Structural properties of pyrazolyl-bridged diiridium complexes

Ron D . Brost

Bachelor o f Applied Science (Chem), British Columbia, 1984

A DISSERTATION SUBM ITTED IN PARTIAL FULFILLM ENT OF THE REQUIREMENTS FOR THE DEGREE O F DOCTOR OF PHILOSOPHY

in the Department o f Chemistry

W e accept this dissertation as conforming to the required standard

Dr. T. Ftyles ~ Dr. C. Piccidtto " D r f a Rp.id DrJM . C ow ie ® Ron D . Brost, 1991 UNIVERSITY OF VICTORIA

All rights reserved. This dissertation may not be reproduced in w hole or in part, by mimeograph or other m eans, without permission o f the author.

A C C E P T E D

F A C U L T Y Of- G R A DU A TE S T U D I E S

Dr. S.R . Stobart

DATE.

A bstract

The x-ray structures o f several alkyl halide, alkyl dihalide, and hydrogen adducts to pyrazolyl-bridged diiridium com p lexes [Ir(L)(L')(|i-pz) ] 2 (L=CO, L'=CO, PPI13, pzH = pyrazolyl) are determined. The diiridium(bis-pyrazolyl) core o f these com plexes enables contact between the two centers so that metal-metal bond formation m ay occur, exem plified by a short iridium -iridium distance o f 2 .7 8 (1 ) A in the diiridiu m (II) co m p lex [Ir2(CO)4(p-p z)2(CH(CH3)?,)I]. O xidation mechanisms are postulated based on reaction kinetics. The oxidative addition o f m ethyl iodide to [Ir(CO)4(p-p z) ] 2 (3) is observed to occur by a two-step mechanism, where a high positive AS* term m ay be due to a highly ordered intermediate. This is proposed as evidence for an S n 2 addition, where coordination o f the alkyl halide is follow ed by halide dissociation and migration to a tra n s diaxial coordination site. Different kinetics o f the reaction are observed in THF and benzene, w hich is also attributed to a polar S n 2 intermediate. O ccupation o f the 3,3' and 5,5' positions o f the pyrazolyl ligand decreases the reaction rate by an order o f magnitude or greater, w hich indicates steric. inhibition o f the reaction by the bridging ligands. Experimental evidence for a competing light-induced reaction that corresponds to a radical- chain m echanism rather than the dark Sjs}2 reaction is also presented. O xidative isom erization o f an iodo(iodom ethylene) com plex to the m ethylene-bridged isom er is determ ined to be an intramolecular process based on isotope labelling experim ents and kinetics. N eg lig ib le isom erization to the bridging m ethylene com p lex under am bient c o n d it io n s is a ttr ib u ted to c o o r d in a tiv e sa tu ra tio n ; th e s t a b ilit y o f [Ir2(PPh3)2(C O )2(lt-pz)2(|i-C H2)(I)2] is likew ise due to coordinative saturation o f the m etal centers. The addition o f hydrogen or hydride to [Ir(PPh3)(CO)(|Li-pz) ]2 is possible through a number o f synthetic routes, but the stereochemistry o f the iridium(II) hydrido

hydride induces structural rearrangements in substitution reactions so that stereochemistry o f parent com plexes is not conserved. This is dem onstrated by the x-ray structures o f [ I r ( P P h 3 ) ( C O ) ( n - p z ) H ] 2 , [ I r 2 ( P P h 3 ) 2 ( C O ) 2 ( f i - p z ) 2 H C n , and [Ir2(P P h3)2(C O )2(|i-p z )2HI]. The hydride ligand prom otes nucleophilic attack on an electron-rich iridium center; thus water and other L ew is bases are found to react with the cationic diiridium hydride com plex [Ir2(PPh3)2(CO)2(M--pz)2l(B F4).

Examiners: Df. S.R . Stotfart^ Dr. G .W ?Bushnell Dr. T. EyJ$ Dr. C. Picciotto Dr. R. Peid" ) ^ Dr. M . Cow le

Table o f Contents

I. THE METAL-METAL BOND IN OXIDATIVE A D D IT IO N S...1

II. STERIC A N D ELECTRONIC EFFECTS IN A L K Y L H ALID E A D D IT IO N ... 11

Introduction... 11

Results... 13

Kinetics o f Methyl Iodide Addition to 3 ... 23

D i s c u s s i o n ...47

III. STRUCTURAL REORGANIZATION IN THE FORM ATION OF METHYLENE COM PLEXES... 62

Introduction... ... 62

R esults...67

D i s c u s s i o n ...96

IV. STRUCTURAL FEATURES IN TH E H YDROG ENATIO N O F DIIRIDIUM CO M PLEXES... 104

Introduction... 104

R esults... 109

Crystallography o f Pyrazolyl-bridged Diiridium H ydrides...115

D i s c u s s i o n ... 170

S y n t h e s e s ...188

[Ir(C O D )(p-pz))2 ( 2 ) ... 188

[Ir2 ( C O ) 4 ( l i - p z )2( M e ) ]( P F 6) ( 1 2 ) ... 189

[Ir(PPh3)(13C O )(p-pz)]2 ( l a ) ... 189

[Ir2(PPh3)2(C O )2(li-pz)2(CH2l)IJ (1 8 ) ... 190

[Ir2(PPh3)2(C O )2(li-pz)2(li-CH2)l2l (1 9 )...190

[Ir(PPh3)(C O )0i-p z)I]2 ( 1 5 ) ... 192

Reaction o f Diazomethane with [Ir(PPh3)(CO)(|.i-pz)Ij2 (1 5 )... 193

Reaction o f Ethyldiazoacetate with fIr(PPh3)(CO)((.t-pz)I|2 ( 1 5 ) ...193

R eaction o f 1,1-Diiodoethane with [Ir(PPh3)(C O )(p -p z )l2 (1 )... 193

Reaction o f Dibromomethane with [Ir(PPh3)(C O )(|i-pz)]2 ( 1 ) ... 194

[Ir(PPh3)(C O )(|i-pz)(I I)]2 ( 2 3 ) ... 194

IIr2 (P P h 3)2( C O ) 2(H -p z )2(H )Cl] ( 2 6 ) ... 195

[I r (P P h 3) ( C O ) ( j i - P7 ) C l ] 2 ( 2 7 ) ... 195

[Ir2(PPh3)(CO)(ji-pz)HI] ( 2 8 ) ... 196

[Ir2(PPh3)2(C O )2(li-pz)2H](BF4) ( 2 4 ) ...196

[Ir2(PPh3)2(CO)2(li-pz)2(H-OH)(H)2l(BF4) (2 5 )... 197

Reaction o f Methanol with [Ir2(PPh3)2(CO)2(H-pz)2H ](BF4) ( 2 4 ) ... 198

Reaction o f Hydrogen Sulfide with [Ir2(PPh3)2(CO)2(H-pz)2HJ(BF4) ( 2 4 ) ...199

3 2 200 Reaction o f Sodium Hydride with 3 2 ... 200

[Ir(PPh3)(CO )(^-pz)(H )(CF3CO O )]2 (3 3 )...202

Kinetics Experim ents...202

Kinetics o f Methyl Iodide Addition to [Ir(CO)2(H-pz)]2 ( 2 ) ...202

Kinetic Study o f the Oxidative Isomerization o f (18) to ( 1 9 ) ...205

K inetics o f Water Addition to [Ir2(PPh3)2(C O )2(|i-p z)2(H )](B F4) ( 2 4 ) ... 2 06 Iso to p e L a b ellin g E x p erim en ts... ..2 0 8 Investigation o f Triphenylphosphine Exchange in 2 4 ... 208

