent decomposition pathways can account for the observed product mixture (Scheme 3.1). Firstly, the cation can abstract a phenyl group from the BPI14 anion to yield the observed chromium(III) aryl halide product 20, as well as BPh3. Alternatively, the cation can abstract a halide from the starting material to yield the product dichloride 19 and the chromium(II) cation B, which reacts with the counterion to yield CrPh[N(SiMe2CH2Ph2)2] (21) and BPh3. Activation of a BPI14- counterion by a chromium(III) cation has been previously observed.5 The chromium(III) dichloride 19 was the only isolable product from this reaction; its X-ray crystal structure is shown in Figure 3.1. The room temperature solution magnetic moment (by Evans' method6,7) of 19 is 3.5 B.M., consistent with a high-spin d 3 Cr(III) centre. The vast majority of chromium(III) complexes are octahedral.30,8 The amidodiphosphine ligand is bound in a meridional fashion; the two phosphines and two chlorides are both trans-oriented; the P(l)-Cr-110 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes P(2) and Cl(l)-Cr-Cl(2) bond angles of 167.49(3)° and 172.76(4)° illustrate this point. The other trans-angle is between the amide and the THF molecule; the N(l)-Cr-0(1) bond angle is 179.43(10)°. Scheme 3.1 [PNP]Cr(Ph)Cl + B P h 3 20 {[PNP]CrCI(S)n}+BPh4- (A) • {[PNP]Cr(S)n}+BPh4- (B) + 19 [PNP]CrCI 1 [PNP]CrPh + B P h 3 21 The Cr(HI)-P bond lengths (Table 3.1) of 2.491(1) and 2.487(1) A are typical of high-spin Cr(III)-P bonds. Other examples include Cr-P bonds ranging from 2.429(1) to 2.444(1) A in [CrCl{N(CH2CH2PMe2)2}2],9 2.414(2) A in [CpCrCl2](dppe)10 (dppe = Ph 2PCH 2CH 2PPh 2) and 2.426(2) A in Cp*CrMe2(PMe3).u The Cr-Cl bond lengths of 2.3248(9) and 2.3254(9) A in 19 are slightly longer than typical high-spin Cr(III)-Cl bond lengths. For example, the Cr-Cl bond lengths in CpCrCl 2 ( T l 1 - P M e 2 C H 2 P M e 2 ) are 2.281(2) and 2.295(2) A , 1 0 in [CrCl{N(CH2CH2PMe2)2}2] it is a long 2.401 A (due to a trans-amide being present),9 and 2.303-2.325 A in CrCl3[H2N(CH2)2NH(CH2)2NH2].12 The Cr-N bond length in 19 of 2.023(3) A is relatively long compared to other Cr(III)-amide bond lengths, examples including 1.996(2) and 2.017(2) A in [CrCl{N(CH 2CH 2PMe 2)2}2L 9 1.932(3) and 1.931(3) A in Cp*Cr(quinolinediamide),13 and a very short 1.87 A in Cr(N'Pr2)3.14 Finally, the Cr-THF bond length of 2.143(2) A in dichloride 19 can be compared with the shorter 2.042(5) and 2.043(5) A in cationic [Cp*Cr(THF)2Me]+BPh4,-5 with 2.123(3) A in Cr(CO)5(THF)1 5 and with 2.046(8), 2.044(7) and 2.214(7) A in (p-CH3C6H4)CrCl2(THF)3.16 Note the two different Cr-THF bond 111 References begin on page 147 Table 3.1 Chapter 3: One-Electron Oxidation Reactions ofCr(H) Complexes Selected bond lengths and angles in CrCl2(THF)[N(SiMe2CH2PPh2)2] (19). Atom Atom Distance (A) Atom Atom Distance (A) Cr P(l) 2.491(1) Cr Cl(l) 2.3248(9) Cr P(2) 2.487(1) Cr Cl(2) 2.3254(9) Cr N(l) 2.023(3) Cr CKD 2.143(2) N(l) Si(l) 1.721(3) N(l) Si(2) 1.721(3) Atom Atom Atom Angle (°) Atom Atom Atom Angle (°) Cl(l) Cr Cl(2) 172.76(4) P(l) Cr P(2) 167.49(3) O(l) Cr N(l) 179.43(10) Si(l) N(l) Si(2) 123.0(1) Cr O(l) C(31) 126.4(2) Cr O(l) C(34) 125.7(2) C(31) Oil) C(34) 107.8(2) Cr N(l) Si(l) 118.6(1) C25 Figure 3.1 Molecular structure (ORTEP) and numbering scheme for CrCl2(THF)[N(SiMe2CH2PPh2)2] (19). 112 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes lengths in the latter complex; the long bond corresponds to the THF trans to the alkyl group. As has been observed before,15"17 the bond angles around the coordinated oxygen of 126.4(2)°, 125.7(2)° and 107.8(2)° (Cr-0(1)-C(31), Cr-0(1)-C(34) and C(31)-0(l)-C(34) respectively) in 19 indicate a trigonal-planar oxygen, although it has been noted that this is not sufficient evidence to establish THF-metal 7t-donation.18 (ii) Synthesis of a cyclopentadienyl Cr(iii) complex (22) The metathesis of one chloride in CrCl2(THF)[N(SiMe2CH2PPh2)2] (19) by NaCp»DME causes a rapid colour change from purple to dark green. After workup, green CrCpCl[N(SiMe2CH2PPh2)2] (22) was isolated in high yield (equation 3.4). M © v Me [3.4] 19 Complex 22 has a solution magnetic moment of 3.6 B.M., consistent with a high-spin Cr(III) centre. Note that the high-spin nature of 22 implies that either one phosphine is not bound or the Cp is T]3-bound; a fully ligated complex would be a 17-electron species and low-spin.19,20 The similar complexes CpCrCl2(P-P) (P-P = dmpm, dmpe and dppe) were determined to be octahedral 15-electron high-spin d 3 complexes with r(5-Cp ligands and dangling phosphines.10 The structure of 22 is hence proposed to be similar to this. The lack of adduct formation has been 113 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes attributed to the barrier associated with electron pairing upon binding of an extra phosphine.20 Unlike the CpCrCl2(P-P) complexes, no 3 1P{ lH) NMR spectrum was observed for 22, possibly a result of fast exchange of phosphine. In addition, no resonances were observed in either the NMR or ESR spectra of 22. In the absence of a crystal structure, a definitive structural assignment is difficult, but a piano stool structure is highly probable. 3 . 3 Reaction of Chromium(II) Methyl Complex 5 with PhSSPh One type of reagent that easily undergoes one-electron oxidation reactions with transition metals is a disulfide, RSSR. 2 1 The S-S bond in diphenyldisulfide in particular is known to be easily cleaved, resulting in the formation of PhS» radicals. Complexes which undergo this formal one-electron oxidation with PhSSPh include [CpM(CO)3]2 (M = C r , 2 2 - 2 3 Mo , 2 2 W 2 4 , 2 5 ) , [Cp*Cr(CO) 3] 2 2 6 and W0Pr3P)2(CO)3.2 7 Accordingly, addition of one half an equivalent of PhSSPh to a brown solution of CrMe[N(SiMe2CH2PPh2)2] (5) at 0 °C resulted in a rapid colour change to purple; reaction at room temperature yielded a dark brown solution from which no product could be isolated. After workup, a solid with the empirical formula {CrMe(SPh)[N(SiMe2CH2PPh2)2]}x (23) was obtained (equation 3.5). Ph 2 [3.5] Ph 2 5 23 114 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes Compound 23 was expected to be dinuclear, in part due to the overwhelming prevalence of octahedral chromium(III) geometry and also due to the high propensity of thiolates, in particular phenylthiolates, to bridge metal centres. Examples of bridging chromium.thiolates include [CpCr(CO)2(SPh)]2 and [CpCr(SPh)]2S,23 Cp2Cr2(u.-SPh)(p3-S)2FeCp and CpCr(u.-SPh)3Fe(u-SPh)3CrCp,28 and Cp2W(p-SPh)Cr(CO>4.29 However, the solution magnetic moment (Evans' method6'7) of 3.8 B.M. was consistent with a mononuclear high-spin Cr(III) complex. The mass spectrum indicated only a monomer fragment at mle 689 (M + - Me) and the *H NMR spectrum consisted of a series of very broad, paramagnetically shifted resonances from which no structural information could be ascertained. An X-ray crystal structure of thiolate 23 was obtained in order to determine the nuclearity of the system. The X-ray crystal structure, shown in Figure 3.2, revealed that thiolate 23 was, in fact, a //ve-coordinate chromium(III) complex, with a terminal phenylthiolate ligand. However, the geometry is difficult to assign to a standard five-coordinate arrangement. For example, the complex could be considered as a distorted square-pyramid, in which the methyl group, C(37), occupies the apical position. The trans angles of the square base are then described by P(l)-Cr-P(2) and S(l)-Cr-NQ), which are 166.30(6)° and 150.0(1)° respectively. Alternatively, a distorted trigonal-bipyramidal geometry can be considered, in which the phosphines are axial and the equatorial angles are defined by S(l)-Cr-C(37), S(l)-Cr-N(l) and C(37)-Cr-N(l). These angles are 92.8(2)°, 150.0(1)° and 117.2(2)° respectively, obviously substantially distorted from the ideal 120° for trigonal-bipyramidal coordination. An empirical parameter x, has been defined in order to assist in the description of distorted five-coordinate compounds.30 The geometric parameter x = ([3 - a)/60, where (3 and a are the two largest angles in the structure, and which do not incorporate the putative apex of a square-pyramidal structure. Values of x range from 0-1, where x = 0 corresponds to a square-pyramidal structure (the possibility of the metal being raised out of the plane is accounted for) and x = 1 corresponds to a trigonal-bipyramidal structure. Intermediate values describe the continuum of 115 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(ll) Complexes structures between these two extremes. The use of this parameter to describe distorted five-coordinate complexes, while perhaps instructive, is a rather simplistic method of characterization. One primary concern of the use of the x value is that it doesn't account well for distortions in the equatorial plane of a trigonal-bipyramid. In any case, the r value for thiolate 23 is 0.26, which implies that a distorted square-pyramid is a reasonable description of the structure, at least using this criterion. C23 Figure 3.2 Molecular structure (ORTEP) and numbering scheme for CrMe(SPh) [N(SiMe 2CH 2PPh 2)2] (23). The few structurally characterized examples of five-coordinate chromium(III) complexes are trigonal-bipyramidal CrCl3 (NMe3) 2 , 3 1 ' 3 2 distorted trigonal-bipyramidal Na 2CrPh5»3Et 20«THF, 3 3 square-pyramidal Cr(tmtaa)Cl (tmtaa = tetramethyl-dibenzotetraaza[14]annulene) and recently, two-legged piano-stool [r| 5-Me4C5SiMe 2-r) 1-N f B u ] C r C H 2 S i M e 3 . 3 4 Other non-octahedral complexes of chromium(III) include trigonal 116 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes C r ( N * P r 2 ) 3 1 4 and Cr[N(SiMe3 ) 2] 3 3 5 and trigonal-monopyramidal Cr[(^uMe 2Si)NCH 2CH 2] 3N]. 3 6 Table 3.2 Selected bond lengths and angles in CrMe(SPh)[N(SiMe2CH2PPh2)2] (23). Atom Atom Distance ( A ) Atom Atom Distance ( A ) Cr P(l) 2.449(2) Cr S ( D 2.371(2) Cr P(2) 2.479(2) Cr C(37) 2.054(5) Cr N(l) 2.017(4) S(l) C(31) 1.771(7) N(l) Si(l) 1.732(4) N(l) Si(2) 1.731(4) Atom Atom Atom Angle O Atom Atom Atom Angle O S(l) Cr N(l) 150.0(1) P(l) Cr P(2) 166.30(6) S(l) Cr P(l) 93.90(7) S ( D Cr P(2) 99.76(6) S(l) Cr C(37) 92.8(2) P(l) Cr N(l) 85.6(1) P(l) Cr C(37) 88.4(2) P(2) Cr N(l) 83.1(1) P(2) Cr C(37) 89.9(2) N(l) Cr C(37) 117.2(2) Cr S(l) C(31) 101.9(2) Si(l) N(l) Si(2) 118.6(2) The Cr-P bond lengths of 2.449(2) and 2.479(2) A and the Cr-N bond length of 2.017(4) A in thiolate 23 are very similar to those found in the octahedral dichloride 19. Note that the change in coordination number and geometry does not affect these bond lengths to any great extent. The Cr-S bond length of 2.371(2) A is unremarkable and can be compared with terminal Cr(III)-S bonds of 2.389(5) A in [(en)Cr(SCH2CH2NH2)2]C104 3 7 2.364(5) A (average) in (PPh4)Na[Cr3(SCH2CH20)6]38 and 2.396(2) A (average) in Cr(CS2NEt2)3.3 9 Interestingly, the Cr(III)-S bond lengths in bridging phenylthiolates are not that different. The Cr-S bond lengths in 117 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes Cp2Cr2(p-SPh)(p3-S)2FeCp and CpCr(p-SPh)3Fe(u-SPh)3CrCp28 range from 2.336(3) to 2.383(8) A and in [CpCr(p-SPh)]2S the range is from 2.365(1) to 2.383(1) A . 2 3 Note that the Cr-S(l)-C(31) bond angle of 101.9(2)° implies that the second lone-pair on the thiolate is not interacting with the metal; a bond angle of closer to 180° would be expected in that situation. The Cr(III)-C(37) bond length of 2.054(5) A is quite short and can be compared to Cr(ffl)-Me bond lengths of 2.09(2) and 2.14(2) A in octahedral, neutral CrMe3[fBuSi(CH2PMe2)3],40 2.067(5) A in Cp*CrMe(PMe 3), n 2.073(3) A in [CpCrMe]2(p-Cl)241 and 2.087(2) A in [Cp*CrMe]2(p-Cl)242 A very long Cr-C bond length of 2.300(15) A in Li 3CrMe 6»3 C4H 8 0 2 can be attributed to the trianionic nature of the complex.43 The only other comparable Cr(III)-C bond length can be found in the cationic complex [Cp*CrMe(TFfF)2]BPh4, with a Cr-C bond length of 2.056(8) A; 5 it was noted that this was the shortest Cr-CMe bond length observed in a series of complexes prepared in the Theopold laboratory.44 The electronically unsaturated nature of complex 23, due to its pentacoordination, could be a factor in rationalizing this short bond length. The mechanism of formation of thiolate 23 was not examined. The Cr(I) radical Cp*Cr(CO)3» was shown to react with PhSSPh by attack of the chromium radical on the disulfide to yield Cp*Cr(SPh)(CO)3 and thiolate radicals (»SPh) 2 3 , 2 6 Diphenyldisulfide was also shown to react with Cp2Ta(p>CH2)2CoCp via initial attack of the cobalt(II) centre on the disulfide; in the same paper, PhSSPh was shown to react with cobaltocene, Cp2Co, via an outer-sphere electron transfer mechanism.45 Note that although chromium(III) thiolates have been structurally characterized, it appears that this is the first example of the one-electron oxidation of a chromium(n) complex with disulfides. This is particularly significant as the starting chromium(II) complex contains a metal-alkyl group and both the starting material and product have been structurally characterized. Finally, this reaction confirms that the chromium(II) complex CrMe[N(SiMe2CH2PPh2)2] (5) can undergo one-electron oxidation reactions with an appropriate substrate. 118 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes 3.4 Reaction of Chromium(II) Complexes with Alkyl Halides (i) Introduction One very important group of substrates that undergoes oxidative addition reactions is alkyl halides.1 Two-electron oxidative addition of substrates such as Mel to metal complexes L n M x to give LnMx+2(Me)(I) have been well studied.1'2 Chromium(II) and cobalt(II) complexes, however, are more likely to undergo one-electron oxidative addition, and this is indeed observed for reactions with alkyl halides. Reaction of [Cr(H20)6]2+ with alkyl halides and other radical sources to generate aqueous organometallic Cr(III) cations have been well studied.46^8 Although there are little structural data, mechanistic and kinetic data abound49"52 and all support a radical-based atom abstraction mechanism,53,54 shown in Scheme 3.2. Scheme 3.2 LnM(ll) + RX LnM(lll)X + R-LnM(ll) + R« - LnM(lll)R 2L nM(ll) + RX LnM(lll)R + LnM(lll)X The reaction of chromium(II) complexes with alkyl halides is thus expected to give two products, the Cr(III) alkyl and the Cr(III) halide complex. The stoichiometry of the reaction is expected to be two moles of chromium(II) complex per one mole of alkyl halide. As expected, the reaction of [Cr(H20)6]2+ with alkyl halides RX yields [CrR(H20)5]2+ and [CrX(H 20) 5] 2 + , as shown in equation 3.6.47'48 2 [Cr(H 2 0) 6 ] 2 + + RX [CrR(H 2 0) 5 ] 2 + + [CrX(H 2 0) 5 ] 2 + + 2 H 2 0 [3.6] 119 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes More recently, 17-electron Cr(I) radical reactions with alkyl halides to give Cr(U) products have also been examined.55"58 In this case, the chromium starting material, commonly a metal-metal bonded dimer such as [CpCr(CO)3]2, reacts cleanly only in cases where the metal-centred radicals are stable with respect to dimerization. Detailed mechanistic studies utilizing other stable 17-electron radicals such as that produced from flash photolysis of [CpM(CO)3] (M = Mo, W) 5 9 and Re(CO)4L60 have been reported. In every case studied, however, M L n has been a coordinatively saturated octahedral complex (n = 6), as have been the metal-containing products. There are also no examples of organometallic Cr(II)Ln reactants which undergo this reaction. The square-planar, high-spin Cr(II) mesityl complex Cr(C6H2Me3)2(PMe3)2 reacts with methyl iodide not to give one-electron oxidation chromium(III) products but rather a substitution with the Cr-R fragment to give organic products (C6H2Me4> and C1I2.61 The chromium(II) complex CrMe[N(SiMe2CH.2PPh2)2] (5) prepared in Chapter 2 has been shown to undergo one-electron oxidative addition with PhSSPh. The reactivity of this highly coordinatively and electronically unsaturated square planar Cr(II) organometallic complex with alkyl halides was therefore examined. (ii) Reaction of CrMe[N(SiMe 2CH2PPh2) 2] (5) with methyl iodide and bromide A red-brown toluene solution of the Cr(II) methyl complex CrMe[N(SiMe2CH2PPh2)2] (5) reacts with MeX (X = Br, I) in a 2:1 stoichiometry to give a purple solution from which the chromium(III) alkyl halide Cr(Me)X[N(SiMe2CH2PPh2)2] (X = Br (24), I (25)) was isolated in 40% yield (equation 3.7). 120 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes M e ^ i Me2Si + MeX Toluene -78 °C MesSi MegSi X = Br 24 I 25 [3.7] These Cr(III) complexes are paramagnetic, with a solution magnetic moment of 3.8 B.M. (Evans' method),6-7 consistent with a high-spin d^ complex.62 Note that this reactivity is different from that observed for Cr(C6H2Me3)2(PMe3)2-61 Evaporation of a saturated 1:1 benzene/hexanes solution of 24 gave purple prisms. The X-ray crystal structure of 24 is shown in Figure 3.3, along with some pertinent bond lengths and angles in Table 3.3. C17 C16 Figure 3.3 Molecular structure (ORTEP) and numbering scheme for Cr(Me)Br[N(SiMe2CH2PPh2)2] (24). 121 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes The structure reveals a distorted five-coordinate Cr(III) centre; the T v a l u e 3 0 for this complex is 0.5, which is halfway between trigonal bipyramid and square pyramidal geometries. The complex could be considered as a square pyramid with the methyl C(31) in the apical position. The P(l)-Cr-P(2) angle of 170.88(7)° and the Br(l)-Cr-NQ) angle of 141.0(1)° define the distorted square base in this case. Alternatively, the phosphines can be considered the trans-axial ligands in a trigonal-bipyramid, with the Br(l)-Cr-N, Br(l)-Cr-C(31) and N(l)-Cr-C(31) angles of 141.0(1)°, 99.6(2)° and 119.4(2)° defining the equatorial plane. The coordinative unsaturation of this complex can be compared with other complexes involving chromium(III) centres with alkyl and halide ligands; such complexes are invariably octahedral or dinuclear with bridging halides. Examples include octahedral ( r i 3 - l , 3 , 5 - t r i a z a c y c l o h e x a n e ) C r ( C H 2 S i M e 3 ) 2 C l , 6 3 Cr(«Bu )2Cl [ (Me 2 PCH 2 )3CMe] , 6 4 RCrCl 2 (THF) 6 5 - 6 6 and C r M e C l 2 ( d i p p e ) ( T H F ) 6 7 and dinuclear [Cp 'CrR] 2(p-Cl)2 (Cp' = Cp, R = M e ; 4 1 Cp' = Cp*, R = Me 4 2 C H 2 P h 6 8 ) complexes. Table 3.3 Selected bond lengths and angles in Cr(Me)Br[N(SiMe2CH2PPh2)2] (24). Atom Atom Distance (A) Atom Atom Distance (A) Cr P(l) 2.464(2) Cr Br(l) 2.602(1) Cr P(2) 2.452(2) Cr C(31) 2.181(7) Cr N(l) 2.009(4) N(l) Si(l) 1.712(5) N(l) Si(2) 1.725(4) Atom Atom Atom Angle (°) Atom Atom Atom Angle (°) Br(l) Cr N(l) 141.0(1) P(l) Cr P(2) 170.88(7) Br(l) Cr P(l) 92.60(5) Br(l) Cr P(2) 95.40(5) Br(l) Cr C(31) 99.6(2) P(l) Cr N(l) 84.7(1) P(l) Cr C(31) 96.0(2) P(2) Cr N(l) 86.3(1) P(2) Cr C(31) 87.0(2) N(l) Cr C(31) 119.4(2) 122 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes The Cr-P and Cr-N bond lengths in Cr(Me)Br[N(SiMe2CH2PPh2)2] (24) are very similar to those found in thiolate 23. The Cr-C bond length of 2.181(7) A, however, is substantially different and is in fact longer than most Cr(III)-C bond lengths;44 it is also longer than the bond in the starting CrMe[N(SiMe2CH2PPh2)2] (5) complex (2.151(3) A)._ On the other hand, the Cr-N bond length of 2.009(4) A in 24 is significantly shorter than in 5 (2.117(3) A). This fact, coupled with the observed lengthening of the Si-N bonds in the product 24 (1.712(5), 1.725(4) A) implies that the amide lone pair is substantially more involved with stabilizing the Cr(ffl) metal centre than was the case for Cr(II). The Cr-Br bond length of 2.602(1) A is long compared to other examples such as 2.478(2) A in [CpCrBr] 2(|i-OCMe 3)2, 6 9 2.579(1) A in [CrBr2(H20)2]*2[Hpy]Br,70 2.496(1) A in cationic [CrBr2(cyclam)]Br (cyclam = 1,4,8,11-tetraazacyclotetradecane)71 and 2.518(1) A (average) in anionic (4-bromoanilinium)6CrBr6.72 This long bond length in methyl bromide 24 implies that there is little or no 7t-donation from the bromide to the metal in this complex; short Cr-Br bond lengths of 2.393(4) and 2.375(5) A in Cp*CrOBr2 were interpreted as being due to extensive bromide to metal rc-donation.73 Still, the most unique feature of the structure is its pentacoordination. This coordinative unsaturation is vital for polymerization catalyzed by metal complexes5'68'74 and hence the available open site, coupled with the present Cr-C bond and an abstractable halide, makes this complex an excellent starting point for further research. Although the formation of the isolated product Cr(Me)X[N(SiMe2CH2PPh2)2] (X = Br (24); X = I (25)) is consistent with the mechanism outlined in Scheme 3.2, where 24 and 25 are the products of halide atom-abstraction from RX, the product of alkyl radical addition to CrMe[N(SiMe2CH2PPh2)2] (5), namely CrMe2[N(SiMe2CH2PPh2)2] should also have been formed, but this product was not detected. Instead, a substantial amount of brown material was extracted from the reaction mixture. GC-MS head-space analysis failed to detect either MeH or MeMe (radical solvent abstraction or coupling products) and therefore attempts were made to synthesize the Cr(III) dimethyl compound by another route in order to test its stability. Addition of two equivalents of MeLi or MeMgBr to CrCl2(THF)[N(SiMe2CH2PPh2)2] (19) gave only brown 123 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes intractable material from which no viable compounds could be isolated, suggesting that the desired chromium(III) dimethyl complex was unstable (equation 3.8). P h 2 •P MeoSi .CI \ Cr T H F M e 2 S i ^ Q / THF, -78 °C M'R Decomposition at rt [3.8] •P P h 2 19 M'R = LiMe, MeMgBr, K C H 2 P h Reaction with two equivalents of KCH2PI1, Mg(CH.2Ph)2 or (PhCH2)MgBr gave a green product at -78 °C which decomposed to a brown colour upon warming to room temperature. Although the addition of a strong base such as pyridine might stabilize these highly reactive 13-electron complexes, this addition would effectively eliminate the available coordination site, therefore this route was not examined. It seems that although the reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with MeX (X = Br, I) does proceed according to the classical mechanism, one product is isolated and the other product decomposes. The mechanism of this decomposition and the ultimate fate of the chromium centre are unclear. (iii) Synthesis of a chromium(III) dialkyl complex Although a chromium(IU) dimethyl complex could not be synthesized, the observation of a green intermediate upon benzylation of the dichloride 1 9 at -78 °C prior to decomposition led to the consideration that bulky alkyl groups could be used to stabilize and prepare chromium(III) dialkyl complexes. The use of bulk to sterically crowd a metal centre and kinetically stabilize products is 124 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes well estabilished: the preparations of M(CH2SiMe3)4 (M =V, Cr) 7 5 - 7 6 and M[CH(SiMe3)2l3 (M = Ti, V, Cr) 7 7 , 7 8 are good illustrations of this concept in action. Accordingly, the bulky chromium(II) alkyl complex, Cr(CH2SiMe3)[N(SiMe2CH2PPh2)2] (9) was utilized as a starting material for the preparation of a chromium(III) dialkyl complex. The reaction of neosilyl 9 with four equivalents of benzyl chloride resulted in a colour change from purple to orange/brown. After workup, the chloride abstraction product, Cr(CH2SiMe3)Cl[N(SiMe2CH2PPh2)2] (26) was obtained in high yield (equation 3.9). Note that the use of several equivalents of benzyl chloride in this case allows for the high-yield synthesis of the chromium(ffi) alkyl halide complex. This result does seem to imply that the benzyl radical does not combine with any chromium(II) starting material, as it is known that the halide abstraction step is slow while the radical coupling step is fast. 4 7 , 4 8 , 5 3 In particular, the lack of chromium(UI) dialkyl or of decomposition products suggests that perhaps halide abstraction is relatively fast in these coordinatively unsaturated systems. Neosilyl chloride complex 26 is similar to the methyl bromide complex 24 in that it is also a high-spin, spin-only chromium(III) complex, with a solution magnetic moment of 3.8 B.M. Me 2 Si Me 5 Si P h 2 P — C r C H 2 S i M e 3 P P h 2 "78 C . . e . Toluene \ PhCH 2 CI M e 5 S ^CHgSiMea [3.9] r CI 26 The crystal structure of 26, shown in Figure 3.4, reveals another distorted pentacoordinate chromium(III) complex. The x value30 of 26 is 0.26, implying that a distorted square-based 125 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(H) Complexes pyramid is the most appropriate geometric designation. In this situation, the methyl group, C(31) is in the apical position and the trans-angles in the square base are 148.95(6)° and 164.83(13)° for P(l)-Cr-P(2) and N(l)-Cr-Cl(l) respectively (Table 3.4). Note that this orientation is the reverse of that observed in the methyl bromide complex 24, where the trans-phosphine angle was easily the largest. If the geometry is to be considered as a distorted trigonal-bipyramid the equatorial plane is defined in this case by the two phosphines and the methyl group, with the amide and chloride being trans-axially oriented. The equatorial angles are 148.95(6)°, 91.1(2)° and 119.9(2)° for P(l)-Cr-P(2), C(31)-Cr-P(l) and C(31)-Cr-P(2) respectively. The much greater distortions in this complex compared to the methyl bromide complex 24 could be due to the increase in steric interactions with the introduction of a neosilyl group. C(10) Figure 3.4 Molecular structure (ORTEP) and numbering scheme for Cr(CH2SiMe3)Cl[N(SiMe2CH2PPh2)2] (26). 126 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(IJ) Complexes Distortions appear evident in the Cr-P bond lengths of 2.422(2) and 2.525(2) A. Again, this is likely a reflection of increased steric congestion at the metal centre; the actual bond lengths are similar to those observed in the other Cr(III) structures presented here. The Cr-N bond length of 2.022(4) A is also comparable to that observed in thiolate 23 and methyl bromide 24. The Cr-C bond length of 2.110(6) A is in between those measured for thiolate 23 and methyl bromide 24 and is fairly typical of chromium(III)-carbon bond lengths.44 Table 3.4 Selected bond lengths and angles in Cr(CH 2SiMe3)Cl[N(SiMe2CH 2PPh2)2] (26). Atom Atom Distance (A) Atom Atom Distance (A) Cr P(l) 2.525(2) Cr Cl(l) 2.315(2) Cr P(2) 2.422(2) Cr C(31) 2.110(6) Cr N(l) 2.022(4) N(l) Si(l) 1.731(4) N(l) Si(2) 1.715(5) Atom Atom Atom Angle O Atom Atom Atom Angle O Cl(l) Cr N(l) 164.83(13) P(l) Cr P(2) 148.95(6) Cl(l) Cr P(l) 87.96(5) Cl(l) Cr P(2) 92.10(6) C1Q) Cr C(31) 93.5(2) P(l) Cr N(l) 86.63(13) P(l) Cr C(31) 119.9(2) P(2) Cr N(l) 85.32(13) P(2) Cr C(31) 91.1(2) N(l) Cr C(31) 101.5(2) Reaction of neosilyl chloride 26 with MeLi, MeMgBr or KCE^Ph resulted in brown solutions from which no tractable products could be identified. However, metathesis with neosilyl lithium, LiCH2SiMe3, caused a change from orange/brown to a dark green, the same colour that 127 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes was observed at low temperatures prior to decomposition upon dibenzylation of the dichloride 19. Workup of the solution allowed for the isolation of a chromium(III) dialkyl complex, Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27) in moderate yield (equation 3.10). P h 2 •P P h 2 •P Me 2 Si Me 2 Si \ : N ^ C H 2 S i M e 3 Me 2 Si — C r ' \ . *~ : | X)l L i C H 2 S i M e 3 M e 2 S i / / ^ C H 2 S i M e 3 : N — C r [ 3 . 1 0 ] C H 2 S i M e 3 Phe Toluene, -78 °C P h , 26 27 The dialkyl complex 27 is very soluble in alkane solvents. X-ray quality crystals could be grown from the slow evaporation of a solution of hexamethyldisiloxane; the structure is shown in Figure 3.5. It is immediately apparent that there is a great deal of steric congestion around the metal and this manifests itself in the large distortions in this complex. The structure, with a t value30 of 0.2, can be best described as a distorted square-based pyramid, with one neosilyl group (C(35)) in the apical position. The trans angles of the square base are defined by P(l)-Cr-P(2) (164.87(5)°) and N(l)-Cr-C(31) (152.9(2)°). Alternately, the complex could be considered as a very distorted trigonal-bipyramid, with the phosphines occupying the axial sites and the equatorial plane defined by the N(l)-Cr-C(31), N(l)-Cr-C(35) and C(31)-Cr-C(35) angles of 152.9(2)°, 105.0(2)° and 101.5(2)° respectively. As in the neosilyl chloride complex 26 , severe distortions due to steric congestion are manifested in the different Cr-P bond lengths of 2.563(2) and 2.469(2) A. Although one of the Cr-P bond lengths is particularly long when compared to other systems, the origin of the extreme asymmetry is not known. The Cr-N bond length of 2.071(4) A is slightly longer than the 2.009(4) and 2.022(4) A observed in the methyl bromide complex 2 5 or the neosilyl chloride 2 6 128 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes respectively; again, this could easily be due to the steric interactions at the metal centre. The Cr-C bond lengths in bis(neosilyl) 27 of 2.112(5) and 2.090(5) A are fairly typical of high-spin chromium(III) systems.44 C I 4 ) C(37) Figure 3.5 Molecular structure (ORTEP) and numbering scheme for Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27). Chromium(III) dialkyl systems are relatively common; even p-hydrogen-containing n-butyl groups have been incorporated into a chromium(III) system. In almost every case, however, the alkyl complexes are octahedral. Structurally characterized examples of chromium(III) complexes containing more than one Cr-C a-bond include Cp*Cr(py)(CH2Ph)2 and LiCp*Cr(CH2Ph)3,6 8 Cr("Bu)2Cl[(Me2PCH2)3CMe],64 (t]3-l,3,5-triazacyclohexane)CrR2Y (R = CH 2SiMe 3, Y = Cl ; 6 3 R = Y = CH 2Ph 7 9), Cp*CrMe2(PMe3),n CrR3['BuSi(CH2PMe2)3] (R = Me, "Bu)40 and anionic Li 3CrMe6 #3CzuH80 2. 4 3 A substantial number of chromium(III) aryl (particularly phenyl) complexes have been prepared.80-81 Chromium(III) complexes with re-bound carbon fragments have also been well studied.82 Hence, there is no inherent difficulty in preparing chromium(III) alkyl complexes; the restrictions with our system could be due to the ligand system present, or due to the coordinative unsaturation at the metal centre. The crystal structure of 129 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes Cr(CH2SiMe3)2[N(SiMe2CFf2PPh.2)2] (27) illustrates the extreme steric protection around the metal centre which appears to be a factor in preparing dialkyl complexes stabilized by the amidodiphosphine ligand. Table 3.5 Selected bond lengths and angles in Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27). Atom Atom Distance (A) Atom Atom Distance (A) Cr P(l) 2.563(2) Cr C(31) 2.112(5) Cr P(2) 2.469(2) Cr C(35) 2.090(5) Cr N(l) 2.071(4) N(l) Sid) 1.714(4) N(l) Si(2) 1.722(4) Atom Atom Atom Angle O Atom Atom Atom Angle (°) C(31) Cr N(l) 152.9(2) P(l) Cr P(2) 164.87(5) C(31) Cr P(l) 97.9(2) C(31) Cr P(2) 92.6(2) C(31) Cr C(35) 101.5(2) P(l) Cr N(l) 83.19(10) P(l) Cr C(35) 99.4(2) P(2) Cr N(l) 82.48(10) P(2) Cr C(35) 89.0(2) N(l) Cr C(35) 105.0(2) (iv) Survey of alkyl halide reactivity with chromium(II) complexes The complex CrMe[N(SiMe2CH2PPh2)2] (5) was shown to react with Mel and MeBr to yield the halide-transfer product Cr(Me)X[N(SiMe2CH2PPh2)2]. The generality of this reaction with respect to other alkyl halides and other chromium(II) systems was examined. The difficulty in this analysis lies in the fact that two products are produced, one of which (the alkyl-transfer product) is suspected of usually decomposing. The viable product is often inseparable from the 130 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(IJ) Complexes decomposition material. Reaction of chromium(II) complexes with alkyl halides in the presence of suitable trapping agents such as PR3 or pyridine is foiled by the fact that RX reacts preferentially with the trapping agents. In most cases, mass spectrometry of the residue and fractional crystallization is the most useful method to identify the products of a given reaction. Benzyl chloride was found to be a suitable substrate for the formation of the chloride-transfer product Cr(CH2SiMe3)Cl[N(SiMe2CH2PPh2)2] (26), as shown in the previous section. The reaction of benzyl chloride with CrMe[N(SiMe2CH2PPh2)2] (5) also produced the purple chloride-transfer product Cr(Me)Cl[N(SiMe2CH2PPh2)2] (28). In both cases, the benzyl-transfer product was not detected. Note that benzyl chloride does not react with [Cp*Cr(CO)3], a chromium(I) substrate that has been shown to undergo one-electron oxidation reactions.58 The chromium(H) methyl complex 5 was also shown to react rapidly with CF3CH2I to form the purple iodide-transfer product 25. Although the addition of donor ligands to stabilize in situ formation of chromium(III) dialkyls was not feasible, incorporation of alkyl groups capable of donating more than two electrons to form more stable 15- or 17-electron dialkyl complexes was considered a viable option. The use of donor groups (amines, phosphines) appended to ligand frameworks, acting as stabilizing internal bases for reactive species is well estabilished. The substrate 2-methylallyl chloride, with its ability to donate four electrons to a metal by binding in an T|3-fashion, could be considered as a form of internal base where the extra 7t-donating pair of electrons is the putative base. Accordingly, reaction of CrMe[N(SiMe2CH2PPh2)2] (5) with 0.5 equiv of 2-methylallyl chloride at -78 °C caused a rapid colour change to intense green. After workup and recrystallization, two different sets of crystals could be discerned: purple and dark green. The purple complex was found to be Cr(Me)Cl[N(SiMe2CH2PPh2)2j (28); the green complex is tentatively assigned as CrMe(Tl3-C4H7)[N(SiMe2CH2PPh2)2] (29) (equation 3.11). The mass spectrum of these crystals supports the assignments made, but the crystals could not be sufficiently separated to obtain an elemental analysis. It appears, however, that the use of 131 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes an alkyl halide in which the alkyl group can act as an internal base affords the observation of both halide and alkyl-transfer products, as is expected according to the mechanism in Scheme 3.2. Similarly, the use of the 16-electron complex Cr(rj5-C5H5)[N(SiMe2CH2PPh2)2] (H) as a starting material allows for the formation of 15- or 17-electron chromium(ITI) products, although it should be noted that this is a low-spin complex and hence not directly comparable to the other reactions presented. Nevertheless, addition of 0.5 equiv of benzyl chloride to a red solution of 11 gives a rapid colour change to green, from which pale green and dark green crystals (inseparable but distinctly observable) can be isolated. One set of crystals is CrClCp[N(SiMe2CH2PPh2)2] (2 2), identified by mass spectrometry, and the other set is likely CrCp(CH2Ph)[N(SiMe2CH2PPh2)2] (30), also tentatively identified by mass spectral peaks only. Fifteen- and 17-electron complexes containing Ti 5 -Cp and rj3-allyl fragments are quite common in chromium(UI) chemistry20,81'82 so this stabilization is not particularly surprising; such ligands are useful for probing the one-electron oxidation reactivity of our systems as both expected products become stable systems. P h 2 2 [3.11] P h 2 29 132 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes The reactivity of the dinuclear five-coordinate chromium(II) chloride compound {[(Ph2PCH2SiMe2)2N]2Cr}2(|i-Cl)2 (1) with alkyl halides was also examined. Addition of benzyl chloride to 1 resulted in the formation of the halide-transfer product CrCl2[N(SiMe2CH2PPh2)2] (19, minus the THF ligand) but the alkyl-transfer product could not be detected. Similar results were observed for 2-methylallyl chloride. Reaction with the substrate methyl chloroacetate resulted in the expected halide-transfer product but in this case the alkyl-transfer product Cr(CH2C02Me)Cl[N(SiMe2CH2PPh2)2] (31) was also detected by mass spectrometry. Considering that in this situation the alkyl-transfer product should be stable, the fact that they were not observed again may imply the absence of alkyl radical coupling with chromium(II), although the dinuclear nature of 1 and the large excess of alkyl halide may mitigate this secondary reaction pathway. Reaction of chloride 1 with Mel, however, gave a completely different result. No chromium(III) halide-transfer or alkyl-transfer product was observed. Instead, addition of Mel to a purple toluene solution of 1 resulted in the slow bleaching of the solution and the formation of a pale green precipitate. This insoluble product did not yield any recognizable mass spectral peaks, although the elemental analysis indicated that the iodide had transferred. This lack of oxidative reactivity was difficult to explain using the limited information present. In fact, as will be explained in Chapter 4, a similar reaction occurs with the cobalt system to yield a zwitterionic product, which in the case of this chromium product would be formulated as CrI2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (32). A discussion of this reactivity is best postponed until Chapter 4. 133 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes 3.5 Reactivity of Chromium(III) Complexes with Ethylene and Hydrogen Addition of four atmospheres of hydrogen to a purple toluene solution of Cr(Me)I[N(SiMe2CH2PPh2)2] (25) resulted in a slow change to pale green. After workup, a green solid was isolated and identified as the chromium(II) reduction product CrI[N(SiMe2CH2PPh2)2] (33) (Scheme 3.3). Whether the mechanism of formation involves chromium(III) hydride formation is difficult to say but it is worth pointing out that stable, corroborated chromium(IIJ) hydrides are unknown.44'83 Chromium(III) complexes have been shown to be active catalysts for the production of polyethylene. In particular, a coordinatively unsaturated molecule (usually a 13-electron system is considered the active catalyst) that also contains a chromium-carbon bond is necessary in order for catalysis to occur.5'34'68,84 The charge on the system does not seem to be a vital component of the system.74 The five-coordinate alkyl halide complexes Cr(R)X[N(SiMe2CH2PPh2)2] (R = Me, X = Br (2 4), I (2 5); R = CH 2 Ph, X = CI (2 6) and the dialkyl complex Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27) all satisfy these requirements. Unfortunately, addition of one atmosphere of ethylene at room temperature to a toluene solution of methyl iodide complex 25 or the neosilyl chloride complex 26 resulted in no production of polyethylene over one week (Scheme 3.3). This lack of reactivity could be due to the fact that these complexes are not 13-electron systems but are closer to 15-electron systems by virtue of amide and/or halide %-donation. The added electron donation may effectively negate the catalytic ability of the complexes. On the other hand, addition of ethylene to a solution of dialkyl 27 did result in the slow precipitation of a small amount of white solid, presumably polyethylene, but the solution turned brown over time and production quickly ceased. This catalyst deactivation is likely due to the same reaction that is responsible for the decomposition of sterically unencumbered chromium(III) dialkyl complexes; the colour of the final solution is reminiscent of such decompositions and may be due to some chromium(II) species. It appears then that simply the 134 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes presence of an open site of reactivity on a chromium(III) centre is not sufficient to promote the polymerization of ethylene. Scheme 3.3 r p n 2 3.6 Comparison of Five-coordinate Cr(III) and Ir(III) Complexes The X-ray structures of the five-coordinate chromium(III) complexes presented in this chapter illustrate varying degrees of distortion from a regular trigonal bipyramidal geometry. The origin of this distortion is electronic in origin; a perfect D 3h structure for a high-spin d 3 system is Jahn-Teller unstable due to the presence of one unpaired electron in the degenerate pair of orbitals d x 2-y 2 and d x y . Hence, trigonal-bipyramidal high-spin d 3 complexes will distort to remove this degeneracy. The nature of the distortion depends on the ligand set present. In the literature there are two other examples of trigonal-bipyramidal Cr(III) systems, and both show distortions from ideal D3h symmetry. The distortions in the complex CrCl2(NMe3)2 are very small; the trans-N-Cr-N angle is 178.8(5)° and the equatorial angles are 111.4(2)°, 124.3(1)° and 124.3(1)°. 3 2 Larger changes are observed in Na2CrPh5»3Et20»THF; the trans-angle is 161° and the equatorial angles are 104°, 111° and 145°. 3 3 The x value for this complex is 0.27, indicating a substantial distortion 135 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes from trigonal bipyramidal geometry; the angles observed here are reminiscent of those observed in our systems. Steric effects are also important (certainly so in the pentaphenyl system) and must be taken into consideration. Note that for the complexes prepared in this thesis, the ligand set is not ML5. The incorporation of different ligands into a complex reduces the symmetry, resulting in the degeneracy of the two d-orbitals in question being removed to a certain extent. — dz2 — dz2 V--.^ d x y dx2.y2 \ ""--rt — d x y d x 2 - y 2 ^ \ t * \ * v * V+-dxzdyz % # d x z d y z Cr(lll), d 3 lr(lll),d 6 (a) (b) Figure 3.6 (a) High-spin Cr(III) and (b) low-spin Ir(III) d-orbital occupancy in a trigonal-bipyramidal geometry; in both cases Jahn-Teller distortion is necessary. A diamagnetic analogue of high-spin Cr(III) (d3) would be low-spin Rh and Ir(III) (d6). Instead of three orbitals being half-filled, they are doubly occupied in the Rh and Ir(III) systems (Figure 3.6). A substantial amount of work has been done using the amidodiphosphine ligand and these two diamagnetic metals.85 In fact, complexes with the exact ligand set as chromium(III) have been prepared.86"89 This provides an excellent opportunity to compare the geometries of two complexes which differ only in metal centre. The iridium(III) complexes Ir(R)Y[N(SiMe2CH2PPh2)2] (R = alkyl; Y = alkyl or halide) all show substantial distortions from trigonal-bipyramidal geometry and the nature of the distortion has been explained theoretically.90" 136 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes 9 3 Although the chromium(III) and iridium(III) crystal radii are different (0.755 vs. 0.82 A for octahedral geometry),94,95 it is not unreasonable to consider that the predictive theory for d 6 iridium(III) distortions would apply to d 3 chromium(III) systems as well. When the structure of Ir(Me)I[N(SiMe2C H2PPh2)2] is compared to Cr(Me)Br[N(SiMe2CH2PPh2)2] (24) it is clearly obvious the two structures are quite different. The iridium complex is almost perfectly square pyramidal, with the methyl in the apical position.86,88 The trans-angles in the square-base are 174.44(15)° and 170.02(6)°. All other angles are within six degrees of ideal 90°. On the other hand, the chromium(III) complex could not be considered as a square pyramid, with the two largest angles being 170.88(7)° and 141.0(1)°. In the case of the neosilyl chloride complex 26 as well the difference is obvious (trans-angles are 164.83(13)° and 148.95(6)°) although perhaps a case for steric effects could be made here. Not so in the methyl bromide complex. In fact, the iridium methyl bromide complex is more sterically hindered; the Ir-P bond lengths of 2.327(2) and 2.335(2) A are much shorter than those found in the high-spin chromium(III) complexes (2.452(2) and 2.464(2) A in 24) despite chromium(ni) being a smaller metal centre. This implies that metal to phosphine electron-donation is increasingly more facile with diamagnetic iridium(III) than with paramagnetic chromium(III). The fact that different structures are observed for the chromium(III) systems implies that the calculations which predict geometric distortions for d 6 iridium(III) complexes cannot be applied to high-spin d 3 chromium(III) complexes. Dialkyl complexes of iridium(III) were also prepared and a comparison of the structures of Ir(R)R'[N(SiMe2CH2PPh2)2] (R = R' = CH 2 Ph; 8 9 R = CH 2 SiMe 3 , R' = Me8 7) with Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27) reveals substantial differences. The iridium complexes were considered to be trigonal-bipyramidal in nature (the so-called Y-shape), with two large and one very small angle (opposite the amide) in the equatorial plane. As an example, the equatorial angles in the iridium dibenzyl complex are 141.6(1)°, 140.8(1)° and 77.6(1)°; the trans-phosphine angle is 170.2(5)°. In the chromium complex, the trans-phosphine angle is 164.87(5)° 137 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes but the equatorial angles are 152.9(2)°, 105.0(2)° and 101.5(2)°, considerably different and in fact much more square-pyramidal in nature. However, the extreme steric congestion in the chromium dialkyl makes it difficult to ascribe the geometric differences purely to electronic effects. One explanation for the lack of applicability of the d 6 iridium theoretical calculation to d 3 chromium could be the inherent difference between first and third row transition metals: the energy splitting of 3d orbitals is much smaller than that of 5d orbitals.3 As a result, the distortions required to remove the degeneracy of the d x 2 . y 2 and d x y orbitals of a trigonal-bipyramid should be much greater in a first row metal than in a third row metal. Given that, calculations done on 5d iridium(III) d 6 centres may not be totally valid for 3d chromium(III) d 3 systems. Distortions are expected to be exacerbated in first-row metal systems and that is in fact observed; the chromium(III) complexes prepared all show greater distortions than the analogous iridium(III) systems and the Cr(UI) distortions are not necessarily in the fashion predicted for Ir(III). The fact that complexes with exactly the same ligand set have different geometries supports this idea. Calculations done on a chromium(III) centre in a similar manner to that for iridium(III) would likely be able to model the observed distortions. 3.7 Summary and Conclusions Chromium(II) complexes containing a metal-carbon bond have been shown to undergo facile one-electron oxidation reactions to chromium(III) products with a variety of substrates. Diphenyldisulfide participates in one-electron oxidation to a square-planar chromium(II) methyl complex to give a mononuclear five-coordinate chromium(III) thiolate complex, which also contains a metal-carbon bond. The classical reaction of chromium(II) with alkyl halides was shown to proceed for highly coordinately and electronically unsaturated complexes, as well as with chromium(II) complexes containing a metal-carbon bond. However, unless the alkyl groups on chromium are very bulky, dialkyl complexes could not be prepared and hence in many cases the 138 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes details of the reactivity and product formation could not be discerned. The difficulty of separating two products, one of which sometimes decomposed, made a thorough examination of this reaction challenging at best. The nature of the decomposition of the dialkyl complexes was not resolved. Chromium-carbon bond homolysis, a-elimination or reductive elimination of alkane could all be considered. Crystal structure analysis of products was vital and not always obtainable. The structures that were solved were all unusual examples of five-coordinate chromium(III) complexes. Distortions in the geometries could not be related to analogous iridium(III) complexes. Unfortunately, none of the chromium(III) alkyl complexes prepared was an efficient ethylene polymerization catalyst. 3 . 8 Experimental 3.8.1 General Procedures and Materials Unless otherwise stated all procedures were performed as described in section 2.8. Diphenyldisulfide was sublimed prior to use. Mel was distilled under nitrogen and degassed prior to use. Benzyl chloride, 2-methylallyl chloride, methyl chloroacetate and trifluoromethyliodomethane (stabilized with copper) were passed through a column of activated neutral alumina into a bomb and degassed with three freeze-pump-thaw cycles. 3.8.2 Synthesis and Reactivity of Complexes (i) CrCl2(THF)[N(SiMe 2CH 2PPh2)2] (19) To a blue THF solution of {[(Ph2PCH2SiMe2)2N]2Cr}2(u.-Cl)2 (1) (0.25 g, 0.20 mmol) in 20 mL THF was added solid AgBPh4 (0.17 g, 0.40 mmol). The suspended silver salt slowly reacted to give a dark purple solution and a black insoluble precipitate. After five hours of being 139 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes stirred, the solvent was removed in vacuo, the residue extracted with toluene and filtered through Celite to give a dark purple solution. The toluene solution was reduced to a minimum amount and placed overnight in a -40 °C freezer, after which CrCi2(THF)[N(SiMe2CH2PPh2)2] (19) was isolated as a purple powder. Addition of hexanes precipitated a white powder identified by MS (mle 242) as BPI13. Elemental analysis of different samples had varying amounts of ligated THF remaining; extended drying in vacuo failed to remove all THF. X-ray quality crystals were obtained by layering a toluene/THF solution with hexanes. Yield: 0.12 g (40%). Anal. Calcd. for C3oH36Cl2CrNP2Si2*C4H80: C, 56.42; H, 6.13; N, 1.94. Calcd. for 19»0.5THF: C, 55.89; H, 5.86; N, 2.04; Calcd. for 19 (no solvent: C, 55.30; H, 5.57; N, 2.15. Found: C, 55.80; H, 6.15; N, 2.05. ' H NMR (C 6D 6): 5 12.5 (v br, 4H), 3.5 (v br, 12H). MS: mle 650 (M+-THF), 615 (M+-THF-C1). Peff = 3.8 B.M. (ii) CrCpCl [N(S iMe 2 CH 2 PPh 2 ) 2 ] (22) To a purple solution of CrCl2(THF)[N(SiMe2CH2PPh2)2] (19) (0.13 g, 0.18 mmol) in 10 mL THF was added at -78 °C a THF solution of NaCp»DME (0.032 g, 0.18 mmol). No reaction occurred at -78 °C but as the solution was warmed to room temperature the colour changed to dark turquoise. After being stirred overnight, the solvent was removed in vacuo, the residue extracted with toluene, filtered through Celite and the resulting blue/green solution concentrated to a minimum. Addition of hexanes (1 mL) and storage in a -40 °C freezer overnight gave CrCpCl[N(SiMe2CH2PPh2)2] (22) as dark green plates. Yield: 0.10 g (83%). Anal. Calcd. for C35H4iClCrNP2Si2*C7H8: C, 65.22; H, 6.39; N, 1.81. Found: C, 64.88; H, 6.33; N, 1.68. MS: mle 680 (M+), 644 (M+-C1), 615 (M+-Cp). peff = 3.6B.M. (iii) CrMe(SPh)[N(SiMe 2 CH 2 PPh 2 )2 ] (23) To a 10 mL red/brown toluene solution of CrMe[N(SiMe2CH2PPh2)2] (5) (0.16 g, 0.27 mmol) cooled to 0 °C was added a 5 mL toluene solution of PhSSPh (0.03 g, 0.14 mmol). Immediately the solution changed to a dark purple colour. After being stirred for one hour at 0 °C, 140 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes the solution was warmed to room temperature and the solvent removed almost to dryness (0.25 m L toluene remaining). The residue was quickly dissolved in 1 m L hexanes, filtered through Celite and the solvent removed in vacuo. Recrystallization from hexanes/toluene (1 m L : 3 drops) in a -40 °C freezer yielded a thick o i l , which upon agitation gave CrMe(SPh)[N(SiMe2CH2PPh2)2] (3) as purple crystals. Y i e l d : 0.12 g (66%). Ana l . Calcd. for C3 7 H44CrNP 2 SSi2»C7H 8 : C , 63.04; H , 6.29; N , 1.99. Found: C , 62.95; H , 6.36; N , 2.13. lH N M R (C 6D 6): 5 13.0 (v br, nH), 10.6 (br, nH), 10.0 (br, nH), 6.0 (v br, 2nH), 4.7 (v br, 2nH), 1.5 (br, nH). M S : mle 689 (M+-Me), 580 (M+-Me-SPh). p.eff = 3.8 B . M . (iv) Cr(Me)Br[N(SiMe 2CH 2PPh2)2] (24) A 10 m L purple toluene solution of CrMe[N(SiMe2CH2PPh2)2] (5) (0.12 g, 0.19 mmol) in a bomb was frozen in hquid nitrogen. To this was added one equivalent of M e B r by quantitative vacuum transfer (Note: M e B r boils at 6 °C so the storage bomb must be cooled to -10 "C before opening). Upon removal of the liquid nitrogen bath and melting of the toluene, a rapid reaction resulted in a dark purple solution. After being warmed to room temperature and being stirred for one hour, the solvent was removed in vacuo, the residue extracted with toluene, filtered through Celite and reduced to a minimum volume. Layering with hexanes (5 mL) yielded purple crystals of C r ( M e ) B r [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (24). Y i e l d : 0.050 g (39%). A n a l . Ca lcd . for C3iH39BrCrNP2Si2*0.5 C 7 H 8 : C, 57.41; H , 6.00; N , 1.94. Found: C, 57.57; H , 6.20; N , 2.05. A H N M R (C 6D 6): 8 11.2 (v br), 10.4 (br), 7.0 (br, shoulder), 6.2 (v br, overlap), 5.8 (v br, overlap), 4.2 (v br). M S : mle 661 ( M + - M e ) , 580 (M+-Me-Br). u . e f f = 3.8 B . M . (v) Cr(Me)I[N(SiMe 2 CH 2 PPh 2 ) 2 ] (25) The reaction was performed as per (iv), substituting one equivalent of M e l for MeBr . After workup, C r ( M e ) I [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (25) was isolated as purple crystals. Y i e l d : 0.060 g (43%). A n a l . Calcd. for C 3 7H 4 4CrINP2Si2«0 .5 C 7 H 8 : C, 53.90; H , 5.64; N , 1.82; I, 16.51. 