Ligand Exchange in 18 and the 18 to 19 O xidative Isom erization... 209

Investigation o f CO E xchange in 2 ... 210

Crystallography o f Diiridium C om p lexes... 210

1. Reaction Tim es for the Carbonyl Substitution Reaction...15

2. D ecom position o f Alky! Halide Adducts to [Ir(CO)2(|i-pz)]2 ( 3 ) ... 19

3. Crystallographic Parameters for [Ir2(PPh3)2(CO)2(C3l l7)(I)(|i-pz)2] ( 4 ) ... 31

4. Atom ic Coordinates and Temperature Parameters for 4 ... 32

5. A n isotrop ic Tem perature Parameters for 4 ... 33

6. Interatomic Distances for 4 ... 34

7. Bond Angles for 4 ... 35

8. Activation Parameters for Iodomethane Addition to Iridium C om plexes... 46

9. NMR Data for M ethylene Com plexes o f [Ir(PPh3)(CO)(p.-pz)]2 ( I ) ... 82

10. 3 1 P { 1I I } N M R and IR D ata for M e th y le n e C o m p le x e s or fIr(PPh3) (C O )^ -p Z)]2 ( l ) ...83

11. Concentration Data for the O xidative isomerization o f 18 to 2 0 ... 84

12. Crystallographic Parameters for [Ir2(PPh3)2(CO)2(CH2l)(I)(|i.-pz)2l (1 8 )... 86

13. Fractional Atomic Coordinates and Temperature Parameters for [Ir2(PPh3)2(CO)2(CH2l)(I)ai-pz)2] (18) (non-phenyl a to m s)... 87

14. Fractional A tom ic Coordinates and Temperature Parameters for 18 (phenyl a t o m s ) ... 88

15. Anisotropic Temperature Parameters for 18 ... 89

16. Interatomic Distances for .18... 90

17. Bond A ngles for 1 8 ...91

18. Crystallographic Parameters for [Ir2(PPh3)2(CO)2(|!-CH2)(l)2(|!-pz)2] 0 9 ) ...92

19. Fractional Atomic Coordinates and Temperature Parameters for 1 9 ... 93

vm

21. Interatomic Distances for 1 9 ...95

22. Bond A n gles for 1 9 ...96

23. Infrared Data for Neutral Diiridium Hydrides... 126

24. ^ NMR Data for Neutral Diiridium H ydrides... 127

25. 3 lp{ 1H) NM R Data o f Neutral Diiridium Hydrides... 128

25. Crystallographic Parameters for [Ir(PPh3)(C O )(H )(p-pz) ] 2 ( 2 3 ) ...129

27. Fractional Atomic Coordinates for 2 3 ... . 130

28. Fractional Atomic Coordinates and Temperature Parameters for 2 3 ...131

29. Anisotropic Temperature Parameters for 2 3 ... 132

30. Interatomic Distances for 2 3 ... 133

31. Bond A n gles for 2 3 ... 134

32. Crystallographic Parameters for [Ir2(PPh3)2(CO )2(M--H)(p-pz)2l (2 4 )... 135

33. Fractional Atomic Coordinates and Temperature Parameters for 24 (non-phenyl a to m s )...136

34. Fractional Atomic Coordinates and Temperature Parameters for 24 (phenyl atom s)... 137

35. Anisotropic Temperature Parameters for 2 4 ... 138

36. Interatomic Distances for 2 4 ... 139

37. Bond A n gles for 2 4 ... 140

38. Crystallographic Parameters for [Ir2(PPh3)2(CO )2(H )(C l)(p -p z)2] (2 6 )...141

39. Fractional Atomic Coordinates and Temperature Parameters for 26 (non-phenyl a to m s )...142

40. Fractional Atomic Coordinates and Temperature Parameters for 26 (phenyl atom s)... 143

4 3 . Bond A n gles for 2 6 ... 146

4 4 . Crystallographic Parameters for [Ir(PPh3)(C O )(C l)(|.i-pz) ] 2 (2 7 )...147

45. Fractional Atomic Coordinates and Temperature Parameters for 27 (non-phenyl a to m s)... 148

46. Fractional Atomic Coordinates and Temperature Parameters for 27 (phenyl atom s)...149

47. Anisotropic Temperature Parameters for 2 7 ... 150

4 8 . Interatomic Distances for 2 7 ... 151

49. Bond A n gles for 2 6 ...152

50. Crystallographic Parameters for [Ir2(PPh3)2(CO)2(H)(I)(|.t-pz)2l ( 2 8 ) ... 153

51. Fractional Atomic Coordinates and Temperature Parameters for 28 ... 154

52. Fractional Atomic Coordinates and Temperature Parameters for 28 (phenyl atom s)...155

5 3. Anisotropic Temperature Parameters for 2 8 ...156

54. Interatomic Distances for 2 8 ... 157

55. Bond A n gles for 2 8 ...158

5 6. Crystallographic Parameters for [Ir2(PPh3)2(CO)2(H)(it-pz)2](BF4) ( 2 4 ) ... 159

57. Fractional Atomic Coordinates and Temperature Parameters for 24 (non-phenyl a to m s)...160

58. Fractional Atomic Coordinates and Temperature Parameters for 24 (phenyl atom s)... 161

59. Anisotropic Temperature Parameters for 2 4 ... 162

6 0 . Interatomic Distances for 2 4 ... 163

X

62. Infrared Data o f Cationic Diiridium Hydrides... — 166

63. ]H NM R Data for Cationic Diiridium H ydrides... 167

64. 3 ,P{ *H} NMR o f Cationic Diiridium H ydrides...168

6 5 . Com parison o f the Bond D istances in 2 3 , 2 6 , 2 7 , 2 8 ... 169

Pentacoordinated M etals... 3

2. Molecular Orbital Diagram for a Homobimetallic d7 - d7 S p e c ies...7

3. ORTEP o f [Ir2(C O )4(li-pz)2(C H (C H 3)2)I] ( 7 ) ...22

4. Tim e contours for the addition o f RI to [Ir(CO)2(lt-pz)]2 ( 3 ) ... 36

5. Tim e contours for the addition o f M el to [Ir(CO)2(p.-pz) ]2 ( 3 ) ... 37

6. Representative Tim e-Drive Scans for the Addition o f M el to [Ir(CO)2(|i-pz)]2 (3).... 38

7 . k2 vs [2] for the Addition o f M el to [Ir(CO)2(M.-pz)]2 (3) in TI-IF...39

8. k0bs vs Iodomethane Concentration for the Addition o f M el to [ I r ( C O ) 2 ( | i - p z ) ] 2 (3 ) in T H F ... 40

9. k0bs vs [2]2 ( M el + [Ir(CO)2([i-pz)]2 (3) in b enzene)... 41

10. k0bs v s. Iodomethane Concentration Squared ( M el + fIr(CO)20 i-p z )]2 (3) in ben zen e)... 42