141 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes Found: C, 54.17; H, 5.66; N, 1.60; I, 16.30. *H NMR (C 6D 6): 5 11.2 (v br), 10.4 (br), 6.3 (v br, overlap), 5.9 (v br, overlap), 4.1 (v br). MS: mle 722 (M+), 707 (M+-Me). u. e f f = 3.8 B.M. (vi) Attempted synthesis of CrR2[N(SiMe2CH2PPh2)2] (R = Me, CH 2 Ph) To a 10 mL THF solution of CrCl2(THF)[N(SiMe2CH2PPh2)2] (19) (0.10 g, 0.17 mmol) at -78 °C was added two equivalents MeMgBr (THF stock solution), MeLi (ether stock solution) or KCH 2Ph (THF solution, 0.039 g). Upon addition of the methylating agents, the purple solution rapidly turned dark brown. Warming to room temperature caused no further colour change and workup did not allow for the identification of tractable products. Upon addition of KCH 2Ph, the solution turned a dark green at low temperature and persisted until the solution was warmed to room temperature, during which the colour changed to brown. Workup gave only intractable brown pastes. (vii) Cr(CH 2SiMe 3)Cl[N(SiMe2CH 2PPh2)2J (26) To a 10 mL purple toluene solution of Cr(CH 2SiMe3)[N(SiMe 2CH 2PPh 2) 2] (9) (approximately 0.10 g, 0.15 mmol) was added 2 drops of neat benzyl chloride (approximately 1.5 equiv) at -78 °C. No immediate reaction occurred but upon being warmed to room temperature the solution changed to golden orange/brown. After being stirred overnight, the solvent was removed in vacuo, the residue extracted with a minimum of toluene, filtered through Celite and hexanes added (1:1). Overnight, from the solution, dark orange crystals of Cr(CH 2SiMe 3)Cl[N(SiMe 2CH 2PPh 2) 2] (26) were deposited. Yield: 0.095 g (90%). Anal. Calcd. for C34H47ClCrNP2Si3: C, 58.06; H, 6.73; N, 1.99. Found: C, 58.43; H, 6.79; N, 1.92. MS: mle 701 (M+-H), 615 (M+-CH2SiMe3). n eff = 3.8 B.M. 142 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes (viii) Attempted synthesis of Cr(CH2SiMe3)R[N(SiMe2CH2PPh2)2] (R = Me CH 2 Ph) Addition of one equivalent of MeLi (ether stock solution) or KCH2Ph (0.028 g, 0.11 mmol) to a 10 mL THF solution of Cr(CH2SiMe3)Cl[N(SiMe2CH2PPh2)2] (26) (0.080 g, 0.11 mmol) at -78 °C resulted in a change in darkening of the orange/brown solution to dark brown. Upon warming of the solution, no further reaction was observed. After workup only intractable brown pastes could be isolated in both cases, from which no viable products could be identified. (ix) Cr(CH 2 SiMe 3 ) 2 [N(SiMe 2 CH 2 PPh 2 )2] (27) Crystals of Cr(CH2SiMe3)Cl[N(SiMe2CH2PPh2)2] (26) (0.13 g, 0.18 mmol) were dissolved in 10 mL THF to give a dark orange solution, which was cooled to -78 °C. To this was added dropwise a 10 mL toluene solution of LiCH2SiMe3 (0.017 g, 0.18 mmol), which resulted in an instant colour change to dark green. Upon being warmed the solution turned a darker green and after 30 minutes of stirring at room temperature the THF was removed in vacuo, the residue extracted with 2 mL hexanes, filtered through Celite and then pumped to dryness again. The residue was dissolved in a minimum amount of hexamethyldisiloxane (1.5 mL) and placed in a -40 °C freezer. Overnight, long green bars of Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27) were isolated. Yield: 0.070 (51%). Anal. Calcd. for C38H58CrNP2Si4»0.5 (Me3Si)20: C, 58.88; H, 8.07; N, 1.67. Found: C, 59.20; H, 7.61; N, 1.70. MS: mle 580 (M+-(CH2SiMe3)2). (x) Reaction of CrMe[N(SiMe 2 CH 2 PPh 2 ) 2 ] (5) with PhCH 2CI and C F 3 C H 2 I Addition of two drops of neat PhCH2Cl or CF3CH2I to a toluene solution of CrMe[N(SiMe2CH2PPh2)2] (5) (0.05 g, 0.08 mmol) at room temperature resulted in a rapid colour change to dark purple. After being stirred for one hour, the solvent was removed in vacuo, the residue extracted with toluene, filtered through Celite and the solvent removed again. A mass spectrum of the crude product indicated the formation of Cr(Me)Cl[N(SiMe2CH2PPh2)2] (28) 143 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes (mle 630 (M + ) , 615 (M +-Me)) from the benzyl chloride reaction and Cr(Me)I[N(SiMe2CH2PPh2)2] (25) (mle 707 (M+-Me)) from the CF 3 CH 2 I reaction. (xi) Reaction of CrMe[N(SiMe 2 CH 2 PPh 2 ) 2 ] (5) with 2-methylallyl chloride CrMe[N(SiMe2CH2PPh2)2] (5) (0.14 g, 0.23 mmol) was dissolved in 10 mL toluene, degassed and frozen. To this was added 2-methylallyl chloride (0.12 mmol) by vacuum transfer using a constant volume gas addition cell. The liquid nitrogen bath was removed and upon melting of the toluene, the brown/red solution quickly turned deep green. The solution became a very dark green as it was warmed to room temperature. The reaction was stirred overnight, then the solvent removed in vacuo, the residue extracted with 1 mL toluene, filtered through Celite and 1 mL hexanes added. Overnight a dark green solid and purple crystals were deposited simultaneously. The purple crystals were identified as Cr(Me)Cl[N(SiMe2CH2PPh2)2] (28) (mle 630 (M+), 615 (M+-Me)) and the green solid as OMe(ri3-C4H7)[N(SiMe2CH2PPh2)2] (29) (mle 635 (M+), 580 (M+'Me-QFL?). The solids could not be separated sufficiendy to obtain elemental analysis. (xii) Reaction of Cr(Ti5-C5H5)[N(SiMe 2CH 2PPh 2) 2] (11) with P h C H 2 C l To a deep red solution of Cr(Ti5-C5H5)[N(SiMe2CH2PPh2)2] (11) (0.09 g, 0.14 mmol) in 10 mL toluene at -78 °C was added PI1CH2CI (toluene stock solution, 0.07 mmol). No immediate reaction occurred but as the solution was warmed to room temperature, the solution turned dark green. After being stirred overnight, the dark green solution was reduced to a minimum (1 mL), hexanes added (2 mL). Dark green and light green crystals were deposited from the solution overnight. The two products were tentatively identified as CrCpCl[N(SiMe2CH2PPh2)2] (22) (MS as reported in (ii)) and CrCp(CH2Ph)[N(SiMe2CH2PPh2)2] (30) (mle 735 (M+-H), 670 (M+-Cp-H)). The solids could not be separated sufficiently to obtain elemental analysis. 144 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes (xiii) Reaction of {[(Ph 2 PCH 2 SiMe 2 ) 2 N] 2 Cr} 2 (p-Cl) 2 (1) with PhCH 2 Cl and C l C H 2 C 0 2 M e {[(Ph2PCH2SiMe2)2N]2Cr}2(p-Cl)2 (1) (0.10 g, 0.08 mmol) was dissolved in 10 mL toluene and cooled to -78 °C to give a purple solution. To this was added two drops of neat PhCFf2Cl or ClCH.2C02Me. Addition of benzyl chloride caused an immediate change to dark brown; the methyl chloroacetate did not begin to turn brown until warmed to room temperature. Overnight stirring at room temperature did not cause any further change; the solvent was then removed in vacuo, the residue extracted with minimum toluene (2 mL), filtered through Celite and hexanes added (2 mL). Purple crystals and a brown solid both precipitated from the solution overnight. The purple solid was identified as CrCl2[N(SiMe2CH2PPh2)2] (19-THF) (mle 650 (M+), 615 (M+-C1)). The brown solid was not identifiable, although in the case of the methyl chloroacetate reaction the mass spectrum of the residue showed a peak at mle 652, tentatively assigned as {Cr(CH2C02Me)[N(SiMe2CH2PPh2)2] - H}, consistent with the formation of Cr(CH2C02Me)Cl[N(SiMe2CH2PPh2)2] (31). (xiv) Reaction of {[(Ph 2 PCH 2 SiMe 2 ) 2 N] 2 Cr} 2 (p-CI) 2 (1) with Mel {[(Ph2PCH2SiMe2)2N]2Cr}2(p-Cl)2 (1) (0.075 g, 0.06 mmol) was dissolved in 10 mL toluene in a bomb and frozen using a liquid nitrogen bath. To this was added Mel (0.03 mmol and 0.06 mmol) by vacuum transfer using a constant volume gas addition cell. In the case of the addition of 0.5 equiv Mel, the solution turned dark purple upon warming to room temperature, with concomitant formation of a very fine pale green precipitate. With one equivalent present, the solution colour bleached completely as the precipitate formed. After being stirred overnight, the solvent was decanted and the precipitate washed with toluene and dried. The solution from the 0;5 equiv reaction was light purple and workup gave CrCi2[N(SiMe2CH2PPh2)2] (19-THF) in 40% yield. The precipitate was tentatively identified as Crl2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (32) only after similar complexes were characterized with cobalt starting materials (Chapter 4). 145 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes Yield: 0.02 g (20%). Anal. Calcd. for C3iH39CrI2NP2Si2: C, 43.83; H, 4.63; N, 1.65. Found: C, 44.21; H, 5.02; N, 2.00. MS: mle 707 (M+-Me-I). (xv) Reaction of Cr(Me)I[N(SiMe 2CH 2PPh2)2] (25) with H 2 A 10 mL purple toluene solution of Cr(Me)I[N(SiMe2CH2PPh2)2] (25) (0.12 g, 0.17 mmol) in a bomb was degassed by two freeze-pump-thaw cycles. The solution was placed under 4 atmospheres of hydrogen and warmed to room temperature. After 20 minutes, the purple colour had changed to light green. After being stirred overnight, the solvent was removed in vacuo, the residue extracted with hexanes and filtered through Celite to give a light green solution from which a light green solid was obtained. Yield: 0.030 g (26%). The hexane-insoluble residue was extracted with toluene to yield a brown solution from which intractable brown pastes were obtained. The green solid is tentatively identified as CrI[N(SiMe2CH2PPh2)2] (33). Anal. Calcd. for C3oH36CrINP2Si2»0.5 C 7 H 8 : C, 53.38; H, 5.35; N, 1.86. Found: C, 54.02; H, 5.54; N, 2.00. MS: mle 707 (M+). (xvi) Conditions of attempted reaction of Cr(R)X[N(SiMe 2CH2PPh2)2] with ethylene Addition of one atmosphere of ethylene to a bomb containing a 10 mL toluene solution of C r ( M e ) I [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (2 5) (0.05 g, 0.07 mmol) or Cr(CH2SiMe3)Cl[N(SiMe2CH2PPh2)2] (26) (0.05 g, 0.07 mmol) resulted in no apparent reaction over one week. No polyethylene was produced and no colour change occurred. Mild heating to 60 °C for three days also had no effect. (xvii) Reaction of Cr(CH 2 SiMe 3 ) 2 [N(S iMe 2 CH 2 PPh 2 ) 2 ] (27) with ethylene Addition of one atmosphere of ethylene to a bomb containing a 10 mL toluene solution of Cr(CH2SiMe3)2[N(SiMe2CH2PPh2)2] (27) (0.05 g, 0.07 mmol) caused no immediate colour change but over twelve hours a small amount of white solid, presumably polyethylene, had been 146 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes produced. After 24 hours, the solution had changed from green to dark brown/red and no further polymer formation was observed. 3.9 R e f e r e n c e s (1) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles and Applications of Organotransition Metal Chemistry; University Science Books: Mill Valley, CA, 1987, Ch. 5. (2) Crabtree, R. F£. The Organometallic Chemistry of the Transition Metals; 2nd ed.; John Wiley & Sons: USA, 1994. (3) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry; 5th ed. ed.; John Wiley & Sons: New York, 1988. (a) Ch. 28; (b) pp 1306-1318; (c) Ch. 18-C. (4) Douglas, B. E.; McDaniel, D. H.; Alexander, J. J. Concepts and Models in Inorganic Chemistry; 2nd ed.; John Wiley & Sons: USA, 1983. (5) Thomas, B. J.; Noh, S. K.; Schulte, G. K.; Sendlinger, S. C ; Theopold, K. H. / . Am. Chem. Soc. 1991,113, 893. (6) Evans, D. F. / . Chem. Soc. 1959, 2003. (7) Sur, S. K. / . Magn. Res. 1989, 82, 169. (8) Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements; Pergamon Press: England, 1984, pp 1196. (9) Al-Soudani, A. R. H.; Batsanov, A. S.; Edwards, P. G.; Howard, J. A. K. / . Chem. Soc, Dalton Trans. 1994, 987. (10) Fettinger, J. C ; Mattamana, S. P.; Poli, R.; Rogers, R. D. Organometallics 1996,15, 4211. (11) Grohmann, A.; Kohler, F. H.; Miiller, G.; Zeh, H. Chem. Ber. 1989,122, 897. 147 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes (12) Fowlie, A. D.; House, D. A.; Robinson, W. T.; Rumball, S. S. / . Chem. Soc. (A) 1970, 803. (13) Danopoulos, A. A.; Wilkinson, G.; Sweet, T. K. N.; Hursthouse, M. B. / . Chem. Soc, Dalton Trans. 1996, 271. (14) Bradley, D. C.; Hursthouse, M. B.; Newing, C. W. Chem. Comm. 1971, 411. (15) Schubert, U.; Friedrich, P.; Orama, O. / . Organomet. Chem. 1978,144, 175. (16) Daly, J. J.; Sneeden, R. P. A. J. Chem. Soc. (A) 1967, 736. (17) Edelmann, F.; Trotter, J. / . Organomet. Chem. 1977,131, 65. (18) Luo, X.-L.; Schulte, G. K.; Crabtree, R. H. Inorg. Chem. 1990, 29, 682. (19) Keogh, D. W.; Poli, R. / . Am. Chem. Soc. 1997,119, 2516. (20) Poli, R. Chem. Rev. 1996, 96, 2135. (21) Kice, J. L. In Sulfur in Organic and Inorganic Chemistry; A. Senning, Ed.; Marcel Dekker: New York, 1971; Vol. 1; pp 153. (22) King, R. B. / . Am. Chem. Soc. 1963, 85, 1587. (23) Goh, L. Y.; Tay, M. S.; Mak, T. C. W.; Wang, R.-J. Organometallics 1992,11, 1711. (24) Abrahamson, H. B.; Freeman, M. L. Organometallics 1983,2, 679. (25) Brandenburg, K. L.; Heeg, M. J.; Abrahamson, H. B. Inorg. Chem. 1987, 26, 1064. (26) Ju, T. D.; Capps, K. B.; Lang, R. R ; Roper, G. C ; Hoff, C. D. Inorg. Chem. 1997, 36, 614. (27) Lang, R, F.; Ju, T. D.; Kiss, G.; Hoff, C. D.; Bryan, J. C ; Kubas, G. / . Am. Chem. Soc. 1994,116, 7917. (28) Nefedov, S. E.; Pasynskii, A. A.; Eremenko, I. L.; Gasanov, G. S.; Ellert, O. G.; Novotortsev, V. M.; Yanovsky, A. I.; Struchkov, Y. T. / . Organomet. Chem. 1993, 443, 101. (29) Prout, K.; Rees, G, V. Acta. Cryst. B 1974, B30, 2717. (30) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor, G. C. / . Chem. Soc, Dalton Trans. 1984, 1349. 148 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes (31) Fowles, G. W. A.; Greene, P. T. Chem. Comm. 1966, 784. (32) Greene, P. T.; Russ, B. J.; Wood, J. S.J. Chem. Soc. (A) 1971, 3636. (33) Miiller, E.; Krause, J.; Schmiedeknecht, K. / . Organomet. Chem. 1972,44, 127. (34) Liang, Y.; Yap, G. P. A.; Rheingold, A. L.; Theopold, K. H. Organometallics 1996,15, 5284. (35) Bradley, D. C.; Hursthouse, M. B.; Rodesiler, D. F. Chem. Comm. 1969, 14. (36) Cummins, C. C ; Lee, J.; Schrock, R. R.; Davis, W. D. Angew. Chem., Int. Ed. Engl. 1992,31, 1501. (37) Stein, C ; Bouma, S.; Carlson, J.; Cornelius, C ; Maeda, J.; Weschler, C ; Deutsch, E.; Hodgson, K. O. Inorg. Chem. 1976,15, 1183. (38) Nicholson, J. R.; Christou, G.; Wang, R.-J.; Huffman, J. C ; Chang, H.-R.; Hendrickson, D. N. Polyhedron 1991,10, 2255. (39) Raston, C. L.; White, A. H. Aust. J. Chem. 1977, 30, 2091. (40) Gardner, T. G.; Girolami, G. S. / . Chem. Soc, Chem. Commun. 1987, 1758. (41) Richeson, D. S.; Hsu, S.-W.; Fredd, N. H.; van Duyne, G.; Theopold, K. H. / . Am. Chem. Soc. 1986,108, 8273. (42) Herrmann, W. A.; Thiel, W. R.; Herdtweck, E. / . Organomet. Chem. 1988, 353, 323. (43) Krausse, J.; Marx, G. / . Organomet. Chem. 1974, 65, 215. (44) Theopold, K. H. Acc. Chem. Res. 1990,23, 263. (45) Aubart, M. A.; Bergman, R. G. / . Am. Chem. Soc. 1996,118, 1793. (46) Kochi, J. K ; Powers, J. W. J. Am. Chem. Soc. 1970, 92, 137. (47) Espenson, J. H. Prog. Inorg. Chem. 1983, 30, 189. (48) Espenson, J. H. Acc. Chem. Res. 1992, 25, 222. (49) Kupferschmidt, W. C ; Jordan, R. B. / . Am. Chem. Soc. 1984,106, 991. (50) Sisley, M. J.; Jordan, R. B. Inorg. Chem. 1988, 27, 1963. (51) Zhang, Z.; Jordan, R. B. Inorg. Chem. 1993, 32, 5472. (52) Bakac, A.; Butkovic, V.; Espenson, J. H.; Orhanovic, M. Inorg. Chem. 1993, 32, 5886. 149 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions of Cr(II) Complexes (53) Baird, M. C. Chem. Rev. 1988, 88, 1217. (54) Tyler, D. R. Prog. Inorg. Chem. 1988, 36, 125. (55) Huber, T. A.; Macartney, D. H.; Baird, M. C. Organometallics 1993,12, 4715. (56) Huber, T. A.; Macartney, D. H.; Baird, M. C. Organometallics 1995,14, 592. (57) Goulin, C. A.; Huber, T. A.; Nelson, J. M.; Macartney, D. H.; Baird, M. C. / . Chem. Soc, Chem. Commun. 1991, 798. (58) MacConnachie, C. A.; Nelson, J. M.; Baird, M. C. Organometallics 1992,11, 2521. (59) Scott, S. L.; Espenson, J. H.; Zhu, Z. / . Am. Chem. Soc. 1993,115, 1789. (60) Lee, K.-W.; Brown, T. L. / . Am. Chem. Soc. 1987,109, 3269. (61) Hermes, A. R.; Morris, R. J.; Girolami, G. S. Organometallics 1988, 7, 2372. (62) Carlin, R. L. Magnetochemistry; Springer-Verlag: Heidelberg, 1986. (63) Kbhn, R. D.; Kocoik-Kohn, G. Angew. Chem., Int. Ed. Engl. 1994, 33, 1877. (64) Thaler, E. G.; Folting, K.; Huffman, J. C ; Caulton, K. G. / . Organomet. Chem. 1989, 376, 343. (65) Kurras, E. Monatsber. Dt. Akad. Wiss. 1963, 5, 378. (66) Nishimura, K.; Kuribayashi, H.; Yamamoto, A.; Ikeda, S. / . Organomet. Chem. 1972, 37, 317. (67) Hermes, A. R.; Girolami, G. S. Inorg. Chem. 1990, 29, 313. (68) Bhandari, G.; Kim, Y.; McFarland, J. M.; Rheingold, A. L.; Theopold, K. H. Organometallics 1995,14, 738. (69) Nefedov, S. E.; Pasynskii, A. A.; Eremenko, I. L.; Orazsakhatov, B.; Ellert, O. G.; Novotortsev, V. M.; Katser, S. B.; Antsyshkina, A. S.; Porai-Koshits, M. A. / . Organomet. Chem. 1988, 345, 97. (70) Ladd, M. F. C ; Larkworthy, L. F.; Leonard, G. A.; Povey, D. C ; Tandon, S. S. / . Chem. Soc, Dalton Trans. 1984, 2351. (71) Lisgarten, J. N.; Palmer, R. A.; Hemmings, A. M.; Gazi, D. M. Acta. Cryst. C 1990, C46, 396. 150 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes (72) Halepoto, D. M.; Larkworthy, L. F.; Povey, D. C ; Siddiqui, R. A.; Smith, G. W. Inorg. Chim. Acta 1994, 227, 167. (73) Morse, D. B.; Rauchfuss, T. B.; Wilson, S. R. / . Am. Chem. Soc. 1988,110, 8234. (74) Theopold, K. H.; Heintz, R. A.; Noh, S. K.; Thomas, B. J. In Homogeneous Transition Metal Catalyzed Reactions; W. R. Moser and D. W. Slocum, Eds.; American Chemical Society: Washington, DC, 1992; pp 591. (75) Yagupsky, G.; Mowat, W.; Shortland, A.; Wilkinson, G. / . Chem. Soc, Chem. Commun. 1970, 1369. (76) Mowat, W.; Shortland, A.; Yagupsky, G.; Hill, N. J.; Yagupsky, M.; Wilkinson, G. / . Chem. Soc, Dalton Trans. 1972, 533. (77) Davidson, P. J.; Lappert, M. F.; Pearce, R. Acc. Chem. Res. 1974, 7, 209. (78) Barker, G. K.; Lappert, M. F. / . Organomet. Chem. 1974, 76, C45. (79) Kdhn, R. D.; Kociok-Kohn, G.; Haufe, M. / . Organomet. Chem. 1995, 501, 303. (80) Hein, F.; Tille, D. Z. Anorg. Allg. Chem. 1964, 329, 72. (81) Sneeden, R. P. A. Organochromium Compounds; Academic Press: New York, 1975. (82) Jolly, P. W. Acc. Chem. Res. 1996,29, 544. (83) Heintz, R. A.; Koetzle, T. F.; Ostrander, R. L.; Rheingold, A. L.; Theopold, K. H.; Wu, P. Nature 1995, 378, 359. (84) Bhandari, G.; Rheingold, A. L.; Theopold, K. H. Chem. Eur. J. 1995,1, 199. (85) Fryzuk, M. D. Can. J. Chem. 1992, 70, 2839. (86) Fryzuk, M. D.; MacNeil, P. A.; Rettig, S. J. / . Am. Chem. Soc. 1985,107, 6708. (87) Fryzuk, M. D.; MacNeil, P. A.; Ball, R. G. / . Am. Chem. Soc. 1986,108, 6414. (88) Fryzuk, M. D.; MacNeil, P. A.; Rettig, S. J. Organometallics 1986,5, 2469. (89) Fryzuk, M. D.; MacNeil, P. A.; Massey, R. L.; Ball, R. G. / . Organomet. Chem. 1989, 368, 231. (90) Thorn, D. L.; Hoffmann, R. New J. Chem. 1979,3, 39. (91) Jean, Y.; Eisenstein, O. Polyhedron 1988, 7, 405. 151 References begin on page 147 Chapter 3: One-Electron Oxidation Reactions ofCr(II) Complexes (92) Rachidi, I. E.-L; Eisenstein, O.; Jean, Y. New J. Chem. 1990,14, 671. (93) Riehl, J. F.; Jean, Y.; Eisenstein, O.; Pelissier, M. Organometallics 1992,11, 729. (94) Shannon, R. D.; Prewitt, C. T. Acta. Cryst. B 1969, B25, 925. (95) Shannon, R. D. Acta. Cryst. B 1976, A32, 751. 152 References begin on page 147 Chapter Four Reactivity of Cobalt(II) Halide and Alkyl Complexes with Alkyl Halides 4.1 Introduction Due to the similarity of one-electron redox behaviour of cobalt(II) to chromium(II),1"5 attention was turned to the reactions of cobalt(II) halide and alkyl complexes (Chapter 2) with alkyl halides. As will be seen, there are some real differences with the cobalt system that helped shed light on the Cr(II)/Cr(III) processes described in Chapter 3. Electron transfer and radical processes figure prominently in many different areas of chemistry.610 In organic chemistry, convenient and versatile sources of alkyl radicals are in demand to initiate a wide variety of reactions.1112 In transition metal chemistry, cobalt complexes in particular have been examined in the light of single electron transfer and radical-based chemistry,1319 in large part due to the discovery that the active site of Vitamin B 1 2 153 References begin on page 194 Chapter 4: Reactivity of Cobalt(H) Alkyl and Halide Complexes with Alkyl Halides contains a readily homolyzable Co(III)-carbon bond.20 This revelation sparked a wide interest in Co(III)-C bond forming reactions.21 Several different reagent types, including direct addition of in situ generated radicals to Co(II) systems22, have been described, but one of the most ubiquitous alkyl radical sources used is alkyl halides. Both one- and two-electron processes involving cobalt complexes and alkyl halides have been described.23,24 For example, the two-electron oxidative addition of alkyl halides to Co(I) derivatives (equation 4.1) is a classic route to Co(III) species:25,26 L nCo(l) + RX - L n Co(l l l )RX [4.1] More importantly, and directly analogous to the chromium(II) reaction is the addition of alkyl halide to Co(II) complexes to give Co(III) halide and Co(III) alkyl complexes in the same one-electron redox, radical-based bimolecular reaction mechanism as for the chromium(II) reaction (Scheme 4.1/Scheme 3.2):27 Scheme 4.1 L nM(ll) + RX - L nM(lll)X + R-L nM(ll) + R« - L nM(lll)R 2L n M(l l ) + RX L nM(lll)R + LnMfllOX Generally, the Co(II) systems studied have been low-spin, rigidly square planar systems (porphyrins, salen-type ligands) as these compounds are considered suitable models for Vitamin g 1 2 )5,28 although the classic report of addition of RX to Co(CN)53" shows that this reaction can be generalized beyond traditional planar tetradentate ligand type complexes.3,4,29 Despite this last example, however, studies involving alkyl halide addition reactions to »o«-tetrachelating or 154 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides non-macrocyclic Co(II) systems are not common. In one case, addition of alkyl halides to cobaltocene has been reported.30"33 Other than this, one-electron oxidation reactions of alkyl halides with a high-spin Co(II) species or a Co(II) compound containing a cobalt-carbon bond would appear to be unknown.25 High and low spin cobalt(II) complexes were prepared in Chapter 2 and this chapter examines their reaction with alkyl halides, the structure of the resulting five-coordinate cobalt(III) complexes and a brief look at their utility for alkyl radical generation. 4.2 Reactivity of Cobalt(II) Halide Complexes with Benzyl Halides The course of the reaction of cobalt(U) halide and alkyl complexes with alkyl halides and the products observed were found to vary with the choice of cobalt complex as well as with alkyl halide, resulting in considerable complexity. Each combination of reaction substrates will hence be discussed as separate units. (i) Synthesis and characterization of cobalt (III) dihalide complexes Addition of excess benzyl chloride (10 equiv) to a blue toluene solution of 2 results in a slow colour change to bright red to generate the cobalt(III) complex CoCl2[N(SiMe2CH2PPh2)2] (34) (Scheme 4.2). Similarly, reaction of 2 with benzyl bromide rapidly progresses to give the mixed bromo-chloro derivative CoBrCl[N(SiMe2CH2PPh2)2] (35). If this reaction is allowed to proceed over longer periods of time, a halide exchange reaction occurs to yield the dibromide complex CoBr2[N(SiMe2CH2PPh2)2] (36) and benzyl chloride (detected by *H NMR spectroscopy). 155 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides Scheme 4.2 Me2Si Me2Si Me2Si Me?Si PhCH2CI —Co—-CI PhCH2Br C o — B r Me2Si Me2Si -CI + 0.5 PhCH 2CH 2Ph P Ph 2 34 Me2Si PhCH2CI Me2Si + 0.5 | > S C | PhCH 2CH 2Ph P Ph 2 35 Me 2 S. > ^ PhCH2Br / M e 2 S ; ' Co + 0.5 " N B r PhCH 2CH 2Ph This last step is considerably slower than the one-electron oxidation, but it makes obtaining pure 35 difficult. Complex 36 is visible as an impurity by lH NMR spectroscopy. The dibromide 36 can also be made from CoBr[N(SiMe2CH2PPh2)2] (3) and benzyl bromide; in this case the reaction proceeds to completion within two hours. In all cases, the benzyl group ends up as one half equiv of bibenzyl (PhCFJ^CFI^ Ph), detected by *H NMR spectroscopy (2.85 ppm) and by mass spectroscopy (mle 182, M+). 