11. Eyring Plot for Iodomethane + [Ir(CO)2(|i-p z)]2 (3), in T H F ... 43

12. Eyring Plot for Iodomethane + [Ir(CO)2(M-~Pz)]2 (3), in T H F ... 44

13. Eyring Plot for Iodomethane + [Ir(CO)2(|J.-pz)]2 (3), in b en zen e... 45

14. ORTEP o f [Ir2(PPh3)2(CO )2(li-pz)2(I)(CH2l)] ( 1 8 ) ...69

15. ORTEP o f [Ir2(C O )2(PPh3)2(|i-pz)2(li-C H 2)(I)2] ( 1 9 ) ... 74

16. Calculated vs Experimental m /e for [Ir2(CO)2(PPh3)2(|i-p z)2(CH2l)(I)] ( 1 8 ) ...81

17. FA B M ass Spectra for the O xidative Isom erization o f 18 to 1 9 ...85

18. ORTEP o f [Ir(PPh3)(CO )0i-pz)H ]2 ( 2 3 ) ... 117

19. O RTEP o f [Ir2(P P h 3)2 (C O )2( it -p z )2HCl] ( 2 6 ) ... 119

XU

21. ORTEP o f [Ir2(CO)2(PPh3)2(li-p z)2HI] ( 2 8 ) ... 123 22. ORTEP o f [Ir2(CO)2(PPh3)2(ll-pz)2H](BF4) ( 2 4 ) ...125 23. ! H NM R D ata S h ow in g the Isom erization o f 24 to 2 5 ...165

[Ir(CO D)(|i-pz)]2 (2) [Ir(COD)(|i-M epz)]2 (2a) [Ir(COD)(n-35DM epz)]2 (2b) [Ir(COD)([i-345TM epz)]2 (2c) [Ir(CO)2(fi-Pz)]2 (3)

[Ir(CO)2(fi-M epz)]2 (3a) [Ir(CO)2(|i-35D M epz)]2 (3b) [Ir(CO)2(p.-345TMepz)]2 (3c) [Ir2(CO)4(p.-pz)2(CH3)I] (4) [Ir2(CO)4(|J.-pz)2(CH3)Br] (5) [Ir2(CO)4 (|4-pz)2(CH3CH2)IJ (6) [Ir2(CO)4(ji-pz)2((CH3)2CH)l] (7) [Ir2(CO)4((i-pz)2(CH3(CH2)4)I] (8) [Ir2(CO)4(|i-pz)2(CH3(CH2)7)I] (9) [Ir2(CO)4(p-p z)2(CH3(CH2) i 7)I] (10) [Ir(CO)2(|i-pz)I]2 (11) [Ir2(CO )4(|4-pz)2(CH3)] (PF6) (12) [Ir2(PPh3)2(CO)2(p.-pz)2(CH)3(I)] (13) [Ir2(PPh3)2(CQ )2Qa-pz)2(C H (C H )3)(I)] (14) [Ir(PPh3)(C O )(|i-pz)I]2 (15)

[Ir(PPh3) (13CO)(,u-pz)I]2 (15a) [Ir2(CO)4((i-pz)2iCH 2I)I] (16)

xiv

[Ir2(CO)4(H-pz)2( ^ C H 2)(l)2] (17) [Ir2(PPh3)2(CO)2(|X-pz)2(CH2I)(I)] (18) [Ir2(PPh3-d i5)2(CO)2(p-pz)2(CH2I)(I)] (18a) [Ir2(PPh3)2( 13CO)2()i-pz)2(CD2l)(I)] (18b) [Ir2(PPh3)2(13CO)2(p-pz)2(CH2I)(I)] (18c) [Ir2(PPh3)2(CO )2()i-pz)2(CD2l)(I)] (18d) [Ir2(PPh3)2(CO )2( ^ p z )2()i-CH2)(I)2] (19) [Ir2(PPh3-di5)2(CO)2(p-pz)2()i-CH2)(I)2] (19a) [Ir2(PPh3)2(13CO)2ai-pz)2()i-C H 2)(I)2](19b) [Ir2(PPh3)2(13CO )2(p-pz)2(p-CH 2)(I)2] (19c) [Ir2(PPh3)2(CO )2()i-pz)2(p-CH2)(I)2] (1 9 d )

[Ir2<;PPh3)2(12CO )(13C O )(p-pz)2()i-C H 2)(I)2](19e) [Ir2(PPh3)2(12C 0)(13C 0)(p-pz)2(p-C D 2)(I)2] (19f) [Ir2(PPh3)2(C O )2()i-pz)2(C H 2C H 2l)I](21)

[Ir2(PPh3)2(CO)2()X-pz)2(CH2Br)Br](22) [Ir(PPh3)(C O )(p-pz)H ]2 (23) [Ir2(PPh3)2(CO )2()i-pz)2H ](B F4)(24) [Ir2(PPh3)2(CO)2(|l-pz)2(O H)(H )2](BF4) (25) [Ir2(PPh3)2(CO)2(p-pz)2HCll (26) [Ir(CO)2(|i-pz)C l]2 (27)

[ li2(PPh3)2(CO )2(p-pz)2HI] (28)

[Ir2(PPh3)2(CO)2(p-pz)2(O M e)(H )2](B F4) (29) [Ir2(PPh3)2(CO )2(p-pz)2(SH )(H )2)(BF4) (30) [Ir2(PPh3)2(CO)2()l-pz)2(H)(I)2l(BF4) (31) [Ir(PPh3)(C O )(|i-pz)(H )(C F3CO O )]2 (33)

br broad Bu Butyl COD 1,5 cyclooctadiene d doublet dppm bis(diphenylphosphino)methanc dicp 1,3-diisocyanopropane Et ethyl

FAB Fast Atom Bombardment

m multiplet

bridging

Me methyl

NM R Nuclear Magnetic Resonance ORTEP Oak Ridge Thermal Ellipsoid Plot

Ph phenyl pzH pyrazole q quartet R alkyl group s singlet sh shoulder St strong t triplet, tertiary THF tetrahydrofuran TMP trimethylphosphite TMS tetramethylsilane w weak X halide

3Mepz 3-methyl pyrazolyl 35DMepz 3,5-dimethyl pyrazolyl 345TMepz 3,4,5-trimethyl pyrazolyl

I. T IIE M E T A L -M E T A L B O N D IN O X ID A T IV E A D D IT IO N S

Com m ercial catalytic processes such as hydrogenation, hydroform ylation, and polymerization are among the recent achievem ents o f the application o f the principles o f organometallic chemistry. Common to all o f these processes is the formation and cleavage o f m etal to ligand bonds in a cyclica l fash ion, particularly those bonds in volvin g hydrocarbon fragments. During the last three decades, the principles o f metal-hydrocarbon bonding have becom e sufficiently well understood to m ake many qualitative and som e quantitative predictions concerning the behavior o f m ononuclear com plexes o f this type. H owever, similar chemistry taking place across adjacent metal centers that are capable o f interaction is not fully understood and is the locus o f considerable current research.

M ultiply-bonded binuclear species without bridging ligands are presently w ell- described only for the electron-poor transition m etals early in the d-block, such as m olybdenum .1 A ligand bridging two metal centers, however, will allow a large variation in metal-metal distances and oxidation states for a variety o f transition elements. The choice o f a bridging ligand in the design o f a bimetallic com plex is critical to the chemistry o f the com p lex;2 the ligand must be resistant to bridge-cleavage reactions that w ou ld lead to a mononuclear complex, it must allow metal-metal interactions, and must be itse lf resistant to attack. W hile m any ligands w ill bridge m etals, few w ill fulfill all three criteria; halo- bridged com plexes are easily cleaved by Lew is bases,3 sulfides are them selves reactive,4 and phosphido-bridges tend to undergo further bridging reactions.5

The ch oice o f a bridging ligand w ill predispose a dinuclear system to a certain geometry. For example, the two planes o f a double square d^-d^ bimetallic com plex can be held in a face-to-face orientation, a side to side orientation, or at som e dihedral angle.2 Som e bis-exobidentate ligands can be cis or trans on the metal centers so that a variety o f b r id g in g g e o m e t r ie s a re p o s s i b l e . B i n u c l e a r c o m p l e x e s u s in g

2-(diphenylphosphino)pyridine (dppp) have been used to hold m etal atoms in a face-to- face or side-to-side orientation and have received considerable attention in recent years; in particular, dppm has been found to bridge metal atoms with various coordination numbers and oxidation sta tes.6 Schm idbauer has investigated the chem istry o f the digold bis(phosphido)m ethane, which add halogens and alkyl halides under ambient conditions.7 G ray et al pursued a sim ilar chem istry for the fa ce -to -fa ce dirhodium com plex [tetra(|i-diisocyanopropyl)dirhodium (I)] with particular attention to the photoelcctron spectroscopy related to orbital hybridization energies.** Poilblanc et al synthesized and characterized a thiolato-bridged diiridium complex, where the sulfide bridges act as a hinge; typically the com p lex had a square-plane dihedral angle (a ) o f 133° (M-M distance = 3 .2 1 6 A ).9 Reactant access to the intermetallic region m ay be influenced by the bridging ligand in a face-to-face or side-to-side bimetallic complex, but a hinged arrangement allows better approach o f reactants inside the intermetallic region and a greater variation in metal- m etal distances as a consequence o f variability in a . A disadvantage o f this system is the reactivity o f the sulfide bridge; how ever, a choice o f bulky substituents ( Ph, t -B u , 2 ,4 ,6 -trim eth y lp h en y l) on the sulfide has produced a series o f binuclear com plexes with reasonable core stability.

d) hinged

Figure 1. B im e ta llic C om pou n ds F o rm ed fro m T e tra c o o r d in a te d P la n a r M e ta ls o r Pentacoordinated M etals2

m etal separation and oxidation state, is inert to m ost oxidants and docs not cause problematic solubility problems. The bis-bridged binuclear unit demonstrates great core stability to m ost reagents. The size and geometry o f the pyrazolyl ligand only very rarely allow s chelation11 so that it is normally considered to be an bridging ligand. The two-atom linkage produced by bridging bis-pyrazolyl ligands does not preclude m etal-m etal interactions as o b serv ed in sin gle atom linkages; indeed, the structurally sim ilar [Ir(PMe3)(C O )(|i-tB uS) ] 2 which is bridged by a single center has a metal-metal distance of 3.22 A ,12 w h ile the metal-metal distance o f [Ir(CO)(PPh3)(p.-pz) |2 (1) which is bridged by pyrazolyl is actually closer at 3.16 A .13

There are a num ber o f synthetic routes to bridging pyrazolyl com p lexes. For e x a m p le, d irect reaction o f sodium pyrazolide with Vaska's co m p lex generates [Ir(CO)(PPh3)((i-p z) ] 2 (1) in reasonable y ield .14 A more versatile synthetic method is the synthesis o f the pyrazolyl-bridged diiridium cyclooctadiene com plex [Ir(COD)(jj.-pz) I2 (2) by cleavage o f the chloride-bridged com plex with pyrazole follow ed by addition o f base. Substitution o f different terminal ligands for cyclooctadiene allows examination o f a variety o f electronic and steric effects on oxidative addition to the bimetallic complex:

5

The d7 configuration o f mononuclear iridium com p lexes has alm ost no genuine e x a m p les.15 H ow ever, i f a metal center with one unpaired spin is in c lo se proxim ity to another center, formation o f molecular orbitals w ill generate a lower energy bonding orbital and a higher energy antibonding orbital. Thus two available electrons (form ally on e from each center) w ill spin-pair in the bonding orbital and result in an increase in bond order. For example, hybridization o f two dz2, pz orbitals in a d7- d 7 bim etallic system (tw o D4h centers side by side) w ill result in two sets o f orbitals; one electron oxidations from d8 at

influences and geometry do not prohibit such spin-pairing stabilization.16 This is easily observed experim entally, sin ce spin pairing results in a diam agnetic, rather than a paramagnetic com plex. The bimetallic com plex is formally o f d7 configuration and resists the further oxidation to the d6 state that would be expected in a mononuclear com plex.17

A w e ll-e sta b lish e d concept in organom etallic chem istry is that o f oxidative addition, where a group, A - B , adds to and oxidizes a single metal center; the metal thus behaves sim ultaneously as a Lew is acid and as a Lew is b a se.18 For all such reactions, both the oxidation state and the coordination number o f the metal increase, so that the oxid a tiv e addition results in a formal configuration change from d n to dn"2 and a coordination number increase from n to n+2. Oxidative additions can proceed by a variety o f m echanism s that depend on the addenda, the metal, the ligands, or the solvent system. T w o electron oxidative additions are known for virtually all dn (n = even) com plexes; how ever, they are far more prevalent with the electron-rich elem ents to the right o f the transition series, the m ost notable being iron(O), ruthenium(O), osmium(O), rhodium(l), iridium (I), nickel(O), palladium (II),(0), and platinum (ll),(0). Higher oxidation states are usually more stable for the heavier metals than for the lighter; thus Rh(III) is generally less stable than Ir(III). In general, factors that tend to increase the availability o f electrons on the metal tend to increase the oxidizability.19 Steric properties o f the ligand must also be considered since bulky ligands tend to decrease the ease o f oxidative addition by shielding the metal and distorting the surrounding ligands.

Figure 2. M olecular O rbital D iagram f o r a H om obim etallic d 7 - d 7 Species assum ing D4ft or 0 ,1(1 sym m etry 16

orbitals in a bimetallic com plex o f appropriate symmetry may allow m etal-m etal bonding, thus permitting the bim etallic species to accommodate unconventional formal oxidation states. Reaction o f a substrate A-B to a bridged bimetallic species w ill result in either cleavage o f a metal-metal bond with a one-electron oxidation o f each metal:

t o

M n + AB --- ► A — h

formation o f a metal-metal bond with a one-electron oxidation o f each metal:

9

formation o f a cationic complex with a terminal or bridging ligand:

M n + AB

>* A —

h

r

B “

or formation o f a dicationic complex:

t o

M n + 2AB --- ► •

2AB"

where the last product is actually the result o f an oxidation, rather than an oxidative addition. Direct evid en ce was recently reported that supports a mechanism whereby initial attack at one metal center (one two-electron oxidation) is follow ed by migration o f one o f the incom ing ligands to the other metal center (tw o one-electron oxidations).20 Thus, the final products o f oxidative addition do not necessarily im ply a m echanistic pathway, and steric and lig.md electronic effects can influence the chemistry involved. Although a given m echanism m ay be co n sisten t with the observed products, radical p ro cesses or rearrangements are common and other mechanisms cannot be ruled out.