156 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides The intermediacy of benzyl radicals was substantiated by performing the reaction in the presence of TEMPO, a radical trapping agent (equation 4.2). The benzyl-trapped product, N-benzyloxy-2,2,6,6-tetramethylpiperidine was identified by MS (mle 247, M + ) and by lH NMR spectroscopy.16 Control reactions showed that TEMPO does not react with CoX[N(SiMe2CH2PPh2)2] or with benzyl halides separately. [4.2] T E M P O 1 H NMR 4.82 ppm Five-coordinate Co(III) complexes are not that common. The vast majority of Co(III) complexes are octahedral and generally are diamagnetic (low spin d 6 ) 3 4 In contrast, complexes 34, 35 and 36 are paramagnetic, with solution magnetic moments of 3.6, 3.1 and 3;3 B.M. respectively. A value of 3.1 B.M. for 35 is consistent with a d 6 intermediate spin S = 1 system with some second-order spin-orbit coupling,35 and is within the observed range of the other five-coordinate Co(III) complexes known. The observed value of 3.6 B.M. for 34, however, is substantially higher; the reasons for this are uncertain. Despite the paramagnetism of these complexes, broad, shifted lH NMR spectra can be observed (Figure 4.1) and integration and variable temperature NMR spectra yield chemical shift information. For example, for CoCl2[N(SiMe2CH2PPh2)2] (34) integration easily identifies the SiM<?2 peak at 12.1 ppm. The ortho and meta protons of PPI12 have the same integration but the ortho proton is tentatively assigned to 7.2 ppm and the meta to 0.8 ppm based partly on the greater broadness of the ortho proton and also on the well known phenomenon that resonances 157 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides due to ortho and para protons in a phenyl ring shift in tandem and in the opposite direction to the meta protons.36 The para proton and the backbone methylenes also have the same integration but in this case the resonance due to the para proton, at 9.9 ppm, is sharp and relatively temperature independent while the methylenes, being much closer to the metal centre, are found at -58.4 ppm, much broader and more shifted. Note that only one other cobalt(III) paramagnetic lH NMR spectra has been reported to my knowledge.37 Figure 4.1 lU NMR spectrum of CoCl2[N(SiMe2CH2PPh2)2] (34) in C6D6(*). The X-ray crystal structure of CoBr2[N(SiMe2CH2PPh2)2] (36) was solved and reveals the nearly perfectly trigonal bipyramidal geometry around cobalt (Figure 4.2). The phosphines are trans-axially oriented, with a P-Co-P angle of 173.87(8)° (Table 4.1). The amide and two halides are equatorial, with the in-plane angles of 117.88(5)°, 121.4(2)° and 120.7(1)° an indication of the lack of distortion in this system. Note that an undistorted trigonal bipyramid is the theoretically predicted structure for a five-coordinate intermediate-spin d 6 complex.38,39 An 158 References begin on page 194 Chapter 4: Reactivity of Cobalt(H) Alkyl and Halide Complexes with Alkyl Halides Table 4.1 Selected bond lengths and angles in CoBr2[N(SiMe2CH2PPh2)2] (36). Atom Atom Distance ( A ) Atom Atom Distance (A) Co P(l) 2.264(2) Co Br(l) 2.375(1) Co P(2) 2.282(2) Co Br(2) 2.378(1) Co N 1.926(5) N Si(l) 1.734(5) Atom Atom Atom Angle (°) Atom Atom Atom Angle (°) Ft I) Co P(2) 173.87(8) Br(l) Co Br(2) 117.88(5) Br(l) Co N 121.4(2) Br(2) Co N 120.7(1) P(l) Co N 87.7(1) Br(l) Co P(l) 91.18(5) Si(l) N Si(2) 125.1(3) Br(2) Co P(l) 88.41(6) Figure 4.2 Molecular structure (ORTEP) and numbering scheme for CoBr2[N(SiMe2CH2PPh2)2] (36). 159 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides intermediate-spin d 6 complex half-fills the d x 2 . y 2 and d x y orbitals, thus removing the requirement for Jahn-Teller distortion. The electron occupancy in a low-spin d 6 complex such as Ir(R)X[N(SiMe2CH2PPh2)] or the high-spin d 3 chromium(III) complexes discussed in Chapter 3 will Jahn-Teller distort.4043 Of the few structurally characterized examples of five-coordinate cobalt(III) complexes, the majority are square pyramidal, usually C0L4X, where L4 represents a square planar ligand array such as porphyrin or salen and X is an alkyl or halide group.44"50 Although most of these are diamagnetic, a few are paramagnetic with peff = 2.4 - 3.25 B .M. 5 1 " 5 4 There are, to my knowledge, only three structures of trigonal bipyramidal cobalt(IH) complexes: CoCl3(PEt3)2,55 CoI3(PMe3)256 and Col3(SbPh3)2.57 As well, a series of CoX3(PR3)2 complexes have been studied spectroscopically and by EXAFS to yield structural information.58,59 From the X-ray and EXAFS studies, Co-P bond lengths in trigonal bipyramidal Co(III) complexes range from 2.28 - 2.34 A; in comparison, the Co-P bond lengths in 36 of 2.264(2) and 2.282(2) A are on the short side but otherwise unremarkable. Note that, again, the Co-P distances are sensitive to spin state; Co-P distances in diamagnetic octahedral Co(III) complexes such as CoMe2(PMe3)2(p-(CH2)2PMe2) (2.196, 2.174 A ) 6 0 are significantly shorter than in 36. The Co-Br distances of 2.375(1) and 2.378(1) A are unremarkable. Cobalt(III) amides (excluding porphyrin-type macrocycles) are uncommon. The homoleptic amide, high-spin trigonal Co[N(SiMe3)2]3, with Co-N and Si-N bond lengths of 1.870(3) and 1.754(2) A respectively, has been reported.61 In 36 the Co-N distance of 1.926(5) A and Si-N distances of 1.734(5) and 1.721(5) A indicate a lesser degree of amide lone-pair interaction with the metal centre. 160 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides (ii) Variable-temperature magnetic susceptibility of cobalt(III) dibromide 36 The scarcity of Co(III) compounds in a trigonal bipyramidal geometry is the source of the lack of detailed magnetic data of such systems. The CoX3(PR3)2 series of complexes55'56'58,59 are paramagnetic with [Xeff = 2.93-3.28 B . M . , comparable to data for 35 and 36, but Col3(SbPh3)2 has a room temperature magnetic moment of 4.4 B . M . , possibly consistent with some spin-equilibrium process.57 However, no detailed variable-temperature magnetic study of a trigonal-bipyramidal Co(III) system has been reported, either in the solid state or in solution. The solid state molar magnetic susceptibility (xm) of CoBr2[N(SiMe2CH2PPh2)2]*(0.5 C7H8) (36) was measured from 4.4 - 81.5 K and the results are shown in Figure 4.3 as a plot of magnetic moment vs. temperature. At high temperatures the magnetic moment tends towards the 3.3 B . M . observed at room temperature, consistent with some second-order spin-orbit coupling to an intermediate-spin d 6 S = 1 system. As the temperature is lowered the moment drops to 1.59 B . M . at 4.4 K, a behaviour which may be attributed to zero-field splitting (ZFS) of the S = 1 level into the ±1 Kramers' doublet and a non-magnetic S = 0 state, with a separation of D cm - 1. The data were analyzed using the equation for zero-field splitting of a S = 1 state,62 XCo(,n, = l / 3 [ C . r ^ , + 2 / 3 [ C . e f ^ p ] where C = (NAg2(32/kT) and x = (D/kT) respectively. The experimental data could be accurately fit to the model by the inclusion of a small amount of paramagnetic S = 3/2 impurity, likely the Co(II) starting material. This was accomplished by combining the above expression with the Curie law term, _ NAg2p2S(S + 1) Xpara - 3 k T 161 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides according to, Xm = [l-P]XCo(III) + PXpara where P represents the fraction of paramagnetic impurity (g = 2.3). Fits of the experimental data to the model were achieved using a nonlinear least-squares procedure with the following as the function minimized: 1/2 F = J _ 2 ( X i o b s - X i c a l c ) 2 R i = 1 ( X i ° b s ) 2 The F value provides a measure of the goodness of fit between experiment and theory. 5 T u I 1 ! 1 =—I 1 1 1 1 1 0 10 20 30 40 50 60 70 80 90 Temperature (K) Figure 4.3 Graph of magnetic moment vs. temperature for CoBr2[N(SiMe2CH2PPfi2)2] (36). 162 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides The best fit of the experimental data with theory (Figure 4.3) yields D = 32.6 ± 0.5 cm - 1, g = 2.15 ± 0.006 and P = 0.0523 ± 0.003 (F = 0.0159). A negative value of D was not fit adequately to the model, implying that a positive sign of D is valid. A change in the g-value of the paramagnetic impurity from 2.15 to 2.3 does not change the D or g value of the fit, but merely the amount of paramagnetic impurity P. The splitting value of 32.6 cm - 1 is quite large but as mentioned previously, there is a paucity of data on trigonal-bipyramidal Co(III) systems with which to compare this value. The S = 1 intermediate spin state has been observed in square planar and square pyramidal Co(III) complexes and the zero-field splitting parameter has been measured in a square planar K[Co(III)N4]»2Ff20 system (N4 = quadridentate diaminodiamido ligand) to be 53.7 c m - 1 . 6 3 , 6 4 Variable temperature measurements on square planar [?Bu3N][Co(NS)2] (NS = aminothiophenolato bidentate ligand) were not conducted to low enough temperature to measure the zero-field splitting.65 A series of square-pyramidal Co(III) complexes, Co(III)N4X (N4 = quadridentate diaminodiamido ligand; X = Cl, Br, I), was investigated in detail; the magnetism was found to be particularly complex and no zero-field splittings were reported.52,53 Square planar Fe(II) phthalocyanine is an S = 1 intermediate-spin system and a zero-field splitting value of 69.9 cm - 1 has been reported;66 iron(II) tetraphenylporphyrin is also an intermediate-spin compound but its magnetic behaviour is complicated by mixing of terms of similar energy.67 Although a comparison of dibromide 36 with S = 1 trigonal-bipyramidal Fe(II) complexes could be of interest, no such compounds have been reported. Other metal centres which exhibit S = 1 zero-field splitting include high-spin d 8 Ni(II) complexes and d 2 V(III) complexes.35,68 It would be interesting to be able to correlate zero-field splitting values for Co(III) with geometry and ligand field strength as has been done for high-spin Co(II) systems69"71 but there are insufficient data in different geometries at this time, although it does appear that square-planar coordination enforces greater zero-field splitting than a five-coordinate trigonal-bipyramidal geometry. 163 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides (iii) Kinetic and mechanistic points The UV-vis spectra of these compounds are dominated by an intense band around 500 nm and another band around 330 nm (Table 4.2). Using these bands, each dihalide derivative can be uniquely identified. Absorption spectra of C0X3P2 complexes have been extensively studied5 8'5 9 , 7 2'7 3 and on the basis of these similar compounds, the band observed around 500 nm is assigned as a phosphine to metal (P(G)->M(dXY) x 2- y 2)) charge transfer band; the extinction coefficient of 1770 M^cnr 1 for 34 supports this assignment. Similarly, the higher energy band is assigned as a phosphine to metal (P(cr)->M(dz2)) charge transfer band. An extra charge-transfer band in 36 at 434 nm is assigned as a bromine to metal (Br(7t)->M(dXy, x2-y2)) transition. One can observe d-d bands around 400-434 nm and also at low energy, but these are much weaker in intensity and appear at best as shoulders, or are obscured completely. Table 4.2 UV-vis spectral data for CoXY[N(SiMe2CH2PPh2)2]. X , Y UV-vis spectrum (nm, (e, IVHcnr1)) Cl, Cl (34) 326 (2380), 488 (1770) Cl,Br(35) 330 (2900), 496 (1900) Br, Br (36) 338 (2900), 434 (1370), 510 (2070) Dramatic colour changes and independent knowledge of spectra of complexes led UV-visible spectroscopy to become the method of choice for some preliminary kinetic investigations as well as compound identification. The reaction of PhCE^X (X = Cl, Brj and CoY[N(SiMe2CH2PPh2] (Y = Cl, Br) was convenientiy monitored by following the growth of an absorption due to the product CoXY[N(SiMe2CH2PPh2] around 500 nm. 164 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides The effect of changing the halide on the bimolecular rate constant k of the reaction was probed by reaction of 2 and 3 with benzyl chloride and benzyl bromide (equation 4.3). Ph 2 •P Me 2 Si Me? Si \ : N — C o — Y + PhCH 2X / I ^ Me 2 Si; Ph 2 •P -P Ph 2 Y = CI 2 Br 3 JN—Co" +0-5 / I V PhCH 2CH 2Ph Me 2 Si^ I 2 2 -P Ph 2 X = Cl, Br X = CI;Y = CI 34 X = Cl; Y = Br 35 X = Br; Y = Br 36 [4.3] Plots of kQbs vs. [PI1CH2X] were linear under pseudo-first-order conditions, confirming the bimolecular nature of the reaction. Data give a rate constant of k = 8.4 ± 0.5 x 10 -5 M^s - 1 for PI1CH2CI and k = 1.6 + 0.5 x 10"2 M ' V 1 for PhCH2Br addition. The much faster rate for benzyl bromide, coupled with the observation of bibenzyl, supports the generally accepted radical-based mechanism for this type of reaction.4,5'27'29'74"77 In addition, consistent with a radical-based mechanism, the reaction does not occur with unactivated substrates such as chlorobenzene. A complicating factor in the reaction of 2 with benzyl bromide is the slower ongoing chloride for bromide substitution reaction, i.e., conversion of 35 to 36. To account for this potentially overlapping reaction kinetically, the reaction of 3 with benzyl bromide was also monitored. The rate constant for this reaction is k = 2.0 ± 0.5 x 10"2 M 'V 1 , implying that the halide exchange side-reaction does not greatly affect the overall kinetic data. Note that no peaks characteristic of 36 (e.g. the extra CT band at 434 nm) are observed in kinetic experiments starting with cobalt chloride 2. 165 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides According to the classical mechanism shown in Scheme 4.1, two products are expected: Co(CH2Ph)X[N(SiMe 2 CH 2 PPh2] ( C ) from alkyl transfer and CoX2 [N(S iMe2CH 2 PPh 2 ] from halide transfer. However, no evidence for the former compound was observed, either by N M R or by U V - v i s spectroscopy. To test whether the complex C o ( R ) X [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] (R = M e , D) would be stable, we attempted to metathesize one halide of dichloride 34 with various alkylating agents in order to generate the putative intermediate C o ( R ) X [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] independently. Addi t ion of one equivalent of an alkyll i thium or Grignard reagent ( C H 3 L i , P h C H 2 M g C l , L i C H 2 S i M e 3 ) to the dichloride, C o C l 2 [ N ( S i M e 2 C H 2 P P h 2 ] (34), resulted in the formation of dark blue 2, the Co(II) starting material, identified by its visible spectrum. Further addition of reagent formed the Co(II) alkyl species as would be expected. This result suggests that the five-coordinate Co(III) alkyls are intrinsically unstable and decompose upon formation by homolytic cleavage of the Co(III)-C bond to give a Co(II) product and the a lkyl radical (Scheme 4.3). Attempts to stabilize the cobalt(III) a lkyl halide complex by performing the reaction at lower temperatures were not successful; thus addition of alkyll i thium reagents to C o C l 2 [ N ( S i M e 2 C H 2 P P h 2 ] (34) at -78 °C caused a colour change from red to yellow/orange, which upon being warmed turned to the blue colour of the Co(U) species. This low-temperature colour change may indeed be the intermediate a lkyl halide complex but this proved unisolable despite attempts at adding trapping agents (phosphine, pyridine, CO) at low temperature. Wi th regard to the one-electron oxidation of C o X [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] by R X , the stoichiometry of this reaction is therefore 1:1, and not 1:0.5 as is found in classical systems (Scheme 4.1). This is due to the fact that alkyl radicals, R», do not form stable complexes with the Co(II) present. 7 8 The facile homolyzability of Co(III)-carbon bonds is recognized in the literature; macrocyclic ligands and/or the proper choice of axial base tend to stabilize these bonds. 7 9 " 8 2 Stable non-macrocyclic a lkyl complexes have been prepared but al l are octahedral, diamagnetic and therefore sterically and electronically saturated. Examples include CoMe3(PMe3)3 , 8 3 C o R 2 ( a c a c ) ( P R ' 3 ) 2 8 4 and C p C o L R R ' . 8 5 Surpr is ingly , addition of triethylphosphine to the dihalide complexes 34 or 35 showed no change in either the * H N M R or 166 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides UV-vis spectra and no alkyl species could be trapped by performing metathesis reactions in the presence of donor ligands. The stability of the five-coordinate, 16-electron complexes toward neutral donor addition could be due to sterics, but an examination of the X-ray crystal structure suggests that such steric crowding is not present. A more likely explanation lies in the stability of the five-coordinate triplet state vs. the six-coordinate geometry, which would have to undergo spin pairing to form a singlet state. Poli has observed that in many situations,86'87 in particular in fifteen-electron Cr(III) systems,88 the addition of a donor ligand is insufficient to overcome the electronic cost of spin pairing and despite electronic unsaturation at the metal, donor binding is not observed. Scheme 4.3 Ph5 Me2Si Me2Si \ i : N — CcL / X Ph 2 X = Cl (34), Br (36) M'R = MeLi, MeMgBr, KCH 2Ph, LiCH 2SiMe 3 M'R -78 °C Me2Si Ph 2 •P R MeoSi :N—Cc* / I ^ X Ph 2 R = CH 2Ph C Me D warm to rt Ph 2 •P Me2Si Me2Si \ I : N — C o — X + R* / | Ph 2 X = Cl (2), Br (3) Note that addition of an organolithium or Grignard reagent to 34 or 35 effectively produces the alkyl radical due to a one-electron reduction to Co(II). Monitoring the reaction of 167 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides LiCH2SiMe3 with dichloride 34 in an NMR tube resulted in the formation of Me4Si and PhCH2CH2Ph (bibenzyl), presumably from a neosilyl radical (Me3SiCH2#) abstracting a proton from the toluene solvent. The resulting benzyl radicals then coupled to form bibenzyl (Scheme 4). The observation of these products supports the intermediacy of alkyl radicals in this chemistry. This implies that complexes 34-36 can be used to generate alkyl radicals in situ given the appropriate organolithium or Grignard reagent. Co(III) complexes have been used as stable radical sources; the complexes here have the potential to be generators of a wide range of alkyl radicals. Scheme 4.4 Co(R)X[N(SiMe 2CH 2PPh 2)2] CoX[N(SiMe 2CH 2PPh 2)2] + R # R» + P h C H 3 - RH + PhCH 2 « 2 PhCH 2 « ^ P h C H 2 C H 2 P h 4.3 Reactivity of Cobalt(II) Halide Complexes with Methyl Iodide As described above, the reaction of the high-spin Co(II) halides 2 and 3 with benzyl halides yields red cobalt(III) products by a one-electron oxidation mechanism. In contrast, substitution of the benzyl halides by methyl iodide, however, yielded no Co(III) oxidation product. Instead, addition of excess methyl iodide to a blue toluene solution of CoCl[N(SiMe2CH2PPh2)2] (2) causes a purple precipitate to form over one hour. The precipitate was identified by spectroscopic analysis as a phosphonium dihalocobaltate zwitterion CoXY[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (X = Cl, Y = I (37); X = Y = I (38)), effectively the product of Mel addition to 2 (equation 4.4). A crystal structure of a similar zwitterion (described later) aided in this identification. 168 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides P h 2 P P h 2 P M e 2 S i \ xs Mel M e 2 S i I N — Co—CI [4.4] M e 2 S i / Me2Si P P h 2 MePh2P0 2 X = CI 37 I 38 In the MS the molecular ion peak is unobservable but the peak for M + - X is strong. The MS also indicates that an inseparable mixture of halide-containing species is present; these likely include diiodo, iodochloro and perhaps even dichlorocobaltate anions; hence, halide exchange is also occurring. Products containing two methyl phosphonium halides may also be present; essentially, this is a complicated reaction with many products, all zwitterionic and hence mostly insoluble in toluene. The precipitate is soluble and stable in CH2CI2 and hence X H NMR and 3 1P{ 1H} NMR spectra could be observed. The ! H NMR spectrum is complex, with multiple peaks (no coupling) due to the presence of multiple products; in the 3 1P{ lH) NMR spectrum, the non-coordinated, tetrahedral phosphonium can be seen at +70 to +90 ppm, depending on the halide groups. Attempts to purify single compounds from this mixture were unsuccessful, but the important feature to recognize is that upon changing alkyl halide reagent from benzyl bromide/chloride to methyl iodide, no redox chemistry and no Co(III) products are observed. The key to this change in reactivity lies in the reactivity of alkyl halides with free phosphines, which has been well studied89 and predicts a reactivity trend based on X of I>Br»Cl for the reaction shown in equation 4.5: P R 3 + R'X [R'R 3P]X [4.5] 169 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides Hence, free phosphines react with alkyl halides, and the reaction proceeds most quickly with alkyl iodide substrates. In CoX[N(SiMe2CH2PPh2)2] the chelating phosphines are bound to the metal centre, precluding this line of reactivity but in many chelating phosphine systems an on-off fluxionality is observed.90 Me 2 Si MeoSi \ : N -P h 2 •P •Co—iCI ^ P h 2 2 Me 2 Si P h 2 •P ^ N — C o — C I [4.6] MeoSi / P h 2 P : Systems containing this ligand on diamagnetic metals have also exhibited this property.90 As a result, an equilibrium could exist (equation 4.6) and the dangling phosphine then is available to react with the large excess of alkyl halide present. Rate data available89 for the reaction of Mel with PPh3 indicate that this side reaction would be competitive and in the case of Mel and 2, the side reaction dominates the chemistry. The analogous rates for benzyl bromide and chloride are slower and, in these cases, redox chemistry predominates. In fact, the rate for PhCH2Br and PPh3 reaction is only approximately three times slower than with Mel and with a very high concentration of PhCH2Br, some CoBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2CH2Ph)] zwitterion (39) is produced. On the other hand, even in neat benzyl chloride, no zwitterion was observed. 170 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides Essentially, this zwitterion, although an interesting product in its own right, is a side-product that can be minimized by avoiding very large excesses of alkyl halide, and by avoiding the use of Mel as a substrate with 2. 4.4 Reactivity of Cobalt(II) Methy l Complex 6 with Methy l Hal ides (i) Synthesis of cobalt(II) halide complexes by redox substitution Addition of one equivalent of CH3I to a yellow toluene solution of 6 results in the slow formation of a green solution with visible spectral bands at 538, 634 and 806 nm. This is similar to the bands of 2 and 3 (Table 2.3) and the product is identified by the visible spectrum, NMR spectrum, elemental analysis and mass spectrum as CoI[N(SiMe2CH2PPh2)2] (4). Similarly, 3 can be prepared by addition of CE^Br to 6 (equation 4.7). M e 2 S MeoS — C o — M e 1 equiv M e 2 S k MeX \ Toluene Me 2 Si rt / P h 2 P + 2 MeH : N — C o — X •P P h 2 X = Br 3 I 4 P h C H 2 C H 2 P h [4.7] The final product, then, is still a Co(II) product: apparently, a halide for alkyl metathesis has occurred and not an oxidation. This result is seemingly contrary to what one would expect based on the classical mechanism. Thus, halide abstraction by the cobalt(II) alkyl 6 should be the first step, to give Co(Me)X[N(SiMe2CH.2PPh2)2 (D); however, this complex was shown (in 171 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides 4.2 (i)) to rapidly decompose by Co-R bond homolysis to generate the observed product CoX[N(SiMe2CH2PPh2)2] and presumably methane from the methyl radical abstracting a proton from toluene (bibenzyl is also observed). For this reason, although a halide for alkyl substitution has occurred, an oxidation followed by a fast reduction is likely responsible mechanistically. The more interesting point is the fact that this route is followed at all; recall that Mel is the reagent that reacts with cobalt(II) chloride 2 to form exclusively zwitterion. Now, with cobalt(II) methyl complex 6, Mel is undergoing redox chemistry. The inherent and fundamental difference between 2 and 6 that provides for the different observed reactivity is the fact that CoX[N(SiMe2CH2PPh2)2] (2) is tetrahedral and high-spin while CoMe[N(SiMe2CH2PPh2)2] (6) is square planar and low-spin. Why is this important? It was suggested earlier that zwitterion formation occurred via alkyl halide reaction with a dangling, free phosphine from the chelate arm. High-spin Co-P bonds are substantially longer and weaker than low-spin Co-P bonds (2.373(3) A in 4 vs. 2.191(2) A in 6) and hence have a higher propensity to dissociate. This easily dissociable phosphine in the high-spin system 2 can react with exogenous methyl iodide more readily than the relatively tightly bound phosphine in low-spin 6, and this is one reason for the reactivity difference observed. In addition, the geometry difference may be important as it affects the orbital availability and accessibility of the unpaired electrons. The unpaired electrons on the metal must be accessible to external substrates in order for one-electron redox chemistry to occur. In high-spin, tetrahedral 2, although there are three unpaired electrons, they reside in the d x y , d x z and d y z orbitals, which in a tetrahedral geometry may be relatively sterically blocked, limiting their accessibility to external substrates. On the other hand, in low-spin, square planar 6, there is only one unpaired electron but it is located in the d x y orbital, which is wide open on either face to external substrate attack. Hence it may be easier for square-planar 6 to undergo redox chemistry than it is for tetrahedral 2. This factor, combined with Co-P bond strength differences can readily rationalize the observed differences in reactivity towards Mel between CoX[N(SiMe2CH2PPh2)2] and CoMe[N(SiMe2CH2PPh2)2].87 172 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides (ii) Variable temperature magnetic susceptibility of cobalt(II) iodide complex 4 The solid state magnetic susceptibility (%m) of iodide 4 was measured from 4.2 - 80 K and the results are shown in Figure 4.4 as a plot of magnetic moment vs. temperature. At high temperatures the magnetic moment is 4.2 B.M. and is greater than u. s o., consistent with some second-order spin-orbit coupling to an excited S = 3/2 state35 and is comparable to other tetrahedral Co(II) complexes.91,92 As the temperature is lowered the moment drops to 3.45 B.M. at 4.2 K. This behaviour at low temperatures is typical for S = 3/2 d 7 systems such as Co(II) and is indicative of zero-field splitting (ZFS) of the ground S = 3/2 level into the ±3/2 and ±1/2 Kramers' doublets, with a separation of D cm - 1 . 3 5 , 7 1 The data were analyzed using the equation for zero-field splitting of a S = 3/2 state:62 where x = (D/kT) and C = (NAg2P2/kT) respectively. Fits of the experimental data to the model were achieved using a nonlinear least-squares procedure as described in section 4.2 (ii). Experimental susceptibility vs. temperature data for iodide 4 are compared with the best fit results from theory (Figure 4.4) with D = 20 ± 1 cm'1 and g = 2.24 ± 0.01 (F = 0.0122). A value of -20 cm - 1 for D yielded an identical fit; hence the sign cannot be determined. The splitting magnitude is large but of the order observed for other tetrahedral Co(II) systems, which generally have 2D values around 7-11 cm - 1 . 6 8 The zero-field-splitting in CoCi2(PPh3)2 was determined as -3 c m 4 in the solid state, and +7.5 cm - 1 in frozen ethanol glass. Similar data is obtained from CS3C0X5 (X = Cl, Br) complexes.69,93 Studies have correlated the geometry around high-spin Co(II) with the magnitude of the zero-field splitting; a distorted tetrahedron has the smallest ZFS, with five coordinate and octahedral ZFS being larger and largest respectively.69 The relatively large D value observed for iodide 4 compared to other tetrahedral Xm = 1/3 [C • 1 + 9e"2x -] + 2/3 [C < 4 + (3/x)(l - e-2x). 4(1+ e-2x) 4 (1 + e-2x) 173 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides systems could be due to the fact that the D value is observed to increase with substitution of a larger halide. Orbital contributions to the moment vary inversely with the strength of the ligand field, and hence, within the same geometry weaker-field ligands should give rise to larger zero-field splittings. The magnetic moments of tetrahalocobaltate(II) anions, C0X42", reflect this trend, ranging from 4.59 B.M. for X = CI to 4.77 B.M. for X = I. 3 4- 6 8 There seem to be no measurements of D using Co(II) tetrahedral complexes containing iodides, but if the above trend is followed, a relatively large value of D compared to chloride and bromide containing species would be expected; this is observed in iodide 4. A direct observation of this trend has been reported in the series of square-pyramidal intermediate-spin d 5 (S = 3/2) FeX(S2CNEt2)2 (X = CI, Br, I) complexes where D ranges from 3.4 cm - 1 (X = CI) to 19.5 cm - 1 (X = I).94 Another factor contributing to the large D value could be the much reduced symmetry of 4 compared to most compounds studied; 4 is a C0A2XY system with at best C s symmetry as opposed to other complexes which are usually Td, D2d or D3 symmetry.95 5.0 -, 1 Figure 4.4 Graph of magnetic moment vs. temperature for CoI[N(SiMe2CH2PPh2)2] (4). 174 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides 4 . 5 Reactivity of Cobalt(II) Methyl Complex 6 with Benzyl Halides (i) Synthesis and structure of a zwitterionic cobalt(II) complex Addition of one equiv of benzyl chloride to 6 gives two products over several hours, their proportions depending on the relative concentration of 6 and RX. One product is 2, the halide for alkyl metathesis product. A second product precipitates out of toluene as a purple powder. If the reaction is done with excess benzyl chloride, after the formation of 2, the reaction continues on to form 34 the halide transfer product, but this step is slow enough that the formation of 2 is readily observable. That is not the case for benzyl bromide addition, where the primary product CoBr[N(SiMe2CH2PPh2)2] (3) rapidly continues reacting with benzyl bromide to give CoBr2[N(SiMe2CH2PPh2)2] (36), both identified by UV-visible spectroscopy. In this reaction, over time, a purple powder also precipitates out of solution (Scheme 4.5). Scheme 4.5 r Ph 2 Phs C o — M e : N — C o — X / / 6 X = CI 2 Br 3 Ph 2 v. xs PhCH2X Ph, Ph 2 / : N — C o — X X MePh 2P0 x = Cl 41 Br 42 Ph 2 X = CI 34 Br36 175 References begin on page 194 Chapter 4: Reactivity of Cobalt(H) Alkyl and Halide Complexes with Alkyl Halides Table 4.3 Selected bond lengths and angles in CoCl2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (41). Atom Atom Distance (A) Atom Atom Distance (A) Co P(l) 2.3593(6) Co N 1.968(2) Co Cl(l) 2.2784(6) Co Cl(2) 2.2747(6) N SKD 1.719(2) N Si(2) 1.706(2) Atom Atom Atom Angle O Atom Atom Atom Angle (°) N Co Cl(l) 115.06(5) Cl(l) Co Cl(2) 110.90(2) N Co Cl(2) 117.28(6) N Co P(l) 92.20(5) Cl(l) Co P(l) 104.77(2) Cl(2) Co P(l) 114.73(2) Figure 4.5 Molecular structure (ORTEP) and numbering scheme for CoCl2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (41). 176 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides These purple powders, on the basis of previous observation, were readily identified as zwitterion species, but in this case a pure, single compound could be isolated. Single crystals of the zwitterion CoX2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (X = CI ( 4 1 ) ; X = Br ( 4 2 ) ) formed from the reaction of 6 with benzyl chloride were obtained by performing the reaction without stirring, allowing single crystals of zwitterionic product to slowly deposit over several days. The X-ray crystal structure is shown in Figure 4.5 and selected bond lengths and angles in Table 4.3. The structure reveals a tetrahedral cobaltate centre and also a tetrahedral, unbound phosphonium cation that is dangling away from the metal. The zwitterionic nature of the product accounts for its relative insolubility in non-polar solvents. The Co-P bond distance of 2.3593(6) A is typical for a high-spin Co(II) centre, although the Co-N bond distance of 1.968(2) A is on the long side for other comparable complexes. The N-Si bonds of 1.719(2) and 1.706(2) A are substantially more asymmetric then in other compounds presented; this is in keeping with the general asymmetric nature of the complex. The Co-Cl bond distances of 2.2747(6) and 2.2784(6) A are considerably longer than those found in Co(PPh3)2Cl2;93 likely this is due to the presence of a formal negative charge on the cobalt. The key feature of this complex, other than confirming the nature of the purple precipitate beyond all doubt, is the identity of the phosphonium cation: the added alkyl is methyl, despite the reaction occurring in a large excess of benzyl halide. This implies that an intramolecular mechanism must be responsible for this zwitterion formation, contrary to the zwitterion formation observed previously. Mass spectrometry is an excellent tool for identifying these zwitterionic species and it was confirmed that this intramolecular transfer is indeed occurring; the M+-X peak is quite clear in the spectrum and one can hence easily differentiate which phosphonium alkyl is present. The reaction has some generality; reaction of Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] ( 1 0 ) with excess benzyl chloride gave some intramolecular zwitterion CoX2[N(SiMe2CH2PPh2)-177 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides (SiMe2CH2PPh2CH2SiMe3)] (40) with the neosilyl group present at the phosphonium centre (equation 4.8), as well as the expected Co(II) halide species. Me 2Si \ Ph 2 -P © Ph 2 •P Me 2Si : N — C o — B r 40 Me 2Si MeoSi \ j — C o — C H 2 S i M e 3 P Ph 2 10 xs PhCH 2Br \ ^ Br 0PPh2CH2SiMe3 + Ph, [4.8] Me 2Si Me 2Si \ i — C o / Br 36 Br Ph, Interestingly, the reaction with Co(CH2Ph)[N(SiMe2CH2PPh2)2] (8) proceeds rapidly to give only CoX[N(SiMe2CH2PPh2)2]; no zwitterion product is observed at all in this case (equation 4.9). Ph 2 P Me 2Si Me 9Si \ xs d 7 - PhCH 2Br / • N — C o — C H 2 P h 36 only [4.9] •P Ph 2 8 178 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides In order to determine the fate of the alkyl group on the cobalt in the case of halide-for-alkyl substitution, the reaction of Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] (10) with excess benzyl bromide was followed by lH NMR spectroscopy. In the presence of excess benzyl bromide, the formation of Me3SiCH2Br was clearly observed. Bibenzyl was also produced. The presence of neosilyl bromide can be attributed to the formation of Me3SiCH2» radicals, which can abstract a bromide from the excess benzyl bromide present. This result also illustrates the inherent instability of the corresponding cobalt(III) complex Co(CH2SiMe3)Br[N(SiMe2CH2PPh2)2] (E), which hence decomposes by rapid homolysis of the cobalt-carbon bond, as shown in other experiments. In addition, if the reaction of 6 with benzyl chloride is conducted in a sealed NMR tube, bibenzyl is produced and peaks assignable to 2 are observed. Some zwitterionic 41 precipitate also forms. (ii) Mechanism of intramolecular zwitterion formation A possible mechanism for intramolecular formation of 41 is presented in Scheme 4.6 that encompasses the observations discussed in the previous section and is based on the concept of radical cages and phosphorus radical species. Other mechanisms that involve cobalt(I) or cobalt(IV) intermediates must at some point invoke nucleophilic attack of the phosphine on the coordinated methyl-group, an improbable situation with no precedent. Attack of a coordinated alkyl group on a phosphenyl cation (PR2+) has been documented96 but it is difficult to imagine such species in this system. In any mechanism proposed, the first step is invariably one-electron oxidation and halide-transfer from benzyl halide to the cobalt(II) methyl complex, to form the cobalt(III) alkyl halide complex Co(Me)X[N(SiMe2CH2PPh2)2] (D). This complex has already been shown to 179 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides Scheme 4.6 PhCHoX M e 2 S i P h 2 •P PhCH 2 X P h 2 - P M e 2 S i ^ : N — C o M e 2 S i ^ * \ / MePh2P0 X0 Halide transfer Me 2 Si \ : N — C o / I V -P Ph 2 Me P h 2 -1 M e 2 S i ^ : N — C o — X M e 2 S i ^ | Me* P h 2 Radical Cage Electron transfer X = CI 41 Br 42 180 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides decompose via Co-Me bond homolysis to yield the cobalt(II) halide complex and methyl radicals. It is proposed that prior to this decomposition, Co-C bond homolysis leads to the formation of an organometallic radical cage species, with the alkyl (methyl) radical and the cobalt(II) complex making up the cage. Cage effects in organometallic chemistry have been by and large overlooked,97"100 but there is precedent for such effects being important in reaction mechanisms and in some cases being the dominant factor in determining observed products.97,101 The radical cage pair has several modes of decomposition and it is at this point where different products can be observed. Firstly, cage escape yields the observed cobalt(II) halide product along with the methyl radical, which undergoes further secondary reactions (with solvent, Scheme 4.4). The two radicals can, of course recombine to form the cobalt(III) alkyl halide complex. Finally, it is postulated that the methyl radical can add to one phosphine arm to form a four-coordinate phosphoranyl radical F. The reaction of three-coordinate phosphines with radicals has been well studied and it was determined that methyl radicals react reversibly with phosphorus(III) centres,102 ,103 and that benzyl radicals are generally found to be unreactive to phosphine.102,103 This important point explains the lack of zwitterion formation when starting with the cobalt(II) benzyl complex; the resulting benzyl radical does not react with phosphine and hence only cage escape products are observed. The addition of methyl radicals to triphenylphosphine to give the phosphoranyl radical PPh3Me» has been illustrated in a key paper that describes the photolytic reaction of CpW(CO)3Me in the presence of PPh3 to give [PPh3Me]+[CpW(CO)3]-, a final product that is reminiscent of that observed in the cobalt system here.104 The tungsten was proposed to progress through a phosphoranyl radical; in the cobalt system, the extension is that the radical is tethered to the complex by a chelate arm, while in the tungsten example, the phosphine is exogenous. The fact that the formation of the methylphosphoranyl radical F is reversible is important as it implies that in the absence of a suitable trapping agent, no zwitterion should be produced. That is in fact the case; alkylation of CoCl2[N(SiMe2CH2PPh2)2] (34) to give the cobalt(III) 181 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides alkyl halide complex D followed by decomposition does not form any zwitterion. Excess benzyl halide is necessary to trap the phosphoranyl radical species F, hence alkylation of 34 at low temperatures, addition of benzyl halide and warming to room temperature results in exclusively zwitterion formation. The preferential formation of zwitterion at low temperatures is also consistent with the mechanism being radical-cage-based. Radical cages are more stable at low temperatures98 and the longer the cage lifetime, the more the zwitterionic pathway should be preferred. The nature of the trapping of the phosphoranyl species F by benzyl halide is difficult to determine. There are three possible mechanisms for such trapping. One-electron oxidative halide-transfer by benzyl halide at the cobalt(TJ) metal centre to give a cobalt(III) dihalide species H, followed by electron transfer from the phosphoranyl radical RMePi^P* would yield the final product. Alternately, one-electron oxidative halide-transfer of benzyl halide to the phosphoranyl radical RMePl^P* directly to give an endogenous phosphonium halide salt G, followed by halide transfer to the cobalt(II) metal centre is a viable route. Initial electron-transfer of the phosphoranyl radical to cobalt to yield a cobalt(I) species, followed by benzyl halide one-electron oxidative halide-transfer is also possible. The difficulty of determining the trapping route was noted in the CpW(CO)3Me reaction with PPI13, in which a number of routes to the final product exist.104 This proposed mechanism (Scheme 4.6), while speculative, accounts for the observations in the previous section. The intramolecular transfer of the methyl group to phosphorus is explained and occurs in the presence of a large excess of benzyl halide. When no benzyl halide is present, no zwitterion is formed as no trapping agent is present. The longer lifetime of radical cages at lower temperatures can explain the preferential formation of zwitterion at low temperatures. Finally, the lack of zwitterion formation from the cobalt(II) benzyl complex can be rationalized by the lack of reactivity of benzyl radicals with phosphines. 182 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides 4.6 Summary and Comparison with Chromium(LT) Redox Chemistry An overview of the reaction with alkyl halides of cobalt(II) complexes incorporating the tridentate ligand -N(SiMe2CH2PPh2)2 may now be elucidated (Scheme 4.7). The starting Co(II) halides and Co(II) alkyls can easily be prepared and are linked with an unusual paramagnetic five-coordinate Co(III) dihalide and to each other via a series of alkyl halide reactions. In general, the reactivity of these species does follow the expected classical mechanism, with a few twists. The instability of the species Co(R)X[N(SiMe2CH2PPh2)2] to cobalt-carbon bond homolysis is a vital factor in this reaction outline; this instability results in different stoichiometrics observed relative to those in classical systems. A side-reaction with the chelating phosphine arms to form unusual zwitterionic species is sometimes observed, depending on the alkyl halide reagent used and also depending on the geometry and spin state of the cobalt(II) system in question. The observation that high-spin tetrahedral cobalt(II) halides are more susceptible to intermolecular zwitterion formation than low-spin square planar alkyls stresses the importance of spin state and geometry with respect to the reactivity observed.87 It has also been demonstrated that these complexes can act as radical generators; the Co(III) complexes can convert organolithium or Grignard reagents into their respective radicals and, perhaps more interesting, the Co(II) complexes can in some cases generate alkyl radicals from a given alkyl halide. The reaction involving intramolecular zwitterion formation from organometallic cobalt(II) complexes, mechanistically invoking phosphoranyl radicals and radical cage pairs, deserves further investigation because of its unusual nature. It is now possible to look back at the chromium(II) reactions with alkyl halides and interpret the results with a new perspective. Firstly, it is obvious that the same classical mechanism is followed in both the chromium(II) and cobalt(II) systems. The unexplainable precipitate formed upon addition of Mel to the chromium(II) chloride complex 1 is likely a zwitterionic species (section 3.4 (iv)); the elemental analysis fits quite well with a methylphosphonium diiodochromate species. Although Mel did show redox behaviour with the 183 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides chromium(II) methyl complex 5 (section 3.4 (i)) large excesses of Mel led to precipitates that now can be labelled as zwitterionic species. Hence, in both the cobalt and chromium systems excesses of alkyl iodides led to zwitterionic products. Scheme 4.7 Ph 2 184 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides The decomposition of chromium(III) dialkyl complexes is a difficult issue; the nature of the decomposition was not addressed. In the light of the cobalt results, the homolytic cleavage of Cr-R to give a chromium(II) a lkyl complex and an alkyl radical must be considered. Although chromium(II) alkyls were synthesized by separate procedures, they are not easily observable by mass spectrometry and in the absence of a pure crystalline product are difficult to assign definitively (Chapter 2). They are also generally brown in colour, the same colour observed during the decompositions. However, while a clean homolytic cleavage may be occurring, there are l ikely other, less simple decompositions also active. The fact that decomposition products often had organic-type mass spectra with no characteristic ligand peaks, and the messy nature of some reactions (large excesses of a lkyl halide are problematic, for example), imply a more complex situation. The nature of this decomposition could be examined in more detail in future studies. Upon comparison of the two systems, the different stabilities of the trivalent a lkyl c o m p l e x e s are o b v i o u s . T h e c h r o m i u m ( I I I ) a l k y l h a l i d e c o m p l e x e s Cr(R)X[N(SiMe2CH2PPh2)2] are stable systems; crystal structures of two of these complexes (R = M e , X = B r (24); R = CH2SiMe3 , X = CI (26)) were presented in Chapter 3. On the other hand, the analogous cobalt(III) complexes are unisolable, decomposing by cobalt-carbon bond homolysis. That is, 16-electron cobalt(III) a lky l halide complexes are unstable, while the electron-deficient chromium(III) complexes are isolable solids. In addition, 16-electron cobalt(III) dia lkyl systems CoRR'[N(SiMe2CH2PPh2)2] are completely inaccessible but in the chromium case, the use of bulky a lkyl groups allows for the isolation (and crystal structure analysis) of such species (R = R' = CH2SiMe3 (27)). The difference in stabilities is difficult to ascribe to a single source. The inherent ease of cobalt(III)-carbon bond homolysis (relative to chromium) could be a factor. The potential for a spin-state change in the cobalt system to a diamagnetic S = 0 system upon alkylation could also account for the instability of such systems; cobalt(III)-carbon bond homolysis has been observed to occur preferentially from the cobalt(III) S = 0 state. 1 0 5 Such a spin-state change is impossible in a chromium(III) system. 185 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides Finally, an appropriate literature example serves to conclude this discussion. The paper, entitled "Preparation of Alkylchromium Reagents by Reduction of Alkyl Halides with Chromium(II) Chloride under Cobalt Catalysis", involves the in situ preparation of chromium(III) alkyl complexes which are used as reagents to add the alkyl group to aldehydes.106 It was found that the addition of alkyl halides to CrCl2 often resulted in halide exchange rather than oxidative addition. It is known that chromium(II) reacts with alkyl radicals directly to form chromium(III) alkyls1,2 and so a cobalt(II) catalyst (either Vitamin B12 itself or Co(phthalocyanine)) was added to generate alkyl radicals; the cobalt(III) alkyl complex rapidly homolyzes and the resulting radical adds to chromium(II) to form the desired product. The elegant interplay between cobalt and chromium metal centres in the +2 and +3 oxidation states and the transfer of the alkyl group through the system from cobalt(III) to chromium(II) to yield a chromium(III) alkyl are shown in Scheme 4.8 (from reference 106). Scheme 4.8 Cr 11 R-X ^ [ R-Co " ' X ] ^ > [ R. ] ^ > [ R - C r m ] C o 1 — C o n X R'CHO R ? H R ' + C r n C r n i X C r n r OH (H 2 0 workup) This example illustrates many of the features of chromium(II) and cobalt(II) reactions with alkyl halides in general: the relative stability of chromium(III) alkyl complexes over cobalt(III), and the relative simplicity of cobalt(II) systems compared to chromium(II) systems. 186 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides It is interesting to note that the study required a cobalt(II) catalyst to generate alkyl radicals for chromium(II) to react with; in the systems reported in this thesis, each metal undergoes one-electron oxidation reactions independently. On the basis of the work described here, it is clear that the ancillary ligands exert a profound influence on the course of one-electron redox reactions with cobalt(II) and chromium(II). 4.7 Experimental 4.7.1 General Procedures and Materials Unless otherwise stated all procedures were followed as described in section 2.8. Alkyl halides were either distilled under N 2 or passed through a column of activated alumina, and then degassed by three freeze-pump-thaw cycles. All other reagents were obtained from commercial sources and used as received. 4.7.2 Variable Temperature Magnetics Measurements. Magnetic susceptibility measurements on powdered samples of iodide 4 and dibromide 36 were made at an applied field of 7500 G over the temperature range 4.6 to 80 K using a PAR Model 155 vibrating-sample magnetometer as previously described.107 Magnetic susceptibilities were corrected for the diamagnetism of all atoms using Pascal's constants. 4.7.3 Kinetic Measurements Kinetic measurements on alkyl halide addition to cobalt(II) halides were performed by monitoring the growth of an appropriate wavelength absorbance of the product in the UV-vis spectrum; the band around 500 nm was generally used. Typical metal-complex concentrations were approximately 10 - 4 M, yielding AOD values of approximately 1.5 absorbance units. 187 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides Reactions were conducted under pseudo-first order conditions; the alkyl halide used was present in at least a ten-fold excess. The rate constant k 0b s was determined from the slope of the In A vs. time plot. The final second-order rate constant k was determined from the slope of the straight line plot of kobs vs alkyl halide concentration; pseudo-first order rates at at least three different concentrations of RX were measured. 4.7.4 Synthesis and Reactivity of Complexes (i) Synthesis of CoCl2[N(SiMe2CH2PPh2)2] (34) To a blue 15 mL toluene solution of CoCl[N(SiMe2CH2PPh2)2] (2) (0.15 g, 0.24 mmol) was added excess (0.5 mL) benzyl chloride neat. No immediate colour change occurred but overnight stirring gave a brown/red solution. After being stirred for four more days a bright red solution had formed. The solvent was removed in vacuo and the residue was extracted with minimum toluene, filtered through Celite and hexanes were added carefully (1:1). Overnight, red crystals of CoCl2[N(SiMe2CH2PPh2)2] (34) formed. Yield: 0.090 g (57%). Anal. Calcd. for C30H36CI2C0NP2S12: C, 54.71; H, 5.51; N, 2.13. Found: C, 54.82; H, 5.60; N, 2.20. lH NMR (C 6D 6): 5 12.1 (br, 12FL SiMe2), 9.9 (s, 4H, para-Ph), 7.2 (v br, 8H, ortho-Ph), 0.8 (br, 8H, meta-Ph), -58.4 (v br, 4H, CH2). UV-vis (C 7H 8): 326 (e = 2380 M-'cirr1), 488 (e = 1770 M" icnr 1) nm. MS: mle 657 (M+), 622 (M+ - CI). peff = 3.6 B.M. Reaction of 2 in neat benzyl chloride solvent proceeded more rapidly to give the identical product. (ii) Synthesis of CoBrCl[N(SiMe2CH2PPh2)2] (35) Method 1. To a blue 10 mL toluene solution of CoCl[N(SiMe2CH2PPh2)2] (2) (0.085 g, 0.14 mmol) was added excess (0.3 mL) neat benzyl bromide. Within 5 minutes the solution had become dark red. After being stirred overnight the solvent was removed in vacuo, the residue extracted with minimum toluene and filtered through Celite. Hexanes were added carefully (4 188 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides hexanes: 1 toluene) and left to stand. Overnight red crystals of CoBrCl[N(SiMe2CH2PPh2)2] (35) formed. To obtain X-ray quality crystals, hexanes were added carefully to a minimum of toluene containing 35/36 and allowed to stand, preferentially generating crystals of 36 • O.5C7H8 over two days. Yield: 0.075 g (78%). Anal. Calcd. for C3oH36BrClCoNP2Si2*C7H8: C, 51.25; H, 5.16; N, 1.99. Found: C, 51.29; H, 5.08; N, 2.08. *H NMR (C 6D 6): 8 11.1 (br, 12H, SiMe2), 10.0 (s, 4H, para-Ph), 8.5 (v br, 8H, ortho-Ph), 1.1 (br, 8H, meta-Ph), -58.2 (v br, 4H, CH2). UV-vis: 330 (e = 2900 M - W 1 ) , 496 (e = 1900 M^cnr1) nm. MS: mle 668 (M+ - CI), 622 (M+ - Br). p e f f = 3.0 B.M. Method 2. To a blue 10 mL toluene solution of CoBr[N(SiMe2CH2PPh2)2] (3) (0.085 g, 0.13 mmol) was added excess (0.3 mL) neat benzyl chloride. Overnight the solution became dark red and after being stirred for four days the solvent was removed in vacuo, the residue extracted with minimum toluene and filtered through Celite. Hexanes were added carefully (4 hexanes: 1 toluene) and left to stand. Overnight red crystals of CoBrCl[N(SiMe2CH2PPh2)2] (35) formed. (iii) Synthesis of CoBr 2[N(SiMe 2CH 2PPh 2) 2] (36) To a blue 10 mL toluene solution of CoCl[N(SiMe2CH2PPh2)2] (2) (0.085 g, 0.14 mmol) or CoBr[N(SiMe2CH2PPh2)2] (3) was added excess (0.3 mL) benzyl bromide neat. Within 5 minutes the solution had become dark red. After being stirred for four days (2) or overnight (3) the solvent was removed in vacuo, the residue extracted with minimum toluene and filtered through Celite. Hexanes were added carefully (4 hexanes: 1 toluene) and left to stand. Overnight red crystals of CoBr2[N(SiMe2CH2PPh2)2>0.5 C 7 H 8 (36) formed. Yield: 0.20 g (84%). Anal. Calcd. for C3oH36Br2CoNP2Si2*0.5C7H8: C, 50.70; H, 5.08; N, 1.77. Found: C, 51.45; H, 5.08; N, 1.77. *H NMR (C 6D 6): 5 11.1 (br, 12H, SiMe2), 10.0 (s, 4H, para-Ph), 8.5 (v br, 8H, ortho-Ph), 1.2 (br, 8H, meta-Ph), -58.0 (v br, 4H, CH2). UV-vis: 338 (e = 2900 M - W 1 ) , 432 (e = 1370 M-icm-1), 510 (e = 2070 M-'cnr1) nm. MS: mle IM (M+), 666 (M+-Br). p e f f = 3.3 B.M. Reaction in a very high concentration of benzyl bromide (1/3 of total solvent) gave rapid 189 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides formation of product as well as some zwitterionic, toluene insoluble purple precipitate identified asCoBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2CH2Ph)]. MS: mle 756/758 (M+-Br). (iv) Reaction of CoBr[N(SiMe2CH2PPh2)2] (3) with Benzyl Bromide in the Presence of TEMPO To a blue 10 mL toluene solution of CoBr[N(SiMe2CH2PPh2)2] (3) (0.085 g, 0.14 mmol) was added 10 mg TEMPO and then excess (0.3 mL) benzyl bromide neat. Within 5 minutes the solution had become dark red. After being stirred overnight the solvent was removed in vacuo, the residue extracted with minimum toluene, filtered through Celite and dried again. A NMR spectrum of the residue confirmed the presence of N-benzyloxy-2,2,6,6-tetramethylpiperidine (4.82 ppm, OC//2Ph); a MS was also obtained (mle 247 (M+)). No bibenzyl was observed. (v) Reaction of CoCl[N(SiMe2CH2PPh2)2] (2) with Mel To a blue 15 mL toluene solution of CoCl[N(SiMe2CH2PPh2)2] (2) (0.15 g, 0.24 mmol) was added excess (0.3 mL) Mel neat. No immediate colour change occurred but over an hour a purple precipitate crashed out and the solution colour slowly bleached. After being stirred overnight, the precipitate was collected, washed with toluene and hexanes and dried. Analysis identified the product as a mixture of predominantly CoIX[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (X=C1,1). Yield: 0.14 g (76% based on Mel addition only). Anal. Calcd. for C31H39C0INP2S12X (X = Cl, 80%; X = I, 20%): C, 47.53; H, 5.02; N, 1.79. Found: C, 47.57; H, 4.99; N, 1.76. MS: mle 729 (M+-C1), 637 (M+-I). 