T he reaction o f a metal com plex with a substrate m ay proceed by m ore than one m echanism . D ep en d in g on the conditions o f the reaction, there are six principle possibilities: (a) purely ionic m echanism s (usually observed with ionizable m olecules in polar solvents), (b) concerted three or four center additions (usually for substrates with low polarity), (c) Sjsi2-type reactior" where the m etal center behaves as the nucleophile

p rocesses, (e) radical non-chain reactions (where generation o f the radical for each molecular reaction is required) and (f) template reactions.

This chapter provides the basis for discussion concerning additions to binuclcar com plexes, including oxidative additions, special properties o f bimetallic com plexes, and choice o f a bridging ligand. Chapter 2 begins with an extension o f the recent work o f Harrison and Fjeldsted on the mechanism o f oxidative addition o f alkyl iodide to pyrazolyl- bridged diiridium com plexes. The results o f a kinetic study o f the addition o f alkyl halide to [Ir(CO)2(|i-p z) ] 2 (3) is presented and discussed. Chapter 3 describes the chem istry, crystallography, and the kinetics o f addition o f dihalogenoalkyl to [Ir(CO)(PPh3)(p.-pz) |2 (1), and Chapter 4 considers the character o f hydrogen-iridium bonds, particularly with respect to the site trans effect. Chapter 5 outlines the major conclusions, while Chapter 6 is the Experimental section, in w hich full details o f each experiment are given. Com plete spectroscopic and analytical data are included for the convenience o f future workers in this area.

1 1

II. S T E R IC A N D E L E C T R O N IC E F F E C T S IN A L K Y L H A L ID E A D D IT IO N

I n tr o d u c tio n

O xidative addition o f alkyl halides (RX; R = alkyl, X = halide) to platinum-group metal com plexes has been the subject o f particular interest over the past three decades for several reasons, including the importance o f these reactions to the synthesis o f transition- metal alkyls,21 couplin g reactions involving transition-metal reagents,22 and catalytic processes.23 The use o f transition metal com plexes for the activation o f carbon bonds by reaction with alkyl halides is particularly relevant to the last o f these ideas because o f the potential applications to hydrocarbon rearrangements; for exam ple, the Monsanto Process is a commercial application o f R X addition to a Group 8 metal com plex that leads to the production o f acetic acid from methanol and carbon m onoxide. A key step in the process is the addition o f methyl iodide to a rhodium(I) dicarbonyl diiodide anion, which is follow ed by carbonyl insertion, carbon m onoxide addition, and reductive elim ination o f acetyl iodide. The reaction is zero order with respect to methanol and carbon m onoxide and first order with respect to the rhodium catalyst and iodide promoter 24 The kinetic profile o f this m ononuclear rhodium system has been thoroughly exam ined because o f its industrial significance, but it has been found that the kinetics o f alkyl halide addition to other mono- and polynuclear com pounds often differs from this system or is uncertain 25 For example, different sets o f experimental data for the addition o f RX to Group 9 compounds has led to two very different postulated m echanism s.15

Three different m echanism s are com m only addressed in discu ssions o f the oxidative addition o f RX to low valent G roup 8 transition metal com plexes.26 The earliest k in e tic s tu d ie s o f the a d d ition o f m eth yl io d id e and b e n z y lic h alid es to

consistent with an Sn2 nucleophilic displacement o f halide from carbon by an electron-rich m etal center. T hese studies suggested that the m echanism o f addition to m ononuclear com plexes proceeds as follow s, based on activation parameters, solvent coordination, and substituent influences for substituted benzyl halides and methyl iodide:28

This m echanism w as also consistent with certain stereochemical features o f the reactions such as inversion at the carbon center 26 An alternative interpretation proposed that metal insertion into the carbon-halogen bond led to a three-centered (concerted) transition state, where the C-X a molecular orbital rehybridizes with a vacant metal orbital. Electron back- donation from a filled dXz or dyz metal orbital into the C-X a * orbital would then lead to bond scission and reformation:29

M + R ,C X ₀ --- ► M , .c r3 * X M / CR3 •X

Limited evidence for this pathway exists for mononuclear transition-metal com p lexes.30 A third proposed pathway in volves hom olytic carbon-halogen bond cleavage leading to interm ediary carbon radicals, in either a chain or non-chain process; for exam ple, stereochem ical and kinetic data indicated that the addition o f certain non-saturated alkyl

13

halides to trans -Ir(PMe3)2(CO)Cl (the PM e3 analogue o f Vaska's Com plex) proceeded as a radical-chain o f the form:31

M + R X - 4 M +*RX- - » MX* + R‘ -> M (X )(R)

It is generally observed that the kinetics o f alkyl halide additions to mononuclear com p lexes can be explained by one o f the three m echanism s described above under appropriate conditions; by contrast, no sim ple m odels have been found to consistently account for oxidative addition behavior o f polynuclear centers. For exam ple, enhanced alkyl halide reactivity towards polynuclear co m p lex es com pared to corresponding mononuclear com plexes has been observed and explained by anchimeric effects between tw o metal centers.32 Alternately, a decrease in the reactivity o f binuclear system s with alkyl halides has also been observed and has been rationalized by the increased steric hindrance present in these system s.32 In an effort to reconcile the apparently contradictory e v id en ce, it has been show n that both radical chain and Sn2 m ech a n ism s e x ist sim ultaneously in som e binuclear system s and that minor electronic or steric changes will have a considerable effect on the dominant mechanism o f alkyl halide oxidative addition.32

This chapter exam ines the iridium(I) to iridium(II) transformation resulting from oxidative addition o f saturated alkyl halides to [Ir(CO)2( |i- p z) ] 2 (3 ) in terms o f the observed chem istry, m olecular structure, and kinetics; with this evid en ce, an attempt is m ade to distinguish the effects o f metal-metal distance, electron distributions, and steric factors on com peting mechanisms.

R e s u l t s

Form ation o f alkyl-halide adducts with the binuclear co m p lex 3 w as entered through the d ikidiu m (I) com p lex [Ir(C O D )(|i-pz) ] 2 (2). T yp ically, 2 (or a pyrazolyl

deep red solution w hich was exposed to less than one atmosphere o f carbon monoxide in 2 to 1000 fold e x c e ss o f stoichiometric requirements. The flask w as sealed and stirred, with the carbonylation being accom panied by a change in the initial red colour o f the [Ir (C O D )(p -p z) ] 2 (2 ) fam ily o f compounds to an iridescent yellow in all cases. This required ca 2 m in for the unsubstituted pyrazolyl com plex (Table 1), but up to 2 d for the corresponding 3,4,5-trim ethylpyrazolyl com plex, where the reaction rate was found to be inversely related (qualitatively) to the number o f methyl-groups on the pyrazolyl bridge.

15

C o m p o u n d P r o d u c t [C O ] R e a c t io n

T im e [Ir(COD)(p-pz) ] 2 (2)

[Ir(COD)(p-3-M epz) ] 2 (2a) [Ir(COD)(p-35DM epz) ] 2 (2b)

[Ir(COD)(|i-345TM epz) ] 2 (2c)

[Ir(CO)2(p-pz)]2 (3) [Ir(CO)2((i-3-M epz) ] 2 (3a) [Ir(CO)2(p-35D M epz) ] 2 (3b) [Ir(CO)2(|i-345T M epz)]2(3c)

1 atm 1 atm 1 atm 1 atm 1 2 0 s 18 h 4 0 h (est'd) 48 h

follow ed by redissolution in dichloromethane and recrystallization by dropwisc addition o f hexanes to g iv e a yield between 85-95% . W hile all solutions o f the diiridium tctracarbonyl com plex were y ello w , the solids resulting from the carbonylation o f 2 and 2a were blue- black33, and the solids resulting from the carbonylation o f 2b and 2c were orange.

Synthesis o f the alkyl halide adducts w as carried out by dissolving the appropriate bispyrazolyl(tetracarbonyl)diiridium(I) com pound in THF, follow ed by dropwise addition o f an excess o f alkyl halide (from 10% excess for C18H3 7I to 100 fold ex cess for CI-I3I) to this solution. The reaction vessel was shielded with aluminum foil to exclude ambient light throughout the synthesis. The reaction mixture was stirred until an orange colour was ea sily discem able, and the solvent was rem oved in vacuo to 3 mL. H exane was layered onto the solution and the mixture was cooled to -35° C for 2-3 d. Y ellow (or orange) solids were collected in moderate y ield (40-80% ), any losses being due to the high solubility o f the alkyl iodide product. Separation o f the unreacted octadecyl iodide and octyl iodide from the iridium adduct w as difficult because o f the high solubility o f both the substrate and the adduct; in those cases, purification was incom plete, but NM R spectroscopy was used to establish that less than 5% ex cess alkyl halide remained in admixture with the longcr- chain alkyl halide adducts.

R eactions and yield s o f [Ir(CO !2( |i- p z) ] 2 (3) with m ethyl iodide (to give 4 ), m ethyl brom ide (to g iv e 5), ethyl iodid e (to g iv e 6), propyl iodid e (to give 7 ), pentyl iodide (to give 8), octy l iodide (to g iv e 9), and octadecyl iodide (to give 10) have been previou sly describ ed. 34 A ll com pounds were characterized by using NM R and IR spectroscopy through com parisons w ith previous work by Harrison. T he reactions o f m ethyl io d id e with [Ir(CO)2(M--3M e p z) ] 2 (3 a ), [Ir(CO )2(|i.-3,5D M e p z) ] 2 (3 b ), and [Ir(C O )2(|i-T M e p z) ] 2 (3 c) are also included in the Experim ental section. Carbonyl- stretching frequencies in the IR show ed a decrease in frequency with an increase in the

17

degree o f m ethyl substitution o f the pyrazolyl bridge. D ifferences betw een the N M R resonances o f the pyrazolyl 3 ,4 ,5 protons or the proton resonances o f the 3,5 m ethyl groups indicated that all protons were chem ically inequivalent, and the com p lex w as therefore o f lo w symmetry.

In 1987 Fjeldsted demonstrated that the reaction o f methyl iodide with the iridium(I) com plex [Ir(COD)(|!-pz)]2 (2) results in an equilibrium constant for the forward reaction o f the order o f 7 .4 L m ol'1, i.e. not strongly favouring oxidative addition.30 This observation is in marked contrast to the [Ir(CO)2(H-pz) ] 2 (3) system , where it is found that iridium(II) alkyl halide adducts o f 3 and its bridge-substituted congeners are very stable and resistant to reductive elimination. For example, when stoichiometric amounts o f methyl iodide and 3 w ere sealed in an N M R tube and allow ed to react for 2 4 h, the reaction proceeded to completion with no trace o f starting diiridium com plex. Dissolution o f the adduct 4 in THF or dichloromethane does not result in reductive elimination o f the alkyl halide com plex after 48 h, demonstrating that the equilibrium between the iridium(I) and iridium(II) species lies w ell towards the latter.

The phosphine-substituted diiridium(I) com plex [Ir(CO)(PPh3)(|i-p z) ] 2 (1) show s sim ilar reactivity to that o f the [Ir(CO)2(|i-p z) ] 2 (3) system . Stoichiom etric addition o f propyl iodide to 1 in cfe -dichlorom ethane occurs within 4 h to g iv e the diiridium (II) c o m p le x [Ir2(C O )2(P P h3)2( |i- p z )2(C H (C H3)2l] (1 4 ), w hich strongly favours the diiridium(II) com plex with no remaining 1 observed. The contrast betw een the oxidative addition behavior o f 1 and a sim ilar m ononuclear co m p lex is dem onstrated by an experim ent in which a 5:1 stoichiom etric ratio o f iso-propyl iodide to Vaska's com p lex (trans -Ir(CO)(PPh3)2Cl) in dichloromethane is sealed in an N M R tube and left for 10 d in the dark. A t the end o f this tim e, a N M R spectrum indicated that no change had occurred. Addition o f two equivalents o f methyl iodide to the same N M R sam ple resulted

2 d.

The therm al and p h oto ch em ica l stab ility o f the a lk y l-h a lid c adducts to [Ir(CO)2(|i-p z) ] 2 (3) w as exam ined by dissolving samples o f each adduct in d8-THF and sealing the solutions in an NM R tube (77K under vacuum) in the absence o f light. The h-I NM R spectrum was recorded after warming the sealed tubes to room temperature, then the tubes were heated in an o il bath at 140° C for 36 h; h i NM R spectroscopy indicated that little decom position had occurred after the heating cycle. The tubes were then exposed at close range (25 cm) to a high intensity mercury vapour light source for 10 min and another set o f spectra were recorded. It was found that the irradiation had decom posed the longer chain alkyls halide adducts in a manner consistent with a relationship between the extent o f d e c o m p o s itio n and the a lk y l ch a in le n g th , from 5% d e c o m p o sitio n for [ I r2( C O )4 ( | i - p z )2 l ( ( C H 2 )7C H 3)] ( 9 ) to 95% d e c o m p o s it io n fo r [Ir2(C O )4( |i- p z )2l(( C H2) i7C H3)] ( 1 0 ) . The decom position o f 10 was not pursued further, but other com pounds were exp osed to U V light for extended periods (5 h or greater) and a final set o f spectra recorded. Black insoluble material was produced during the decom position that contributed to the loss o f N M R resolution, and accordingly it was difficult to use N M R to specifically identify the decom position products. H owever, the irradiation o f the pentyl-iodide addu a for 5 h produced [Ir(CO)2(p.-pz)I]2 (11) in 45% yield and hydrocarbons, o f w hich less than 5 mol% o f p en t-l-en e was identified by 'H NM R: 8 5 .9 m (lH ), 8 4.95m (2H ), 82.05q(2H ), 61.5m (2H ), 80.9m (3H ); m ost resonances attributable to hydrocarbon fragments were observed between 82.0 and 80.7 and were not assigned to a particular formula, although the chem ical shift and splitting patterns suggest an alkane product. A M S analysis o f the decom position product m ixture o f the adduct [Ir2(CO )4(lt-pz)2l((C H3)2CH)] after 6 h o f U V light show ed an m /e o f 41 (C3H5), but this

19

did not correlate with ]H NM R spectrum. Observations o f these experim ents are recorded in Table 2.