31p{lF£} NMR (CD2C12): 8 92.3 (X = Cl), 70.6 (X = I). (vi) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with Mel: Synthesis of CoI[N(SiMe2CH2PPh2)2] (4) To a 10 mL toluene solution of CoMe[N(SiMe2CH2PPh2)2] (6) (0.055 g, 0.090 mmol) was added 1.2 equivalents of Mel by quantitative vacuum transfer. Overnight the yellow 190 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides solution turned light green. After being stirred for one more day the toluene was removed in vacuo to give a green oil which was layered with hexanes to quickly yield green crystals of CoI[N(SiMe2CH2PPh2)2] (4). Yield: 0.060 g (95%). Anal. Calcd. for C3oH36CoINP2Si2: C, 50.43; H, 5.08; N, 1.96. Found: C, 50.87; H, 5.00; N, 1.75. lH NMR (C 6D 6): 5 15.5 (br, 8H), -5 (br, 4H). UV-vis: 538 (e = 227 M^cnr 1), 634 (e = 557 M - W 1 ) , 806 (e = 288 M^cnr1) nm. MS: mle 714 (M+), 587 (M+ -1), 510 (M+ -1 - C 6 H 5 ) . peff = 4.3 B.M. (vii) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with MeBr To a 10 mL toluene solution of CoMe[N(SiMe2CH2PPh2)2] (6) (0.097 g, 0.16 mmol) was added 1.2 equivalents of MeBr by vacuum transfer. Overnight the yellow solution turned light green. After being stirred for one more day the toluene was removed in vacuo to give a green oil which was extracted in hexanes to give a yellow/green solution. Yellow and green crystals formed overnight in the freezer. The yellow solid was identified as starting material by MS and the green solid as CoBr[N(SiMe2CH2PPh2)2] (3) by UV-vis, MS and lH NMR spectral comparisons with known material. (viii) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with Benzyl Chloride: Synthesis of CoCl2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (41) To a 10 mL toluene solution of CoMe[N(SiMe2CH2PPh2)2] (6) (0.13 g, 0.21 mmol) was added excess (0.3 mL) neat benzyl chloride. The solution immediately turned orange and then green over a few minutes. Overnight the green solution turned red-brown and a purple precipitate had formed. After being stirred for one more day the toluene was removed in vacuo to give a red oil and a purple solid. The red oil (major product, 72%) was identified as CoCl2[N(SiMe2CH2PPh2)2] (34) by UV-vis and lH NMR spectroscopy and the purple solid (minor product, 28 %) as CoCl2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (41). X-ray quality crystals of 41 were obtained by performing the reaction in an erlenmeyer flask in the absence of stirring. Purple crystals of 41 were deposited over three days. Yield: 0.042 g. Anal. Calcd for 191 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides C31H39CI2C0NP2S12: C, 55.28; H, 5.84; N, 2.08. Found: C, 55.65; H, 5.82; N, 1.98. lH NMR (CD2C12): 8 27 (br, 6H), 13.4 (s, 4H), 7.2 (s, 2H), 6.6 (s, 4H), 5.6 (v br, 6H), 1.8 (br, 4F£), -0.2 (br, 2or3H), -6.6 (s, 2H), -15.0 (br, 2or3H). 31p{lH) NMR (CD2C12): 870.3. UV-vis: 520 (e = 340M"1cm-1), 656 (e = 320 M^cnr 1), 734 (e = 290 M ' W 1 ) nm. MS: mle 637 (M+-C1) (ix) Reaction of CoMe[N(SiMe2CH2PPh2)2] (6) with Benzyl Bromide: Synthesis of CoBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (42) To a 10 mL toluene solution of CoMe[N(SiMe2CH2PPh2)2] (6) (0.13 g, 0.21 mmol) was added excess (0.3 mL) neat benzyl bromide. The solution immediately turned dark orange and then red over a few minutes. Overnight a purple precipitate formed. After being stirred for one more day the toluene was removed in vacuo to give a red oil and a purple solid. The red oil (major product, 80%) was identified as CoBr2[N(SiMe2CH2PPh2)2] (36) by UV-vis and lH NMR spectroscopy and the purple solid (minor product, 20%) as CoBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2Me)] (42). lH NMR (CD2CI2): 8 26.5 (br, 4H), 13.8 (s, 4H), 7.5 (d, 2H), 7.0 (t, 4H), 6.8 (d, 4H), 6.5 (br, 6H), 1.8 (br, 6H), -0.6 (br, 4H), -6.5 (s, 3H), -14.5 (br,4H). 31p{lF£} NMR (CD2C12): 886.4. UV-vis: 552 (e = 330 M^cnr 1), 658 (e = 310 M-'cm-1), 756 (e = 280 M^cm"1) nm. MS: mle 682 (M+-Br). (x) Reaction of Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] (10) with Benzyl Bromide: To a 10 mL toluene solution of Co(CH2SiMe3)[N(SiMe2CH2PPh2)2] (10) (0.075 g, 0.11 mmol) was added excess (0.3 mL) neat benzyl bromide. The solution immediately turned dark orange and then red over a few minutes. Overnight a small amount of purple precipitate had formed. After being stirred for one more day the toluene was removed in vacuo to give a red oil and a purple solid. The red oil (major product, >90%) was identified as CoBr2[N(SiMe2CF£2PPh2)2] (36) by UV-vis and ^ NMR spectroscopy and the purple solid (minor product, <10%) as CoBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2CH2SiMe3)] (40). MS: mle 740 (M+-CH3-Br). 192 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides (xi) Reaction of Co(CH2Ph)[N(SiMe2CH2PPh2)2] (8) with Benzyl Bromide: To a 10 mL toluene solution of Co(CH2Ph)[N(SiMe2CH2PPh2)2] (8) (0.075 g, 0.11 mmol) was added excess (0.3 mL) neat d7-benzyl bromide. The solution immediately turned green and then red over a few minutes. No purple precipitate formed. After being stirred for one more day the toluene was removed in vacuo to give a red oil, identified as CoBr2[N(SiMe2CH2PPh2)2] (36) by UV-vis and ! H NMR spectroscopy. (xii) Reaction of CoX 2[N(SiMe 2CH 2PPh 2) 2] (34, 36) with alkylating agents For each reaction approximately 0.02 g (0.03 mmol) of CoX2[N(SiMe2CH2PPh2)2] (34 or 36) was dissolved in 10 mL toluene. Three reactions, each with one equivalent of MeLi, LiCH2SiMe3 and CsFLjMgBr, at room temperature, all produced CoX[N(SiMe2CH2PPh2)2] (2 or 3), identified by its UV/vis spectrum. Addition of a further equiv produced a yellow colour, consistent with the formation of CoR[N(SiMe2C H2PPh2)2J. The target molecule Co(R)X[N(SiMe2CFf2PPh2)2] was not detected. In the case of LiCH2SiMe3, the reaction was carried out in an NMR tube (C7D8 and CsD^) and the peaks of 2, SiMe4 and Me3SiCH2CH2SiMe3 were observed at 8 0.0 and 0.1. Alkylations with one equivalent of MeLi or MeMgBr at low temperature (-78 °C) changed the deep red solution to a yellow-orange product, which could potentially be Co(Me)X[N(SiMe2CFf.2PPh2)2]. Upon being warmed to room temperature these solutions turned green and CoX[N(SiMe2CFJ.2PPh2)2] was identified. Addition of PEt3, CO and py at low or room temperature prior to alkylation did not alter the final results. 193 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides 4.8 References (1) Espenson, J. H. Prog. Inorg. Chem. 1983, 30, 189. (2) Espenson, J. H. Acc. Chem. Res. 1992,25, 222. (3) Halpern, J.; Maher, J. P. J. Am. Chem. Soc. 1964, 86,2311. (4) Halpern, J.; Maher, J. P. J. Am. Chem. Soc. 1965, 87, 5361. (5) Schneider, P. W.; Phelan, P. F.; Halpern, J. / . Am. Chem. Soc. 1969, 91, 77. (6) Fossey, J.; Lefort, D.; Sorba, J. Free Radicals in Organic Chemistry; John Wiley & Sons: New York, 1995. (7) Eberson, L. Electron Transfer Reactions in Organic Chemistry; Springer-Verlag: Berlin, 1987. (8) Electron Transfer in Inorganic, Organic, and Biological Systems; Bolton, J. R.; Mataga, N.; McLendon, G. L., Eds.; American Chemical Society: Washington, DC, 1991; Vol. 228. (9) Organometallic Radical Processes; Trogler, W. C , Ed.; Elsevier: New York, 1990. (10) Macromol. Chem., Macromol. Symp. 1987,10-11, 1. (11) Curran, D. P. Synthesis 1988, 417. (12) Matyjaszewski, K.; Patten, T. E.; Xia, J. f. Am. Chem. Soc. 1997,119, 61A. (13) Pattenden, G. Chem. Soc. Rev. 1988,17, 361. (14) Johnson, M. D. Acc. Chem. Res. 1983,16, 343. (15) Scheffold, R.; Rytz, G.; Walder, L. In Modern Synthetic Methods; R. Scheffold, Ed.; 1983; Vol. 3; pp 355. (16) Brown, T. M.; Cooksey, C. J.; Crich, D.; Dronsfield, A. T.; Ellis, R. / . Chem. Soc, Perkin Trans.1 1993, 2131. (17) Patel, V. F.; Pattenden, G.; Thompson, D. M. / . Chem. Soc, Perkin Trans.1 1990, 2729. (18) Branchaud, B. P.; Choi, Y. L. / . Org. Chem. 1988,53,4638. (19) Baldwin, J. E.; Li, C.-S. / . Chem. Soc, Chem. Commun. 1987, 166. 194 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides (20) B12; Dolphin, D., Ed.; John Wiley & Sons: New York, 1982; Vol. 1 and 2. (21) Toscano, P. J.; Marzilli, L. G. Prog. Inorg. Chem. 1984, 31, 105. (22) Gridnev, A. A.; Ittel, S. D.; Fryd, M.; Wayland, B. B. Organometallics 1993,12, 4871. (23) Dodd, D.; Johnson, M. D. / . Organomet. Chem. 1973,52, 1. (24) Witman, M. W.; Weber, J. H. Inorg. Chim. Acta 1977,23, 263. (25) Coleman, W. M.; Taylor, L. T. / . Am. Chem. Soc. 1971, 93, 5446. (26) Farmery, K.; Busch, D. H. Inorg. Chem. 1972,11, 2901. (27) Tyler, D. R. Prog. Inorg. Chem. 1988,36, 125. (28) Blaser, H.-U.; Halpern, J. / . Am. Chem. Soc. 1980,102, 1684. (29) Chock, P. B.; Halpern, J. / . Am. Chem. Soc. 1969, 91, 582. (30) Herberich, G. E.; Bauer, E.; Schwarzer, J. / . Organomet. Chem. 1969,17, 445. (31) Herberich, G. E.; Bauer, E. J. Organomet. Chem. 1969,16, 301. (32) Herberich, G. E.; Schwarzer, J. / . Organomet. Chem. 1972,34, C43. (33) Herberich, G. E.; Carstensen, T.; Klein, W.; Schmidt, M. U. Organometallics 1993,12, 1439. (34) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry; 5th ed.; John Wiley & Sons: New York, 1988. Ch. 18-F. (35) Kahn, O. Molecular Magnetism; VCH: New York, 1993. (36) Lamar, G. N.; Horrocks, W. D.; Holm, R. H. NMR of Paramagnetic Molecules; Academic Press: New York, 1973. (37) Collins, T. J.; Richmond, T. G.; Santarsiero, B. D.; Treco, B. G. R. T. / . Am. Chem. Soc. 1986,108, 2088. (38) Burdett, J. K.; Graham, M. A.; Perutz, R. N.; Poliakoff, M.; Rest, A. J.; Turner, J. J.; Turner, R. F. / . Am. Chem. Soc. 1975, 97,4805. (39) Hay, P. J. / . Am. Chem. Soc. 1978,100, 2411. (40) Jean, Y.; Eisenstein, O. Polyhedron 1988, 7,405. (41) Rachidi, I. E.-L; Eisenstein, O.; Jean, Y. New f. Chem. 1990,14, 671. 195 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides (42) Pviehl, J. F.; Jean, Y.; Eisenstein, O.; Pelissier, M. Organometallics 1992,11, 729. (43) Fryzuk, M. D.; MacNeil, P. A.; Massey, R. L.; Ball, R. G. / . Organomet. Chem. 1989, 365,231. (44) Bruckner, S.; Calligaris, M.; Nardin, G.; Randaccio, L. Inorg. Chim. Acta 1969,3, 308. (45) Costa, G.; Mestroni, G.; Stefani, L. / . Organomet. Chem. 1967, 7, 493. (46) Costa, G.; Mestroni, G.; Tauzher, G.; Stefani, L. / . Organomet. Chem. 1966, 6, 181. (47) Hitchcock, P. B.; McLaughlin, G. M. / . Chem. Soc. Dalton Trans. 1976, 1927. (48) Jaynes, B. S.; Ren, T.; Masschelien, A.; Lippard, S. J. / . Am. Chem. Soc. 1993,115, 5589. (49) Marzilli, L. G.; Summers, M. F.; Bresciani-Pahor, N.; Zangrando, E.; Charland, J.-P.; Randaccio, L. / . Am. Chem. Soc. 1985,107, 6880. (50) Summers, J. S.; Petersen, J. L.; Stolzenberg, A. M. / . Am. Chem. Soc. 1994,116, 7189. (51) Bailey, N. A.; McKenzie, E. D.; Worthington, J. M. / . Chem. Soc. Dalton Trans. 1977, 763. (52) Gerloch, M.; Higson, B. M.; McKenzie, E. D. / . Chem. Soc, Chem. Commun. 1971, 1149. (53) Konig, E.; Kremer, S.; Schnakig, R.; Kanellakopulos, B. Chem. Phys. 1978,34, 379. (54) McKenzie, E. D.; Worthington, J. M. Inorg. Chim. Acta 1976,16, 9. (55) van Enckevort, W. K. P.; Hendricks, H. M.; Beurskens, P. T. Cryst. Struct. Commun. 1977, 6,531. (56) McAuliffe, C. A.; Godfrey, S. M.; Mackie, A. G.; Pritchard, R. G. Angew. Chem., Int. Ed. Engl. 1992,31,919. (57) Godfrey, S. M.; McAuliffe, C. A.; Pritchard, R. G. / . Chem. Soc, Chem. Commun. 1994, 45. (58) Levason, W.; Ogden, J. S.; Spicer, M. D. Inorg. Chem. 1989,28,2128. (59) Levason, W.; Spicer, M. D. / . Chem. Soc. Dalton Trans. 1990, 719. (60) Brauer, D. J.; Kriiger, C ; Roberts, P. J.; Tsay, Y.-H. Chem. Ber. 1974, 707, 3706. 196 References begin on page 194 Chapter 4: Reactivity ofCobalt(II) Alkyl and Halide Complexes with Alkyl Halides (61) Ellison, J. J.; Power, P. P.; Shoner, S. C. / . Am. Chem. Soc. 1989, 111, 8044. (62) O'Connor, C. J. Prog. Inorg. Chem. 1982,29, 203. (63) Konig, E.; Schnakig, R.; Kanellakopulos, B. / . Chem. Phys. 1975, 62, 3907. (64) Birker, P. J. M. W. L.; Bour, J. J.; Steggerda, J. J. Inorg. Chem. 1973,12, 1254. (65) Birker, P. J. M. W. L.; de Boer, E. A.; Bour, J. J.V. Coord. Chem. 1973,3, 175. (66) Dale, B. W.; Williams, R. J. P.; Johnson, C. E.; Throp, T. L. / . Chem. Phys. 1968,49, 3441. (67) Boyd, P. D. W.; Buckingham, D. A.; McMeeking, R. F.; Mitra, S. Inorg. Chem. 1979,18, 3585. (68) Carlin, R. L. Magnetochemistry; Springer-Verlag: Heidelberg, 1986. (69) Makinen, M. W.; Kuo, L. C ; Yim, M. B.; Wells, G. B.; Fukuyama, J. M.; Kim, J. E. / . Am. Chem. Soc. 1985,107, 5245. (70) Makinen, M. W.; Yim, M. B. Proc. Natl. Acad. Sci. USA 1981, 78, 6221. (71) Carlin, R. L. Science 1985,227, 1291. (72) Jensen, K. A.; Jorgensen, C. K. Acta Chem. Scand. 1965,19, 451. (73) Jenson, K. A.; Nygaard, B.; Pedersen, C. T. Acta Chem. Scand. 1963,17, 1126. (74) Huber, T. A.; Macartney, D. H.; Baird, M. C. Organometallics 1993,12, 4715. (75) Huber, T. A.; Macartney, D. H.; Baird, M. C. Organometallics 1995,14, 592. (76) Lee, K.-W.; Brown, T. L. J. Am. Chem. Soc. 1987,109, 3269. (77) Scott, S. L.; Espenson, J. H.; Zhu, Z. / . Am. Chem. Soc. 1993,115,1789. (78) It must be assumed on the basis of precedent that in fact the generated alkyl radicals do combine with the cobalt(JJ) halide complex present, but the resulting alkyl halide is homolytically unstable, releasing the alkyl radical back to the system. (79) Bresciani-Pahor, N.; Forcolin, M.; Marzilli, L. G.; Randaccio, L.; Summers, M. F.; Toscano, P. J. Coord. Chem. Rev. 1985, 63, 1. (80) Costa, G. Pure Appl. Chem. 1972,30, 335. (81) Ng, F. T. T.; Rempel, G. L. J. Am. Chem. Soc. 1982,104, 621. 197 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides (82) Tsou, T.-T.; Loots, M.; Halpern, J. / . Am. Chem. Soc. 1982,104, 623. (83) Klein, H.-F.; Karsch, H. H. Chem. Ber. 1975,108, 956. (84) Ikariya, T.; Yamamoto, A. / . Organomet. Chem. 1976,116, 239. (85) Yamazaki, H.; Hagihara, N. / . Organomet. Chem. 1970,21, 431. (86) Keogh, D. W.; Poli, R. / . Am. Chem. Soc. 1997,119, 2516. (87) Poli, R. Chem. Rev. 1996, 96, 2135. (88) Fettinger, J. C ; Mattamana, S. P.; Poli, R.; Rogers, R. D. Organometallics 1996,15, 4211. (89) Thorstenson, T.; Songstad, J. Acta Chem. Scand. 1976, 30, 724. (90) Fryzuk, M. D.; Carter, A.; Rettig, S. J. Organometallics 1992,11, 469. (91) Cotton, F. A.; Faut, O. D.; Goodgame, D. M. L.; Holm, R. H. / . Am. Chem. Soc. 1961, 83, 1780. (92) Horrocks, W. D.; Hecke, G. R. V.; Hall, D. D. Inorg. Chem. 1967, 6, 694. (93) Carlin, R. L.; Chirico, R. D.; Sinn, E.; Mennenga, G.; de Jongh, L. J. Inorg. Chem. 1982, 21, 2218. (94) Ganguli, P.; Marathe, V. R.; Mitra, S. Inorg. Chem. 1975,14, 970. (95) Banci, L.; Bencini, A.; Benelli, C ; Gatteschi, D.; Zanchini, C. Struct. Bonding (Berlin) 1982,52, 37. (96) Nakazawa, H.; Yamaguchi, Y.; Mizuta, T.; Ichimura, S.; Miyoshi, K. Organometallics 1995,14, 4635. (97) Covert, K. J.; Askew, E. F.; Grunkemeier, J.; Koenig, T.; Tyler, D. R. J. Am. Chem. Soc. 1992,114, 10446. (98) Koenig, T. W.; Hay, B. P.; Finke, R. G. Polyhedron 1988, 7, 1499. (99) Thomas, M. J.; Shackleton, T. A.; Wright, S. C ; Gillis, D. J.; Colpa, J. P.; Baird, M. C. / . Chem. Soc, Chem. Commun. 1986, 312. (100) Sweany, R. L.; Halpern, J. / . Am. Chem. Soc. 1977,99, 8335. (101) Jacobsen, E. N.; Bergman, R. G. / . Am. Chem. Soc. 1985,107, 2023. 198 References begin on page 194 Chapter 4: Reactivity of Cobalt(II) Alkyl and Halide Complexes with Alkyl Halides (102) Bentrude, W. G. In The Chemistry of Organophosphorus Compounds; F. R. Hartley, Ed.; Wiley: Sussex, 1990; Vol. 1; pp 531. (103) Dockery, K. P.; Bentrude, W. G. / . Am. Chem. Soc. 1997,119, 1388. (104) Goldman, A. S.; Tyler, D. R. / . Am. Chem. Soc. 1986,108, 89. (105) Kruppa, A. I.; Taraban, M. B.; Leshina, T. V.; Natarajan, E.; Grissom, C. B. Inorg. Chem. 1997, 36, 758. (106) Takai, K.; Nitta, K.; Fujimura, O.; Utimoto, K. / . Org. Chem. 1989,54, 4732. (107) Haynes, J. S.; Oliver, K. W.; Rettig, S. J.; Thompson, R. C.; Trotter, J. Can. J. Chem. 1984, 62, 891. 199 References begin on page 194 Chapter F ive Iron(II) and Iron(III) Halide and Alkyl Complexes 5.1 Introduction Previous chapters have explored the rich chemistry of the amidodiphosphine ligand "N(SiMe2CH2PPh2)2 coordinated to chromium and cobalt centres and an extension of this ligand system to iron(II) seemed warranted. A comparison of the reactivity of iron(II) and iron(III) amidodiphosphine complexes with the analogous chromium and cobalt systems was expected to be easily accessible. Furthermore, paramagnetic organometallic iron(II) chemistry has not been well developed in its own right.1,2 Iron(II), d 6 and iron(III), d 5, also show a wide range of magnetic behaviour.3 Hence, an exploration of the coordination chemistry of iron(II) and iron(III) with -N(SiMe2CH2PPh2)2 was initiated. 200 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes 5.2 Synthesis and Reactivity of Iron(II) Halide Complexes (i) Synthesis and characterization of FeX[N(SiMe 2CH 2PPh 2)2] (X = CI, Br) complexes Dropwise addition of the lithium salt LiN(SiMe 2CH2PPh 2) 2 to a slurry of F e X 2 ( X = C l , Br ) in T H F (equation 5.1), fil tration and recrystal l ization from co ld toluene gives F e X [ N ( S i M e 2 C H 2 P P h 2 ) 2 ] ( X = C l (43), B r (44)) as slightly beige powders in high yield. The complexes have a solution magnetic moment 4 ' 5 of 4.9 B . M . , consistent with a high-spin d 6 system (four unpaired electrons). 6 FeX[N(SiMe2CH2PPh2)2] is ESR-silent but broad peaks are visible in the * H N M R spectrum. Unfortunately, it is difficult to assign the spectrum due to extreme broadness (leading to unreliable integration) and missing peaks. A s before, knowledge of the spectrum, however, is useful in compound identification. THF, rt M 6 2 S i \ F e X 2 + L iN(S iMe 2 CH 2 PPh 2)2 • :N-M e 2 S i P h 2 -P - F e — X [5.1] X = Cl 43 X = Br 44 -P P h 2 Despite extensive efforts, X-ray quality crystals could not be obtained. However, based on spectroscopic evidence halides 43 and 44 are l ike ly mononuclear and tetrahedral. The measurement of a high spin d 6 system rules out the possibility of square planar coordination as square planar iron(n) complexes such as Fe(C6Cl5)2(PEt2Ph)2 are generally low spin. 7 A dimeric species can be excluded as it would be expected to show lower solubility in toluene. The UV-vis spectrum of 43 is also consistent with a tetrahedral geometry around iron. For the tetrahedral 201 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes [FeX4] 2 _ ( X = CI, Br, I, N C S ) complexes, absorption bands in the IR region at around 4000 c m - 1 due to the expected d-d transition were observed. 8 These compounds are typically very pale beige in colour. Other tetrahedral iron(II) phosphine halide complexes include FeCl2(dippe) , 9 FeCi2(PEt n Ph3-n)2, 1 ( ) FeCl2(depe) n and FeX2(dppe) ; 1 2 ' 1 3 only the first example is structurally characterized. A tetrahedral, structurally characterized Fe(U) amide, FeCl [NR 'Ar f ] (TMEDA) (R' = C(CD3)2CH3, C(CD3)2Ph; Ar f = 2,5-C6H3FMe) has also been reported. 1 4 A l l of these examples are high-spin Fe(II) compounds, with magnetic moments ranging from 4.8 to 5.3 B . M . (ii) Reaction of FeCl[N(SiMe 2 CH 2 PPh 2 ) 2 ] (43) with donor ligands Addition of weak donor solvents such as acetonitrile or T H F to iron chloride 43 caused no change in the UV-vis spectrum. Addition of an atmosphere of C O to a toluene solution of 43 also yielded no spectral change. Only addition of pyridine, a strong o-donor, had any effect at all . This is in stark contrast to the reaction of the chromium chloride dimer 1 with donor ligands (section 2.2); it reacts to form five-coordinate adducts even with weak donors. The tetrahedral cobalt chloride complex 2 was found to be unreactive to pyridine but bound C O without a change in spin-state (section 2.3). The ligation of pyridine to 43 can be monitored by UV-vis spectroscopy due to the growth of a band at 370 nm (Figure 5.1) upon addition of pyridine. Further addition caused a shift of this band to 388 nm; hence a two step equilibrium involving a mono-pyridine adduct and a bis-pyridine derivative is proposed (equations 5.2 and 5.3). Ki FeCI[N(SiMe 2 CH 2 PPh 2 ) 2 ] + py — FeCI(py)[N(SiMe 2CH 2PPh 2) 2] [5.2] K 2 FeCI(py)[N(SiMe 2CH 2PPh 2) 2] + py _ ^ FeCI(py) 2 [N(SiMe 2 CH 2 PPh 2 ) 2 ] [5.3] 202 References begin on page 246 Chapter 5: Fe(II) andFe(III) Halide and Alkyl Complexes These pyridine complexes are deep orange, but could not be isolated for elemental analysis, even from neat pyridine. However, careful titration with stoichiometric and then excess pyridine yielded crude equilibrium constants for ligation of one and then two pyridines to the tetrahedral iron centre. K i and K2 may be estimated by plots of log [FeCl(py)[N(SiMe2CH2PPh2)2]] / [FeCl[N(SiMe2CH2PPh2)2]] versus log [pyridine] and log [FeCl(py)2[N(SiMe2CH2PPh2)2]] / [FeCl(py)[N(SiMe2CH2PPh2)2]] versus log [pyridine] respectively (equations 5.4 and 5.5). The endpoint of the first equilibrium was assumed to be at the point where all of 43 had reacted to form 43»py and no 43*py2 had yet formed. On Figure 5.1 this point was assumed to be at the loss of the isosbestic point as more pyridine was added. As increasing amounts of pyridine were added the 370 nm (Figure 5.1) charge transfer band shifted to 388 nm (Figure 5.2). The endpoint to the second equilibrium was assumed to be at the point where no increase in absorbance to the 388 nm band was observed (Figure 5.2, last trace) and that only 43»py2 existed in the solution. Preliminary analysis of the above data gave order of magnitude estimates of K i = 2 x 104 and K 2 = 6. Although it is tempting to definitively assign the mono and bis-pyridine adducts as five and six-coordinate species respectively, there is little evidence to support this view. For example, the 203 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes U ' W I • • • • 1 • • • • 1 • ' • • 1 300 400 500 600 Wavelength (nm) Figure 5.1 Spectrophotometric data for the first titration of FeCl[N(SiMe2CH2PPh2)2] (43) with pyridine. Data are obtained at 25 °C. Spectrum #7 from the bottom, where the isosbestic point is broken, is assumed to be the first endpoint. 1.86 1.51 g 1.17 rt JO | 0.82 0.47 H 0.12 300 600 400 500 Wavelength (nm) Figure 5.2 Spectrophotometric data for the first titration of FeClpy[N(SiMe2CF£2PPh2)2] (43»py) with pyridine. Data are obtained at 25 °C. The endpoint is at the top spectrum, where [py] = 1.64 M. 204 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes lH NMR spectrum of chloride 43 in ds-pyridine shows no substantial change from that observed in d6-benzene. If the geometry was changing, pyridine coordination would be expected to cause a change in the spectrum due to the expected change in magnetic moment (octahedral Fe(II) generally has a higher jj^ ff than tetrahedral Fe(II)) but this is not observed. On the other hand, the 3 1P{ lK} NMR spectrum of chloride 43 in ds-pyridine is barren of peaks, implying that the phosphines are still bound to the still-paramagnetic Fe(II) centre. A fast equilibrium wherein the phosphines and pyridines are competing for sites around a tetrahedral Fe(II) could be invoked to rationalize these results. 5.3 Synthesis and Structure of a Bis-Ligand Iron(II) Complex The ambiguity of the coordination number in the pyridine-ligated complexes spurred an interest in further examination of the coordination chemistry of chloride 43. Use of a second equivalent of LiN(SiMe2CFf2PPh2)2 would give the metal centre a choice of octahedral or tetrahedral coordination. In addition, formation of hexanes-soluble bis-ligand complexes of the formula M[N(SiMe2CH2PPh2)2]2 often were suggested to lower the yield of MX[N(SiMe2CH2PPh2)2] so there was an interest in characterizing an example of this type of complex. Accordingly, one equivalent of LiN(SiMe2CH2PPh2)2 was added dropwise to a toluene solution of FeCl[N(SiMe2CF£2PPh2)2] (43) which yielded, after overnight stirring and workup in hexanes, pale green crystals of Fe[N(SiMe2CH2PPh2)2]2 (45), a bis-ligand complex. The same complex could be prepared directly from FeCl2 by addition of two equivalents of ligand (equation 5.6). 205 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes X = 1 FeCI2 + X LiN(SiMe 2CH 2PPh 2) 2 • FeCI[N(SiMe2CH2PPh2)2] 43 [5.6] X = 2 \ ^ ^ LiN(SiMe 2CH 2PPh 2) 2 PPh 2 Ph 2 ' * P : N — F e - « i i N : M e 2 S i ^ | ^SiMa, P' Ph 2 45 A solution magnetic moment of 4.7 B.M. was measured by Evans' method; this value is consistent with a high-spin d 6 tetrahedral system. The nearly colourless crystals are, as previously described, consistent with a tetrahedral geometry; the very pale green colour could be due to very weak spin-forbidden transitions, similar to those observed for M n 2 + systems.153 The mass spectrum shows a molecular ion peak at 1112 mle, loss of a CH2PPh2 arm at 913 mle and also the base peak M + - [N(SiMe2CH2PPh2)2] at 584 mle. There are no signals assignable to halides in the spectrum, consistent with bis-ligand complexation. The A H NMR spectrum for bis-ligand complex 45 contains four broad peaks which can be tentatively assigned using integration and relative broadness to ligand backbone peaks; the backbone methylene CH2 peaks are too broad to be observed. The unique SiMe2 resonance appears at +20.5 ppm, the meta-proton resonance at +10.8 ppm and the ortho and para resonances at +1.0 and -1.0 ppm respectively. The observation of only one set of resonances for the ligand implies a fluxional process in action that makes the two sides of the chelate arm equivalent. The 3 1P{ lH} NMR spectrum of 45 is devoid of peaks, a fact that can also be attributed to a fluxional process involving dissociation and re-association of the chelating phosphines. Note that of the iron(II) complexes reported in this thesis, this is the only 206 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes one with a relatively clean N M R spectrum; other complexes have considerably broader peaks, and many peaks are missing altogether. Most iron(II) compounds are octahedral, although there are a few cases of four-coordinate Fe(II) derivatives, as already ment ioned. 1 5 b ' 1 6 The magnetic and U V - v i s spectral data suggest that the bis-ligand complex 45 is tetrahedral and not octahedral. Structurally characterized examples of four-coordinate, tetrahedral iron(II) complexes include FeCl2(dippe), 9 Fe[S(NMe2)2]2(S-2,4,6-* ' P r 3 - C 6 H 2 ) 2 , 1 7 [ F e ( S e P h ) 4 ] 2 - , 1 8 F e ( S e - 2 , 6 - ' * P r 2 - C 6 H 3 ) 2 ( P M e 2 P h ) 2 1 9 and F e C l ( N R ' A r f ) ( T M E D A ) (R' = C ( C D 3 ) 2 C H 3 , C ( C D 3 ) 2 P h ; Ar f = 2,5 -C6H3FMe). 1 4 There are some tetrahedral iron(II) alkyls and aryls as well; these wi l l be discussed later. X-ray quality crystals of 45 were prepared by slow evaporation of a hexanes solution to near dryness. The crystal structure, shown in Figure 5.3, confirms the tetrahedral coordination around the iron centre, with each potentially tridentate ligand bound through the central amide and one phosphine, that is, in a bidentate fashion. Selected bond lengths and angles are found in Table 5.1. The complex probably remains tetrahedral due to the steric congestion around the metal; octahedral coordination is unlikely. The distortions in the tetrahedral array manifest themselves in the coordination sphere bond angles, which range from 90.97(7)° (P( l ) -Fe-N(l ) ) to 142.50(9)° (N(l)-Fe-N(2)). As in other metal phosphine complexes, iron-phosphine bonds are very sensitive to both coordination geometry and to spin state. Typical octahedral low-spin Fe(II)-P bonds range from 2.24-2.30 A while octahedral high-spin complexes have Fe-P bonds which are much longer, typically from 2.58-2.68 A . 2 0 These fluctuations reflect the difference in size between high and low-spin metal centres and the fact that greater electronic unsaturation at the metal centre reduces metal-phosphine back-bonding strength (and hence weakens and lengthens the bond). The Fe-P bond lengths of 2.4580(7) and 2.4696(7) A in the bis-ligand complex 45 are intermediate between these two extremes. These Fe-P bond lengths compare well with other tetrahedral iron(II)-P bond 207 References begin on page 246 Chapter 5: Fe(H) and Fe(III) Halide and ALkyl Complexes Table 5.1 Selected bond lengths and angles in Fe[N(SiMe2CH2PPh2)2]2 (45). Atom Atom Distance (A) Atom Atom Distance (A) Fe P(l) 2.4580(7) Fe N(l) 1.981(2) Fe P(2) 2.4696(7) Fe N(2) 1.987(2) N(l) Si(l) 1.725(2) N(l) Si(2) 1.721(2) N(2) Si(3) 1.717(2) N(2) Si(4) 1.722(2) Atom Atom Atom Angle (") Atom Atom Atom Angle (°) N(l) Fe N(2) 142.50(9) N(l) Fe(l) P(D 90.97(7) N(2) Fc P(l) 114.50(6) N(l) Fe P(3) 109.05(6) N(2) Fe P(3) 94.61(6) P(l) Fe P(3) 98.33(3) Figure 5.3 Molecular structure (Chem 3D®) and numbering scheme for Fe[N(SiMe2CH2PPh2)2]2 (45). Phenyl substituents on phosphorus have been removed for clarity. 208 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes lengths such as 2.426(1) A in Fe(Se-2,6-'Pr2-C6H3)2(PMe2Ph)2,19 2.408(2) and 2.447(2) A in Fe(Se-2,6-**Pr2-C6H3)2(depe)19 and 2.47(1) A in FeCl2(dippe).9 There are a number of iron(U) amide complexes with low coordination numbers that have been structurally characterized. Some examples include Fe[N(SiMePh2)2]2»2 1 Fe[N(SiMe3)2]2(THF)22 and FeCl(TMEDA)[NR'Arf] (R' = C(CD3)2CH3, C(CD3)2Ph; Ar f = 2,5-C6H 3 FMe), 1 4 with Fe-N bond lengths of 1.917(2) and 1.927(3), 1.916(5) and 1.918(3) A respectively. The corresponding Fe-N distances in the bis-ligand complex 45 of 1.98(2) and 1.987(2) A are slightly longer than those observed in other iron(II) amides. This elongation is likely due to an attempt to relieve steric crowding. The N-Si bond lengths in 45 range from 1.712(2) to 1.725(2) A and are comparable to those observed in the iron(II) amides in Fe[N(SiMe3)2]2(THF)22 (1.709(5) A) and Fe[N(SiMe3)2]2 (1.723(5) A, gas phase electron diffraction).23 We can now address the question of whether iron(II) can support an octahedral ligand array incorporating one or two ancillary ligands. In the case of two amidodiphosphine ligands, the answer is a definitive no; the X-ray crystal structure clearly shows a tetrahedral array despite a potentially octahedral coordination environment being available. This coordinative unsaturation is probably enforced due to the very high steric requirements of octahedral coordination with two ancillary ligands. What about the original question of the pyridine adduct coordination number, where only one amidodiphosphine is present? It is possible to invoke steric crowding in this case as well, implying a tetrahedral geometry with dangling phosphines. Although one would expect such phosphine arms to be visible in the 3 1 P NMR spectrum, the 3 1P{ lH] NMR spectrum of 45 is devoid of resonances, despite the crystallographic observation of dangling phosphines. A fast fluxional process in which different phosphine arms replace each other at the metal centre can be invoked to make the phosphines equivalent and hence unobservable by virtue of being bound to a paramagnetic Fe(II) centre. In the case of the pyridine adduct, recall that the 3 1 P{ 1 H}NMR 209 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes spectrum of 43 in ds-pyridine is devoid of resonances. Hence, a similar process in which the phosphine arms are competing with pyridine ligands instead of each other is certainly possible. As a result, based on the available data, it is more likely that the pyridine adducts of chloride 43 are in fact tetrahedral and undergoing fast phosphine/pyridine exchange than the alternative structure in which all ligands are bound to iron(II) to yield an octahedral structure. 5.4 Synthesis a n d C h a r a c t e r i z a t i o n of Iron(II) A l k y i C o m p l e x e s (i) Attempted synthesis of iron methyl and benzyl complexes In similar fashion to the alkylations of the chromium and cobalt halide starting materials, the synthesis of the corresponding iron(II) methyl complex FeMe[N(SiMe2CH2PPh2)2l was attempted by addition of MeLi to a toluene or THF solution of chloride 43 (equation 5.7). Dropwise addition at -78 °C resulted in a golden brown solution, which turned dark brown upon being warmed to room temperature. Alternately, addition of MeMgBr at -78 °C to 43 resulted in a bright yellow solution, which again changed to dark brown upon being warmed to room temperature. In both cases removal of the solvent, extraction with hexanes and filtration through Celite gave an oily brown solid which could not be identified. The mass spectrum did not show any recognizable peaks and elemental analysis of multiple attempts at this reaction gave irreproducible and unassignable results. Hence it appears that, unlike the chromium and cobalt cases, FeMe[N(SiMe2CH2PPh2)2l is unstable and decomposes to intractable products. The mechanism of decomposition is unclear. An attempt to form a benzyl complex by addition of benzylpotassium in THF at -78 °C to halide 43 gave a bright yellow solution. The appearance of this colour was encouraging because it is consistent with many other iron(II) alkyl complexes. For example, Fe(CH2Ph)2(dippe),24 FeR.2 (R = C(SiMe3)2C5H4N-2),25-26 FePh2(PEt3)2 2 7 and Fe(Mes*)228'29 are all yellow in colour. The 210 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes yellow colour persisted for a few minutes at room temperature but unfortunately over the course of the workup procedure decomposition to brown intractable materials occurred. P h 2 •P M e2 S i C I -78 °C 25 °C \ . ' r i ^ Yellow/brown Decomp. [5.7] N — F e — C I • . solution M e 2 S i ^ I MeLi or K C H 2 P h P h 2 43 It was considered that the decompositon of the alkyl compounds may be attributed to the coordinative unsaturation of the iron centre. The idea of adding strong donor ligands as base-stabilizing agents was examined. Addition of MeLi or MeMgBr to a pyridine/THF solution of chloride 43 resulted in a promising dark red solution. Unfortunately, decomposition of this red species to a brown material occurred over the course of a day at room temperature. The use of a chelating ligand such as bipyridine as a stabilizing feature could be considered in future. Instead of base-stabilization of low-coordinate alkyls, reaction of the transient Fe-alkyl with CO to form potentially more stable Fe(II)-acyl complexes was examined.30 Hence the addition of MeMgBr or KCH2Ph at -78 "C to chloride 43 gave a yellow solution, as described above, and to this was added either an excess or a stoichiometric amount of CO. Depending on the conditions, colour changes from the yellow putative FeR[N(SiMe2CH2PPh2)2] to red and green compounds occurred. Solution IR spectra of the products showed multiple CO bands from 1818 cnr 1 to 2046 cm - 1, the number and type depending on the reaction conditions. This data implies combinations of acyl products (CO inserted into the iron-alkyl bond) and iron-carbonyl adducts. The identity of these species and their separation and characterization was not resolved. Over time some decomposition was also observed. 211 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes (ii) Synthesis and structure of Fe(ri5-C5H5)rN(SiMe 2 CH 2 PPh 2 )2] (46) Dropwise addition of a solution of NaCp»DME to a THF solution of iron chloride 43 at -78 °C caused a colour change to dark red, and from the resulting solution red crystals of Fe(rj5-C5H5)[N(SiMe2CH2PPh2)2] (46) were isolated and characterized (equation 5.8). This complex is analogous to those prepared with chromium (11) and cobalt (12) metal centres. [5.8] Unlike the inaccessible 14-electron iron(II) methyl and benzyl complexes the iron(II) cyclopentadienyl complex 46 is an eighteen-electron system and, perhaps unique to this thesis, is hence a diamagnetic compound. Of course, therein lies the reason for the stability of 46; it is essentially a ferrocene analogue. A d-d transition is observed in the UV-vis spectrum of 46 at 494 nm; in ferrocene the comparable transition is found at 440 nm.31 The X-ray crystal structure (Figure 5.4) of Fe(Ti5-C5H5)[N(SiMe2CH2PPh2)2] (46) revealed the expected monomeric distorted pseudo-octahedral geometry. This type of complex is commonly referred to as a "half-sandwich" complex and there are many examples in the literature.32 Complex 46 is also isostructural with the chromium analogue 11 (section 2.3 (vi)). The bond angles (Table 5.2) around the iron centre of 123.2°, 124.1° and 124.0° (P(l)-Fe-Cp, 212 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes P(2)-Fe-Cp and N-Fe-Cp) are comparable to those found in the chromium analogue. The Fe-P bond lengths (Table 5.2) of 2.256(1) and 2.242(1) A are comparable to other low-spin octahedral iron(II) complexes (2.24-2.30 A ) 2 0 but are much shorter than those observed for the bis-ligand complex 45 which contains a high-spin tetrahedral iron(II) centre. Also, the bond lengths are substantially shorter than those found in the isostructural chromium complex 11, which has Cr-P bond lengths of 2.353(3) and 2.366(3) A. This shortening is essentially due to the smaller iron(II) centre (no unpaired electrons) vs. the chromium(II) centre, which in addition to being further left in the periodic series (and therefore larger), still contains two unpaired electrons. The crystal radii for octahedral low-spin Fe(II) and Cr(II) centres are 0.75 and 0.87 A respectively,33,34 which almost completely account for the differences observed. Similarly, the Fe-Cp(centroid) distance of 1.71 A is shorter than the 1.86 A observed in the chromium complex. The Fe-N distance of 2.086(3) A is comparable to the chromium complex (2.066(7) A) but longer than in the tetrahedral high-spin iron complex 45 by approximately 0.1 A. Also, the N-Si bond lengths of 1.689(3) and 1.687(3) A in diamagnetic 46 are considerably shorter than in high-spin 45 (1.720 A average). These observations could reflect the minimal interaction of the amide lone-pair with the metal centre in the eighteen electron complex 46 (inclusion of the amide-lone pair would make this a formally 20 electron, or more appropriately an (18 + 8) complex35 compared to the fourteen electron (unsaturated) high-spin 45. (iii) Synthesis and structure of FeR*[N(SiMe 2CH 2PPh 2)2] (R* = bulky alkyl) The observation that the benzyl complex took longer to decompose than the methyl complex implied that perhaps steric bulk could be beneficial in the synthesis of iron(II) alkyl derivatives. This concept is certainly not new; bulky ligands have often been used to shield otherwise reactive metal centres from external attack.32 In extreme cases very low coordination 213 References begin on page 246 Table 5.2 Chapter 5: Fe(II) and Fe(lll) Halide and ALkyl Complexes Selected bond lengths and angles in Fe(ri5-C5H5)[N(SiMe2CH2PPh2)2] (46). Atom Atom Distance (A) Atom Atom Distance (A) Fe P(l) 2.256(1) Fe N 2.086(3) Fc P(2) 2.366(3) Fe Cp 1.71 N SKD 1.689(3) N Si(2) 1.687(3) Atom Atom Atom Angle (°) Atom Atom Atom Angle (") PCD Fc P(2) 100.14(4) P(l) Fe N 81.73(8) PCI) Fc Cp 125.2 P(2) Fe N 89.94(7) P(2) Fc Cp 124.1 N Fe Cp 124.0 Si(l) N Si(2) 135.2(2) Fe P(l) C(l) 104.5(1) C28 C29 ft C30 J5) OS C35 Figure 5.4 Molecular structure (ORTEP) and numbering scheme for Fe(Ti5-C5H5)[N(SiMe2CH2PPh2)2] (46). 214 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes numbers can be enforced, for example, two coordinate iron(II) "super-mesityl" complexes were successfully prepared and structurally characterized.28,29 Tetrahedral iron(II) bis-alkyls were prepared by Hermes and Girolami where the alkyls were all bulky groups-such-as neosilyl, neopentyl and 2-methyl-2-phenylpropyl (neophyl), although, in this case, a bis-benzyl complex was prepared and structurally characterized.24 In keeping with this concept, the reaction of bulky alkylhthium reagents with chloride 43 was attempted (equation 5.9). Addition of LiCH2SiMe3 dropwise at -78 °C to a THF solution of 43 gave a bright yellow solution over 10 minutes which remained yellow after standard workup. Evaporation of a yellow hexanes solution gave a waxy yellow solid which gave an elemental analysis consistent with Fe(CH2SiMe3)[N(SiMe2CH2PPh2)2] (47). The mass spectrum showed a weak M +-Me peak at 663 mle and a strong M+-CH2SiMe3 peak at 584 mle. The solution magnetic moment (Evans' method)4,5 of 4.8 B.M. is consistent with a high-spin tetrahedral iron(II) complex. M e 2 S i M e 2 S \ :N-P h 2 -P • F e — C I + LiFT -P P h 2 THF M 6 2 S i \ :N-" 7 8 ° c > - 0 . / Me 2 Si P h 2 •P • F e — R " [5.9] 43 R* = C H 2 S i M e 3 47 R* = CH(S iMe 3 ) 2 48 •P P h 2 Similarly, reaction of an extremely bulky alkyl, namely LiCH(SiMe3)2, with chloride 43 produced after workup yellow crystals (equation 5.9) of Fe{CH(SiMe3)2}[N(SiMe2CH2PPh2)2] (48). The elemental analysis was consistent with the formulation shown and the solution magnetic 215 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes moment of 5.3 B.M. confirmed the presence of a high-spin iron(II) tetrahedral complex (four unpaired electrons with a second order orbital contribution). In this complex, there was a possibility of an agostic C-H interaction with the metal centre but the IR spectrum showed no evidence of this, that is, no C-H stretches below 2850 cm - 1 were observed.36 The mass spectrum of the bis(trimethylsilyl)methyl complex 48 was very informative, with many identifiable fragments evident (Figure 5.5). The molecular ion peak at 743 mle easily lost a proton (likely the alkyl C-H) to yield a cluster at 742 mle (M+-H) which overlapped with the molecular ion cluster. Also easily observable was the loss of a methyl group at 728 mle (M+-Me), loss of a trimethylsilyl group at 670 mle (M+-SiMe3) and loss of the entire alkyl group at 584 mle (M+-CH(SiMe3)2). too-90-i 80 '-_ 70 -i 60 -j so -i 40: 30-; 20 ^  10-i 0J 550 584 670 619 673 728 / 743 600 650 x40 700 750 Figure 5.5 Mass spectrum of Fe{CH(SiMe3)2}[N(SiMe2CH2PPh2)2] (48). The *H NMR spectra of these iron(II) alkyls are difficult to assign due to extreme broadness and missing peaks. The spectrum of the neosilyl complex 47 has two observable peaks at 11.6 and 0.0 ppm in a 2:1 integration, while the bis(trimethylsilyl)methyl complex 48 shows three peaks at 18.0, 13.4 and -4.2 ppm. Obviously there are many peaks missing from these 216 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes spectra, precluding any assignment. However, these spectra can be used as fingerprints for the alkyl complexes. The neosilyl complex 47, while isolable for elemental analysis purposes, did not produce suitable crystals for X-ray diffraction, and moreover, it was found that complex 47 slowly decomposed over one week to the ubiquitous brown material if left at room temperature. Fortunately, however, the very bulky bis(trimethylsilyl)methyl complex 48 is not only stable over months at room temperature but also grows immense crystals. Slow evaporation of a hexanes solution of 48 to almost complete dryness deposited crystals suitable for X-ray diffraction. The structure is shown in Figure 5.6 and confirms the tetrahedral geometry around the iron(II) centre and also the presence of the Fe-C bond. Selected bond lengths and angles are shown in Table 5.3. There is some distortion from pure tetrahedral; the coordination sphere bond angles (ideally 109.5°) range from 87.03(5)°• (N-Fe-P(2)) to 131.20(7)° (N-Fe-C(31)). This distortion is less than in the sterically crowded bis-ligand complex 45. The Fe-P bond lengths in alkyl 48 of 2.4886(6) and 2.4924(6) A are very similar to those observed in the bis-ligand complex 45 and are similar to those observed in other high-spin tetrahedral iron(II) complexes (as described in section 5.3). Similarly, the Fe-N bond length of 2.002(2) A and the N-Si bond lengths of 1.710(2) and 1.713(2) A are comparable to those observed in complex 45. Compound 48 can be compared with other structurally characterized tetrahedral iron(II) alkyl complexes. A series of alkyls stabilized by l,2-bis(diisopropylphosphino)ethane (dippe) has been prepared and the p-methylbenzyl complex Fe(CH2C6H4Me)2(dippe) structurally characterized; the Fe-C bond length is 2.120(6) A . 2 4 An alkyl-pyridine chelating ligand was used to prepare FeR2 (R = C(SiMe3)2C5H4N-2), with Fe-C bond lengths of 2.154(8) and 2.139(7) ^ 25,26,37 T h e complex Fe(CH2Ph)2(TMEDA) was characterized by lK NMR and 2 H NMR spectroscopy only.38 Paramagnetic, low-coordinate iron(II) aryl complexes are more numerous; 217 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide andALkyl Complexes Table 5.3 Selected bond lengths and angles in Fe{CH(SiMe3)2}[N(SiMe2CH2PPh2)2] (48). Atom Atom Distance (A) Atom Atom Distance (A) Fc P(D 2.4886(6) Fe N 2.002(2) Fe P(2) 2.4924(6) Fe Cpl) 2.069(2) N Si(l) 1.710(2) N Si(2) 1.713(2) Atom Atom Atom Angle (°) Atom Atom Atom Angle O P(l) Fc P(2) 110.39(2) P(D Fe N 92.78(5) P(D Fe 0(31) 109.65(6) P(2) Fe N 87.03(5) P(2) Fe 0(31) 121.39(6) N Fe C(31) 131.20(7) Si(l) N Si(2) 129.93(10) Fe N Si(l) 112.64(8) C(41 Figure 5.6 Molecular structure (ORTEP) and numbering scheme for Fe{CH(SiMe3)2}[N(SiMe2CH2PPh2)2] (48). 218 References begin on page 246 ChapterS: Fe(II) andFe(III) Halide and Alkyl Complexes most use bulky aryls such as mesityl (2,4,6-Me3C6H2) and "super-mesityl" (2,4,6-'Bu3C6H2) ligands to provide steric protection to the metal centre. Bis-mesityliron(II) has been extensively studied as one of the few stable paramagnetic iron(U) aryl complexes; it also has the advantage of being a homoleptic compound.39 The solid state structure is a dimer containing both bridging and terminal mesityl groups with Fe-C bond lengths of 2.117(5) A (average) and 2.022(6) A respectively.40 The corresponding compound with a bulkier mesityl group, namely 2,4,6-z'Pr3C6H2, is still dimeric but the Fe-C distances lengthen to 2.083(9) and 2.104(9) A (terminal) and 2.174(9) (average, bridging), due to the increased steric interaction 4 1 Upon switching the isopropyl groups to the even more bulky r-butyl substituent the compound becomes monomeric and two-coordinate, with an Fe-C bond length of 2.051(5) A . 2 8 , 2 9 Monomeric, tetrahedral adducts of [Fe(Mes)2]2 with phosphines and amines have been reported. Fe(Mes)2(phen)42 and Fe(Mes)2py241 have both been structurally characterized and contain terminal mesityl groups with Fe-C bond lengths of 2.063(4) and 2.148(5) (average) A respectively. A series of dimeric iron(II) mesityl complexes with isocyanides and nitriles have also been reported and an in-depth study of their magnetic interactions was presented 4 1 , 4 3 Other iron aryl complexes include FePh2(PEt3)2 (thermally unstable, no structure)27 and some anionic "ate" complexes44,45 such as Fe(napthyl)42_. In the case of 48 the Fe-C(31) bond length of 2.069(2) A is substantially shorter than the other iron(II) alkyls, despite the apparent steric bulk of the alkyl ligand. In fact, this bond length is shorter than that observed in FeN4(CO)(CH3) (N4 = macrocyclic tetradentate N-donor, C10H19N8), an 18-electron macrocyclic complex with a terminal Fe-CH3 ligand (2.077(5) A ) 4 6 It is also comparable with some of the shorter iron(II)-aryl bond lengths reported, even though metal-alkyl bonds are usually longer than metal-aryl bonds. As well, the metal-carbon bond does not seem to be sensitive to spin state in the same way as metal-phosphine bonds. Of course, back-bonding is not an issue in determining the strength of a M-C bond so perhaps this should not be surprising. There are not enough iron(II)-alkyl complexes structurally characterized to address this point clearly; complex 48 is to my knowledge only the third such compound in the literature. 219 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes (iv) Reaction of Fe{CH(SiMe3)2}[N(SiMe2CH2PPh2)2] (48) with small molecules Iron alkyl 48 does not react with ethylene at one atmosphere (Scheme 5.1) either at room or at elevated temperatures over the course of one week. This result is not surprising; other iron(II) alkyls do not react with ethylene even under more extreme conditions.24 This has been explained as being due to the lack of an available orbital to bind ethylene and the lack of an agostic interaction to facilitate olefin insertion into the Fe-C bond. No reaction Decomp. to Fe metal + organic fragments Iron alkyl 48 reacts slowly with four atmospheres of hydrogen (Scheme 5.1) to produce a deposit of iron metal. Hence, complete decomposition of the complex is observed. With the series of complexes FeR2(dippe) (R = CH2PI1, CH2C6H4Me-/? , CH2SiMe3) , either no reaction with hydrogen occurs or reduction to Fe(0) products are observed, depending on the a lkyl group p resen t . 