The ability o f the 1 and 3a - d metal centers to attain the +11 oxidation state is a consequence o f the potential o f the tw o metal centers to form a^z2 bonding and c * d z2 antibonding orbitals, where electron occupancy o f the bonding orbital w ill lead to an increase in the bond order, while electron occupancy o f the anti-bonding orbital w ill lead to

Compound Observations

140° C, 36 h U V light, 10 min U V light,6 h [Ir2(CC))4(|i-pz)2l(C H (C H3)2)] (7)

[Ir2(CO)4( ^ p z )2((CH2)4CH3)I] (8) [Ir2(CO)4(|i-pz)2((CH2)7CH3)I] (9) [Ir2(CO)4(U-pz)2((CH2) i7CH3)I] (10)

no change no change no change no change no change no change slight decom p, decom posed decom posed decom posed decom posed

Table 2. Decomposition o f Alkyl Halide Adducts to [Ir(CO )2(p-pz)]2 (3)

a decrease in the bond order.16 The relationship between bond order and electronic

configuration focuses special interest on x-ray structure determination for the methyl iodide

adduct to pyrazolyl-bridged diiridium compounds, but the tendency of the adducts to either

co-crystallize with solvent molecules or to precipitate from solution as a powder have

frustrated such efforts in the past.36 Fortunately, it has since been possible to crystallize the

iso-propyl iodide adduct: [Ir(CO)4(ji-pz)2(CH(CH3)2)I]2 (7) was synthesized from

[Ir(CO)

((i-pz

and a 50-fold excess of wo-propyl iodide in

and small (average

saturated dichlorom ethane-hexane mixture at -35° C with the exclusion o f light over two w eeks. The crystals obtained were o f generally poor quality, but a plate o f dim ensions 0 . 2 1 x 0 . 1 2 x 0 .0 1 m m3 was chosen by inspection under polarized light.

X -ray data for 7 w ere c o lle cte d on a N o n iu s-E n ra f C A D 4F diffractom eter controlled by a P D P -1 1 M icro, running the C A D 4 set o f programs. A m onoclinic unit cell w a s determ ined and indexed from 2 0 centered reflection s co llected at random, with 2 0 > 3 O ° , M o K a radiation, to g iv e a unit cell o f dim ensions a = 10.147(5)

A,

A,

A,

Inspection o f the collected data for systematic absences established the ''pace group as P 2i/n . D ensity measurements show ed that Z=4, or 1 m olecule per asymmetric unit. The data were corrected for Lorentz effects, polarization effects, and for decay (less than 8%) follow ing an empirical absorbance correction (p. = 24.16 c m '1). The program SH ELX-7637 w as used for structure solution. Direct m ethods were used to locate the two iridium centers and the iodide; remaining atoms were located by normal w eighted Fourier techniques. The h ea v y atom s w ere refined anisotropically and the rem aining atom s w ere refined isotropically because o f the poor quality (low intensity) data and slight disorder o f the propyl carbons; 1286 reflections were used in the full matrix refinement o f 249 parameters to give a conventional residuals (R) o f 0.1075, with w eighted residuals (Rw) at 0.0954.

A number o f structural features o f 7 could be discerned despite the low -quality crystal that was used in the analysis. The m ost significant o f these was the axial orientation o f the alkyl group and the iodide, and the clo se approach o f the tw o iridium centers

21

(2.739(8)

A)

The m etal-pyrazolyl nitrogen bonds cis to the iodide were sim ilar within error to the metal-nitrogen bonds cis to the alkyl group.The Ir2-I bond length was as expected for an iridium -iodide bond at 2 .7 8 9 (1 0 )

A,

A

A.

23

K in etics o f M ethyl Iodide A ddition to [Ir(C O )2(p z) ] 2 (3)

The kinetic experim ents were approached through U V /v is spectroscopy; initial experiments concerning alkyl halide addition to pyrazolyl-bridged tetracarbonyl diiridium com p ou n d s were con d u cted to id en tify iso sb e stic p oin ts. A T IIF so lu tio n o f [Ir(CO)2(p.-pz) ]2 (3) was placed in a 3 mL quartz cuvette under a nitrogen atmosphere, and an initial spectrum was recorded against a THF blank and plotted for scaling purposes. T he spectrum was then re-recorded after 5 min to confirm that 3 had not reacted. It was a lso established independently that the alkyl halides were not an asorbefacient species in the U V /vis range. A U V /vis cell temperature controller w as used to maintain a constant reaction temperature; for pre-scan experim ents, this was set at 15° C. Pre-distilled alkyl iodide (in 1 0 0 0-fold or greater excess o f stoichiom etric requirements) was then quickly delivered by micropipettor to the cuvette and the timed wavelength scans were started as soon as possible after placing the cuvette in the cell compartment. Trial-and-error was used to determine the optimum delay required for a non-overlapping series o f spectra. Interval times ranging from 15s to 15min were chosen and at the end o f the experiment all data for a particular wavelength interval were overlaid onto a single graph.

The effect o f ambient light on the kinetics o f the reaction w as investigated by exposing the cuvette to laboratory fluorescent lighting for timed periods. Alternatively, the cuvette w as exposed to an imm ersion mercury vapour ultraviolet lamp for fifteen to sixty seconds at 25 cm from the source. This exposure was repeated until there was no change in the absorption spectrum o f the solution. An IR spectrum o f the final product w as recorded to confirm the presence o f the alkyl halide adduct rather than a dihalogeno-adduct; for ex a m p le, after ex ten d ed U V irradiation o f so lu tio n s o f m ethyl brom ide and [Ir(CO)2(|i-p z)]2 ( 3 ) , the IR spectrum was observed to exhibit absorptions at 2 1 0 9 c m '1, 2 0 8 0 c m '1, and 2 0 5 0 cm' 1 (c.f. reference 34). The absorptions corresponding to the

w ere not observed.

T h e o x id a tiv e addition reactions o f the binuclear iridium (I) co m p lex es [Ir(C O )2( |i- p z) ] 2 (3 ) and [Ir(C O )(PPhi''(ii-pz) ] 2 (1) with several alkyl halides were exam ined. Equim olar quantities o f either iodom ethane, bromomethanc, iodoetlmnc, or 2-iodopropane were added to a THF solution o f 3 and the U V /v is spectrum recorded at tim ed intervals, (see Figure 4 , page 36, for 3 + RI, R=M e, iPr, Et). In a similar way, tim ed interval spectra were recorded for iodomethane addition to [lr(CO)2(p -p z) ]2 (3), [Ir(CO)2(|!-3-M epz) ] 2 (3a), [Ir(CO)2(M--3 5DM epz)]2 (3b ), and [Ir(CO)2(p-TM epz) |2 (3c) in THF under both dark conditions and on exposure to fluorescent light.