2 4 , 4 7 To my knowledge there are no known four coordinate neutral iron(II) hydride complexes. Scheme 5.1 H 2 C " ^ C H 2 P h 2 -P Me 2 Si M e 2 S i \ 1 : N — F e — C H ( S i M e 3 ) 2 : / Phe 48 220 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes 5.5 Reaction of Iron(II) Halide and Alkyl Complexes with Benzyl Halides The one-electron redox chemistry found in the reaction of chromium and cobalt(II) alkyls and halides with organic alkyl halides should theoretically be available to an iron(II) centre. In fact, there are few reports of alkyl halide one-electron redox behaviour with iron centres. Reactions of iron complexes with alkyl halides have been studied, but interest has centred on anionic iron(O) complexes to give iron(II) compounds. Substrates examined include [CpFe(CO)2]-,48 [CpFe(COD)]-,38 [(ri3-allyl)Fe(CO)3]- 49and a variety of iron porphyrins.50-51 This is therefore a formally two electron process and is analogous to the reaction of anionic cobalt(I) "superbases" with alkyl halides to give cobalt(III) alkyls.52 The two-electron oxidative addition of Mel to Fe(CO)3(PMe3)2 has been reported.53 In a singular case, the one-electron redox reaction of [(r|5-C5Ph5)Fe(CO)2]2 with alkyl halides RX to give the two products (rj5-C5Ph5)Fe(CO)2R and (r|5-C5Ph5)Fe(CO)2X has been investigated.54 The mechanism is analogous to that postulated for the chromium and cobalt (II)/(III) reaction but in this case the reaction involves an iron(I) to iron(II) redox system. Although the one-electron oxidation of chromium(II) and cobalt(II) to the +3 oxidation state by reaction with alkyl halide has been extensively examined, there seems to be a paucity of information on the corresponding reaction with iron complexes. As this chemistry was probed in detail in the MY[N(SiMe2CH2PPh2)2] (M = Cr, Co, Y = alkyl, halide) system, the reaction of analogous iron(II) complexes with benzyl halides was investigated for comparison. (i) Reaction of FeCl[N(SiMe 2CH 2PPh 2)2] (43) with benzyl bromide Upon addition of five equivalents of benzyl bromide to a toluene solution of the chloride 43, no immediate reaction occurred but over time the solution turned dark red/purple and the concomitant formation of a beige precipitate was observed (equation 5.10). After workup, the 221 References begin on page 246 Chapter 5: Fe(II) andFe(III) Halide and Alkyl Complexes dark red toluene solution yielded FeBr2[N(SiMe2CH2PPh2)2] (49), which was identified by elemental analysis. This is the expected product of halide transfer, analogous with the cobalt reaction. The full characterization (including X-ray structure) and discussion of this product will be presented in the next section. Me 2Si Me2Si \ : N -Ph 2 •P Me2Si \ : N — F e ' Br Me5Si Ph 2 .p IVIC2' xs PhCH2Br •Fe-—CI * / 1 -P Ph 2 43 Ph 2 and Ph, •P Me 2Si: 49 > S B r (<50%) (>50%) \ 1 0 :N Fe Br 50 M e 2 S i ^ 1 \ Br © PPh 2 CH 2 Ph [5.10] This result shows that one-electron oxidation from iron(II) to iron(III) does indeed progress, although it seems to be substantially slower than for the cobalt and chromium systems. A detailed kinetic investigation was not inititated because of the complication of the other product, which, based on the experience in the cobalt system, was assumed to be the zwitterionic species FeBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2CH2Ph)] (50). Unfortunately, in the iron case, this side reaction to form the zwitterionic product competes much more strongly than for cobalt and is in fact the primary reaction route. As a result, attempts to do reactions in large excesses of benzyl bromide resulted in primarily zwitterion formation. No crystal structure is available but 222 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes spectroscopic evidence does point to such a species. The 31P{ iHJNMR spectrum shows a singlet at 27 ppm and the *!! NMR spectrum in CD2CI2 shows a series of broad singlets similar to that observed for the cobalt zwitterion species. The beige colour is consistent with a tetrahedral Fe(II) system, as explained earlier, and the solubility properties conform with that expected for a zwitterionic species. Ionic compounds of iron(II) have been observed to form from reaction of an iron(III) product with chlorinated solvents; the mechanism was not investigated.55,56 (ii) Reaction of Fe{CH(SiMe3)2}[N(SiMe2CH2PPh2)2] (48) with benzyl bromide Addition of an excess of benzyl bromide to Fe{CH(SiMe3)2}[N(SiMe2CH2PPh2)2] (48) gave a "mixed bag" of products, representing all possible combinations of reactivity observed in the cobalt system (Scheme 5.2). The majority of the product (>75%) was the zwitterionic beige precipitate, which contained a mixture of two complexes observable by 3 1P{ 1H} NMR: FeBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2CH2Ph)] (50, 80%) from intermolecular alkyl reaction with the phosphines and FeBr2[N(SiMe2CH2PPh2)(SiMe2CH2PPh2CH(SiMe3)2)] (51, 20%) via an intramolecular mechanism. These species were not easily separated and so were not characterized in depth. In addition, some red/purple product, the iron(III) dibromide 49 was also isolated. This represents the redox pathway, and is comparable to that observed in the cobalt case. That is, a halide for alkyl substitution occurs, forming first iron(JJ) bromide 44, which then reacts either in a zwitterionic reaction (as described in the last section) to form benzylphosphonium 50 or in a redox fashion to form the iron(UI) dibromide 49. This reactivity and its mechanism was not probed in great detail, but these preliminary results suggest that the possible courses of reaction are similar to that observed in the cobalt system. There do not seem to be any examples of trigonal bipyramidal iron(III) alkyl complexes. Stable iron(III) alkyl systems generally involve tetradentate square 223 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes planar ligands such as porphyrins or Schiff bases and are square-pyramidal or octahedral.57"60 Both high and low spin alkyls and aryls are known.61 Scheme 5.2 xs PhCH 2 Br P h 2 •P Me 2 SK \ 1 0 : N Fe iBr M e 2 S i ^ 1 © P P h 2 R Br R = C H 2 P h 50 CH(S iMe 3 ) 2 51 Me 2 Si x s P h C H 2 B r M e 2 S i \ : N -/ P h 2 •P • F e — B r -P 44 P h 2 xs P h C H 2 B r P h 2 -P M e 2 S . v \ I / : N — F e ^ Me2Si<f j ^ ' Phe 49 Br Br The fact that some halide for alkyl substitution occurred and the putative intermediate Fe(R)Br[N(SiMe2CH2PPh2)2] (I) was not isolated implied that the iron(IU) alkyl halide complex was unstable. In order to test this, the iron(IU) dibromide 49 was reacted with one equivalent of MeLi and LiCH2SiMe3 in an attempt to generate an iron (III) alkyl halide complex I (equation 224 References begin on page 246 Chapter 5: Fe(II) andFe(III) Halide and Alkyl Complexes 5.11). As in the cobalt system (but different than the chromium system) this intermediate was unstable and decomposed to the observed product, FeBr[N(SiMe2CH2PPh2)2] (44), an iron(II) species, presumably by metal-carbon bond homolysis. Me 2 Si Me 2 Si \ : N -/ •Fe' c jN B r LiR P h 2 P M e 2 S i ^ : N -| Br 3 MegSi^ P P h 2 49 P h 2 -P • F e — B r + 0.5 P h C H 2 C H 2 P h + RH [5.11] •P P h 2 R = Me, C H 2 S i M e 3 44 5.6 Synthesis, Characterization and Magnetic Properties of Iron(III) Dihalide Complexes The synthesis of FeBr2[N(SiMe2CH2PPh2)2] (49) by one-electron oxidation of FeX[N(SiMe2CH2PPh2)2] (X = CI (43); X = Br (44)) was described in the previous section but this route proceeds in low yield and also has the disadvantage of not being amenable to the introduction of different halides such as chloride. However, the availability of anhydrous FeCl 3 and FeBr3 provided an alternate route to the iron(III) dihalide complexes. Thus, dropwise addition of LiN(SiMe2CH2PPh2)2 to an ether solution of FeX 3 gave the corresponding iron(IU) dihalide complexes FeX2[N(SiMe2CH2PPh2)2] (X = Br (49); X = CI (52)) in excellent isolated yield (equation 5.12). Strictly anhydrous FeX 3 was found to be essential; most commercially available F e C l 3 is often not completely dry. Both FeCi2[N(SiMe2CH2PPh2)2] (52) and 225 References begin on page 246 Chapter 5: Fe(II) andFe(III) Halide and Alkyl Complexes FeBr2[N(SiMe2CH2PPh.2)2] (49) were characterized by elemental analysis, mass spectrometry and for the dibromide, by X-ray crystallography. The X-ray structure of 49, shown in Figure 5.7, confirms the composition of the compound as an iron(III) dihalide, and reveals a nearly perfect mononuclear, trigonal bipyramidal arrangement of ligands around the iron centre. Selected bond lengths and angles are in Table 5.4. The axial P(l)-Fe-P(2) angle of 173.35(8)° and the equatorial angles of 122.8(2)°, 121.0(2)° and 116.10(5)° for N-Fe-Br(l), N-Fe-Br(2) and Br(l)-Fe-Br(2) respectively indicate the minimal distortion around the iron centre. The Fe-N bond of 1.951(6) A in the dibromide 49 is quite long compared to 1.918(4) A in Fe[N(SiMe3)2]362 and 1.896(5) and 1.900(5) A in FeI(d5-py)(NRArf)2 (R = C(CD 3)2CH 3; Ar f = 2,5-C6H 3FMe) 1 4 The N-Si bond lengths in 49 are 1.735(6) and 1.743(6) A. Other structurally characterized trigonal bipyramidal iron(III) compounds include the pentaazidoiron(III) anion Fe(N3)52",63 FeCl3(4-cyanopyridine)264 and FeCl3(NMe3)2.65 More recent and more relevant is the investigation by Walker and Poli of phosphine adducts of FeCl3 and the influence of the phosphine on the spin state.56 Two adducts were structurally characterized, namely FeCi3(PPh3)2 and FeCl3(PMe3)2, both of which are trigonal bipyramidal. The PPI13 adduct as well as the early examples mentioned are all high-spin d 5 compounds with magnetic moments around 5.9 B.M. The trimethylphosphine adduct was described as an "intermediate Ph 2 [5.12] X = Cl 52 Ph 2 226 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and ALkyl Complexes Table 5.4 Selected bond lengths and angles in FeBr2[N(SiMe2CH2PPh2)2] (49). Atom Atom Distance (A) Atom Atom Distance (A) Fc P(l) 2.330(2) Fe N 1.951(6) Fe P(2) 2.350(2) Fe Br(l) 2.4201(13) N Si(l) 1.735(6) Fe Br(2) 2.4263(12) N Si(2) 1.743(6) Atom Atom Atom Angle (") Atom Atom Atom Angle (") P(l) Fe P(2) 173.35(8) N Fe Br(l) 122.8(2) N Fe Br(2) 121.0(2) Br(l) Fe Br(2) 116.10(5) N Fe P(l) 87.3(2) Si(l) N Si(2) 125.5(3) Figure 5.7 Molecular structure (ORTEP) and numbering scheme for FeBr2[N(SiMe2CH2PPh2)2] (49). 227 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes spin" system (S = 3/2, three unpaired electrons, p s . 0 . = 3.87 B.M.) 3 , 6 on the basis of spectroscopic and magnetic data; it showed variable temperature magnetic data with a value of 4.9 B.M. at room temperature, dropping to 4.2 B.M. at -78 "C. The Fe-P bond lengths in the high and intermediate spin systems are substantially different. In high-spin FeCl3(PPh3)2, very long Fe-P distances of 2.654(4) and 2.623(4) A are observed. In contrast, the Fe-P bonds in intermediate spin FeCl3(PMe3)2 are markedly shorter, at 2.342(5) and 2.332(5) A. The Fe-P bonds in dibromide 49 are 2.330(2) and 2.350(2) A, which is consistent with the spin state of the iron(III) centre being S = 3/2, intermediate spin, much like the trimethylphosphine adduct. Accordingly, the solution magnetic moments of the dibromide 49 and dichloride 52 were measured to be 4.6 and 5.0 B.M. respectively by Evans' method.4,5 The trigonal-bipyramidal cobalt(III) dihalide complexes (34-36) presented earlier (section 4.2) were also found to be intermediate spin containing, in this case, S = 1 Co(IU) d 6 centres. Additional evidence for the presence of an S = 3/2 iron centre is the UV-vis spectrum of the dihalide complexes. A pure S = 5/2 spin metal complex, containing five unpaired electrons, would have a very pale colour and no d-d transitions observed in the visible range, due to the lack of any spin allowed transitions from a ground state sextet term ( 6Ai). However, an S = 3/2 state will have spin-allowed transitions available for d-d transitions to be observed. Hence, while the detailed assignment of such bands is not vital, their observation signals the presence of an intermediate-spin Fe(UI) centre. Accordingly, the UV-vis spectra of dibromide 49 and dichloride 52 are shown in Figure 5.8 and it is immediately obvious that, along with high-energy (>350 nm) charge-transfer bands, the visible region also contains spectral features. The dibromide 49 has a dominating peak at 528 (e = 5350 M^cnr1) nm, a weaker band at 716 (e = 1270 M^cnr1) nm and shoulders at 440 (e = 3560 M^cnr1) and 380 (4980 M^cnr1) nm. The dichloride 52 has peaks at 330 (e = 5990 M^cnr 1), 424 (e = 6700 M^cnr 1) and 646 (e = 1220 M-icnr 1) nm respectively. The complex FeCl3(PMe3)2 has a band at 378 (e = 3400 M-icnr 1), 500 (e = 10000 M - W 1 ) and 578 (e = 2800 M-icnr1) nm, while the high-spin FeCl3(PPh3)2 has no spectral features below 360 228 References begin on page 246 Chapter 5: Fe(II) andFe(III) Halide and Alkyl Complexes nm. 5 6 These bands could be assigned to d-d transitions despite their large extinction coefficients as the Laporte selection rule is relaxed for five-coordinate complexes, allowing for greater intensities than in octahedral systems. On the other hand, consideration of these absorbances as halide to metal charge transfer bands, comparable to that observed in CoX2[N(SiMe2CH2PPh2)2] (34-36) cannot be ruled out. 1.14 0.91 i g 0.69 8 0.46 H 0.23-^  0.00 300 500 Wavelength (nm) 80o" 400 500 600 Wavelength (nm) 700 800 Figure 5.8 UV-vis spectra for FeX2[N(SiMe2CH2PPh2)2]. X = Cl (52, top), X = Br (49, bottom). 229 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes The *H NMR spectra of the dihalide complexes featured broad peaks which were in some cases difficult to integrate. For the dichloride 52, two equal intensity peaks at 25.9 and 22.5 ppm and a peak at -7.6 ppm of half the intensity were observed. A very broad peak could also be observed centred around 4 ppm. A variable temperature NMR study showed all peaks sharpening and shifting towards the diamagnetic range, as would be expected, but even at 60 °C assignments could not be made. The dibromide 49, surprisingly, had a substantially different spectrum, with four observable peaks. Broad peaks at 6.9 and 9.0 and a sharper peak at 10.1 ppm integrated 2:2:1 in intensity. There was also a very broad peak centred around 18 ppm, although the integration was difficult to obtain accurately. Most interesting, however, is the variable temperature behaviour of the dibromide 49 (Figure 5.9). Although the dichloride shows regular paramagnetic NMR temperature behaviour, the resonances in the dibromide become broader and more paramagnetically shifted as the temperature is increased; the reverse of what is expected. A measurement of the magnetic susceptibility at increasing temperature confirmed this: the observed moment is u.eff = 4.62 B.M. at 25 °C, 4.68 B.M. at 40 °C and 4.75 B.M. at 60 °C. This change of 0.15 B.M. within the same sample is not attributable to systematic error; the compound is becoming more paramagnetic at higher temperature. This is the only complex reported in this thesis where this behaviour was observed, and it is interesting to note that the dichloride does not show this behaviour. In retrospect, the observed room temperature magnetic moments of 4.6 and 5.0 B.M. for the dichloride and dibromide are really too high to be considered as pure S = 3/2 intermediate spin systems, even with the effects of spin-orbit coupling. Spin-orbit coupling in these systems would have to be second-order since the ground state is orbitally non-degenerate (4A2), so the effect should be relatively small. Hence, the anomalous room temperature moments also offer an insight into unusual magnetic behaviour in these systems. To compare, a true S = 3/2 system for Fe(III) is exemplified by the bis(N,N-diethyldithiocarbamato)Fe(III)X (X = halide) series of complexes.66"71 These distorted square-pyramidal complexes have room temperature magnetic 230 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes c Figure 5.9 Variable Temperature lU NMR spectra for FeBr2[N(SiMe2CH2PPh2)2] (49) from 6-12 ppm. Spectrum (a) 60 °C, (b) 40 °C and (c) 25 °C. 231 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes moments ranging from 3.92 to 4.07 B.M. depending on the halide, values which are quite close to the spin-only S = 3/2 value of 3.87 B . M . 6 7 - 7 0 An unusual, distorted iron(III) porphyrin, with a vinylidene group inserted into an Fe-N bond has also been reported to be an intermediate-spin complex, with pgff = 3.9 B .M. 7 2 These values are found to be essentially temperature-independent (Curie law) except at very low temperatures where zero-field splitting causes a reduction in observed moment.71 Very small intermolecular ferro- and antiferromagnetic interactions at very low temperature have also been observed, but the key point here is that a pure S = 3/2 spin state does not have a room temperature moment of 4.6 B.M., nor is it temperature dependent at high temperatures, as found for 49. There are several possible explanations for the observed magnetic behaviour of the dihalide compounds FeX2[N(SiMe2CH2PPh2)2] (X = Br (49); X = CI (52)). The crystal structure showing a monomeric system rules out any iron-iron exchange interactions so antiferromagnetic coupling of two high-spin S = 5/2 centres can be discounted. Restricting the discussion to mononuclear systems, there are essentially three situations to consider. Firstly, the assignment of the iron centre as a pure intermediate spin S = 3/2 centre with some orbital contribution to the moment can be considered and rejected, as the values of the room temperature moments of 49 and 52 are really too high when compared to validated S = 3/2 compounds. In order to obtain a higher moment, some high-spin S = 5/2 character must be included into the complex. One possibility is that the room temperature moment is fortuitously in the middle of a thermal equilibrium curve describing a S = 3/2 (or even S = 1/2) to S = 5/2 spin transition. The low-high spin equilibrium case is well documented in the literature3'6 but there are only a few examples of spin equilibrium between S = 3/2 and 5/2 Fe(UI).3 One of these is the aforementioned FeCl3(PMe3)2 which has variable temperature magnetic data (as stated earlier) consistent with a thermal equilibrium between a ground state S = 3/2 centre and low-lying S = 5/2 centre. Interestingly, changing the phosphine from PMe3 to P(cyclohexyl)3 yields another thermal equilibrium, this time with the S = 5/2 state being the ground state.56 The key feature of such an equilibrium is a large temperature dependence of the magnetic moment and a sigmoidal-type curve of the moment vs. temperature. 232 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes There is one other way to integrate high-spin character into S = 3/2 complexes. Instead of observing a spin-equilibrium between S = 3/2 and 5/2 states a quantum mechanical mixing of the two states can occur to create a new, discrete spin-admixed ground state that is neither S - 3/2 or 5/2 but has elements of both.3,73"76 Such admixed states give rise to magnetic properties that lie between the extremes of intermediate and high-spin systems (i.e., |ieff = 3.9 - 5.9 B .M. ) . 7 2 , 7 5 , 7 7 Fe(TPP)C104,76-78 Fe(OEP)C1047 6'7 9 , 8 0 and Fe(Pc)Cl81 are the classic examples of complexes exhibiting this spin-admixed behaviour, with room temperature magnetic moments of 5.2, 4.8 and 4.53 B.M. respectively, indicating different amounts of spin-state mixing depending on the ligand set. In addition, the temperature dependence of such a spin-admixed system is quite different than that of spin-crossover complexes. Instead of a sigmoidal curve, essentially Curie-law behaviour with a slight reduction in moment over a large temperature range is observed. 7 3 , 7 6 , 7 8 , 8 1 At low temperatures other factors such as zero-field splitting begin to become important and the moment often decreases faster. However, no sharp transition is observed. Hence, it is possible to differentiate spin-admixed systems from spin-crossover systems on the basis of a variable-temperature magnetics study. A preliminary solid-state variable temperature magnetic susceptibility study of dibromide 49 using a SQUID magnetometer revealed a slow, smooth reduction in moment from 4.70 B.M. at 300 K to 4.45 B.M. at 20 K and not a sharp, spin-crossover curve (Figure 5.10). That is, the magnetic moment more or less obeyed the Curie law over the temperature range discussed. From 20 K the moment begins to drop more rapidly; at 2 K the observed magnetic moment is 3.49 B.M. Hence, it appears that dibromide 49 is a rare example of a spin-admixed Fe(III) system, and is to my knowledge the first documented trigonal-bipyramidal spin-admixed Fe(ni) complex. What factors affect whether a compound will be pure intermediate-spin S = 3/2 or high-spin S = 5/2? Why do some compounds exhibit spin-admixed behaviour while others show thermal spin-crossover behaviour? The theory of spin-admixed systems indicates that essentially 233 References begin on page 246 Chapter 5: Fe(tt) and Fe(III) Halide and Alkyl Complexes 5.0 03 4.5 o o o o o o o o o o o o o o o o o o o o o o o o j 8 ,2 4.0 -I o o o o o 0 3.5 3.0 A , , , . , 1 0 50 100 150 200 250 300 Temperature (K) Figure 5.10 Graph of magnetic moment vs. temperature for FeBr2[N(SiMe2CFJ.2PPh2)2] (49). two parameters can address these questions, namely the energy difference between S = 3/2 and S = 5/2 spin-states (A) and the spin-orbit coupling constant (k or Q.3,73.74 The energy separation A is a function of the ligand field and geometry of a given complex and is thus highly variable. In contrast, the spin-orbit coupling constant £ is relatively insensitive to ligand field changes but its magnitude is usually less in complexes than that tabulated for the free ion. The detailed theory is based on quantum mechanical calculations which show that admixing of spin states occurs via the mechanism of spin-orbit coupling and will only occur if the energy separation between the pure S = 3/2 and S = 5/2 spin state is less than or equal to the spin orbit coupling constant. As a result, the extent of spin-admixture expected can be calculated in terms of the parameter A/£ (Figure 5.II).3-7 3 In the case where A/£ is large and negative, a pure S = 5/2 high-spin complex is expected. If A/£ is large and positive, a pure S = 3/2 intermediate-spin complex will be observed. In these 234 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes pure states, each spin-state can undergo zero-field splitting, which affects the magnetic moment at low temperature. Only in the case where A/£ <1 will spin-admixed systems result, and at the point where A/£ = 0 (i.e. the S = 3/2 and S = 5/2 levels are equivalent in energy) a system with 50% 3/2 and 50% 5/2 character results. For example, an attempt to fit the magnetic susceptibility vs. temperature data for Fe(TPP)C104 using the model described led to a reasonable fit for A = 108 cm - 1 and C, = 188 cm - 1, which is consistent with a ground-state wavefunction of 35% 6 A i (S = 5/2) and 65% 4 A 2 (S = 3/2); the room temperature magnetic moment is 5.04 B . M . 7 6 A/r,Large Neg. 40 i 2D •* 1/2 £ w CM M / 2 « 1 ^ 1 A / t Large Pos. \\ 40 i •*l/2 5! .23/2 « II w 45/2 CM -3/2*5 M/2 " r I r A/C<l I i i t 1_ •i|/2 •*3/2 ±5/2 ±3/2 -1/2 Figure 5.11 Energy levels for iron(III) electronic states in a tetragonal field, (a) A/£ is large and negative; (b) A/£ is large and positive; (c) A/£ <1, spin admixed system. Taken from reference 73. Spin-orbit coupling selection rules prohibit the admixture of states with AS = 2.73 Hence, low and high-spin iron(III) centres can show spin-crossover but not spin admixture. In the case of spin-admixture vs. spin-crossover for S = 3/2 and S = 5/2, if the energy difference between the 235 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes two states is of the order of kT (A ~ kT, thermal population possible) and is much higher in energy than the spin-orbit coupling constant (A/£ is large, which would then preclude admixture) then sharp spin-crossover can be observed. A salient point of this discussion is that a vast range of magnetic behaviour is accessible in Fe(III) systems, and very small changes in ligand field strength and metal geometry can result in radically different magnetic systems. As such, Reed and Guiset recentiy proposed that a "magnetochemical series" could be defined by using the series of complexes Fe(TPP)X, where X is a weakly binding axial anion.7 5 , 7 8 By varying X and observing the extent of spin-admixture by various measurements (magnetic susceptibility, Mossbauer, ESR, lH NMR spectroscopy) one could rank the relative field strengths of X. A similar series was investigated for Fe(Pc)X complexes.81 All of these systems are similar in that they contain macrocyclic, strong field equatorial ligands and a weakly donating axial ligand, to give a square pyramidal complex. Only in one case has the effect of changing ligands on spin state in trigonal-bipyramidal Fe(UI) systems been examined, that being the FeCl3(PR3)2 system.56 In this case, the effect of changing the phosphine on the spin state was examined, with interesting results, but the effect of changing the equatorial halides was not explored. In the FeX2[N(SiMe2CFJ.2PPh2)2] system an exploration of the effect of changing X on the magnetic behaviour and in particular the extent of spin-admixture would be advantageous in order to further understand this phenomenon. On that note, the dichloride 52 has already been prepared; the higher room temperature moment of 5.0 B.M. relative to the dibromide, and the normal variable temperature lH NMR spectral behaviour relative to the dibromide 49 are indications that different magnetic behaviour with halide substitution is occurring. The other compound of interest, Fei2[N(SiMe2CH2PPh2)2] (53) could potentially be prepared from iodine oxidation of FeI[N(SiMe2CH2PPh2)2]. Iodine oxidation of Fe(II) complexes is fairly straightforward; this route was used in the preparation of FeI[NAr f] 2 .(R = C ( C D 3 ) 2 C H 3 ; Ar f = 2 , 5 - C 6 H 3 F M e ) 1 4 Similarly, the reaction of FeCl[N(SiMe2CH2PPh2)2] (43) with elemental iodine was performed. Addition of iodine to a 236 References begin on page 246 Chapter 5: Fe(II) and Fe(III) Halide and Alkyl Complexes pale yellow solution of 43 resulted in the rapid formation of a deep green solution with UV-vis properties consistent with FeIX[N(SiMe2CH2PPh2)2]; intense absorptions at 326, 376 and 608 nm were observed. Alternatively, the reaction of FeCl2[N(SiMe2CH2PPh2)2] (52) with Me3SiI also yields a deep green solution but the resulting product is a mixture of mono- and diiodo-substituted complexes. These preliminary reactions indicate the viability of such synthetic routes. The study could be expanded to include a large range of ligands, including alkoxides, amides, cyanides or thiolates. 5 . 7 Summary and Conclusions The chemistry of iron(II) coordinated with the amidodiphosphine ligand •N(SiMe2CH2PPh2)2 yielded some unexpected results. Although the initial iron(II) halide complexes could be easily prepared, the subsequent metathesis to form iron(II) alkyls proved to be quite difficult; iron(II) alkyls could only be prepared with very bulky alkyl groups. All of the iron(II) complexes (except for the Cp complex) prepared were high-spin tetrahedral systems. Hence, elements of both the chromium(II) and cobalt(II) chemistry were evident. Like the chromium system, alkylation did not cause a change in spin state, but the geometry observed was tetrahedral (as in high-spin cobalt(U)) and not square-planar like in the chromium system. A comparison of the reactivity of the iron(II) system with alkyl halides to the cobalt and chromium systems reveals common features; the general reaction mechanism outlined for these one-electron oxidation reactions (Scheme 4.1) is followed. However, zwitterion formation is much more of a factor for iron(II) than for cobalt(II). On the other hand, the instability and decomposition of iron(JJI) alkyl halides is reminiscent of the cobalt syst