M ethyl iodide addition to [Ir(CO)2([i.-pz)]2 (3) demonstrated unusual behavior; the first set o f spectrophotometric plots at 3 min intervals determined a well-defined isosbcstic point at 393 nm with a local maximum at 412 nm. H ow ever, after 20 min the isosbcstic point lost its definition and at the same time a second isosbestic at 405 nm appeared which persisted until the reaction was complete. This reaction profile was completely reproducible and was not affected by exposure to ambient light. Reactions with longer chain alkyl halides did not show sim ilar kinetics; for exam ple, the reaction with ethyl iodide only showed a single isosbestic point (393 nm) which did not persist after 4 min (see Figure 4, page 36). The reaction o f [Ir(CO)2(|i-p z) ] 2 (3) with /.vo-propyl iodide was too fast to monitor using these techniques, and after 3 min the reaction was complete. The reaction o f methyl iodide and 3 in a non polar solvent such as benzene show ed two isosbestic points at 407 nm and 430 nm (local maximum at 413 nm). R eaction rates were similar to those observed in THF, but changes in the onset and position o f the isosbestic points suggests some solvent dependence o f the meclianism. The reaction o f methyl bromide with i under dark conditions w as very slow , with a half-life on the order o f 1 0 0 min, but exposure to a mercury vapour U V light for 15s intervals, 25 cm from source, increased the rate

25

dram atically and three isosbestic points were observed (3 2 0 nm, 3 3 0 nm, and 353 nm). The presence o f the bromomethane adduct [Ir2(CO)4(p -p z)2(CH3)Br]2 (S) w as confirmed by IR spectroscopy and no [Ir(C O )?(p-pz)B r] 2 was detected, based on the expected position o f the carbonyi stretching frequencies for this com plex. Exposure o f the reaction o f methyl iodide and [ir(CO)2(p -p z )h (3) to ambient light betw een U V /vis scans did not alter the kinetics o f the addition.

R eactions o f ethyl iodid e and propyl iodide w ith [Ir(CO )2(p -p z) ] 2 (3) were monitored by both sequential wavelength scans and time scans; it w as found that the rate o f reaction increased w ith increasing alkyl chain length, so that m ultiple wavelength scans w ere difficult to record before the reaction w as over and isosb estic points could not be observed in either addition. The relative reaction rates o f m ethyl vs ethyl v s isopropyl iodide addition are estimated to be 1:4:10 at 288 K.

T h e w a v elen g th scan observations o f the m ethyl io d id e reaction w ith [Ir(CO)2(p.-pz)]2 (3 ) in THF indicated that further inform ation concerning the reaction would be desirable, and so a complete kinetic study was undertaken to describe the reaction quantitatively. The kinetics o f the same reaction in benzene w ere also investigated. An U V /vis study o f the rate dependence on temperature was com p leted under pseudo-first order conditions w ith respect to methyl iodide in both THF or benzene. A low initial concentration o f 3 (0.18 rnM) was used because o f the high absorption coefficients o f 3 in the U V region (e on the order o f 3000 at the local m axim um ), but the reaction was essentially com plete within 3 0 min when methyl iodide concentrations o f 0.5 M to 1.0 M were used. The spectrum o f the solution o f 3 was checked for changes in absorbance over 10 min at 10° C; w ith no evidence for sample degradation, the room was darkened and the reaction w as initiated by rapid injection o f the required aliquot o f methyl iodide. For the reaction in THF, absorbance was measured at each o f the two isosbestic points in order that the kinetic data pertaining to each o f the points w ould be m easured independently o f the

second part was measured at 393 nm. T he results o f each o f these scans are illustrated in Figure 5, page 37; the first step follow ed first order kinetics for 5 to 6 min after initiation, and while the second step started with tailing that corresponds to the com pletion o f the first reaction, the remainder o f the second curve was m odeled by first order kinetics. The flat part o f th e absorbance v e r s u s tim e plots confirm ed the iso sb estic character at the w avelengths used in the experiments. The first order rate constant k0bs w as determined using the Lambda 4 B software, where the value o f

w as calculated and plotted against tim e at 1 s intervals, the average slo p e o f which corresponded to k 0bs • The program results were checked by manual calculation o f data from selected absorbance versus time plots. The rate constant k 0bs was exam ined as a function o f [Mef] at both isosbestic points, the concentration o f methyl iodide being varied from 0 .0 5 M to 1 M; the first step (monitored at 405 nm) was dependent on methyl iodide in a first-order relationship, with k2 = 0.031 M ' V1 with a correlation coefficien t (r2) o f 0.9985 and a [0,0] intercept (within error). The second step, however, was independent o f methyl iodide concentration ( r2 = 0 .9 9 9 8 ), both in terms o f a near-zero slop e and a non­ origin intercept, where a k j o f 0.0047 s' 1 was determined. Therefore, the first step o f the reaction w as overall second order in methyl iodide and [Ir(CO)2(p.-pz) ]2 (3) (first order in eash), w h ile the second half w as only dependent on 3 concentration. The sam e reaction in benzene was conducted at 4 1 2 nm, the local maximum; it was found that the first-order plots for the benzene experiment were not linear, and follow ed a second-order relationship in both 3 and m ethyl iodide, resulting in an overall fourth-order reaction.

27

A ctivation parameters were determined experim entally by altering the reaction temperature in 5° C increm ents and calculating k.2 , where k2 • T he ra tio s- j - v s ^ were graphed and AH* w as determined from the slope in accordance with the Eyring equation. The ivation entropy AS* was determined from extrapolation o f the graph to the k2/ T intercept. T he data for the first step o f the reaction o f m ethyl io d id e with [Ir(CO)2(ji-pz) ] 2 (3) indicated that AH* for the reaction in THF is 32(2) kJ/mol and AS* is -152 (8 ) J/K/mol; by the G ibbs relation, the free energy o f activation AG* is thus 7 6(l)k J/m ol. Reported errors on the energy terms were determined from the correlation coefficient, but error estim ates were alternatively determined by slope range estimation on the graph. Second step data were determ ined as follow s: AH* = 4 0 (5 ) kJ/mol; AS* = -1 5 7 (1 6 ) J/K /m ol, with AG* = 8 6(l)k J /m o l. The parameters as determ ined for the reaction in benzene were not substantially different, with A H *= 2 7 (5 ) kJ/m ol, AS* = -148 J/K/mol, and AG* = 7 1 (2 ) kJ/mol. (see Figures 8 and 9).

T hese results indicate that the reaction o f methyl iodide (a) involves considerably different o b s e r v a b le kinetics depending on the solvent, and (b) is lik ely to in v o lv e a highly ordered interm ediate, as evidenced by the large negative AS* term. C alculated activation parameters are listed in Table 8, page 46.

An attempt w as made to isolate the intermediate responsible for the appearance o f the first is o s b e s tic p o in t, at 393 nm. A tw en ty -fo ld e x c e s s o f am m onium h ex a flu o ro p h o sp h a te salt w as added to 2 0 m L o f a 4 0m M T H F s o lu tio n o f [Ir(CO)2(|i-p z) ]2 (3); a 10s excess o f methyl iodide w as added to this mixture, w hich was follow ed by 50 mL o f hexanes within thirty seconds to ensure com plete precipitation o f a deep orange product 12. IR spectroscopy established that carbonyl absorptions o f 12 were shifted to a higher energy from both starting material and the m ethyl iodid e adduct (observed absorptions at 2 1 4 0 cm -1, 2098 c m '1, 2050 cm '1, starting material and m ethyl

that a broad absorption corresponding to PF6 w as identified at 840 cm '1. Conductivity m easurem ents w ere recorded on a 5 m M TH F solution o f 12 against an ammonium hexafluorophosphate standard assum ing a stoichiom etry o f 12 corresponding to that o f [Ir(CO)2(H-pz)]2C H3(PF6). A conductivity o f 103 G H c n ^ m o l'1 (vs. 97 C H c n A n o l'1 for a 5 m M THF solution o f NH4PF6, corrected for background THF conductivity) confirm ed a 1:1 electrolyte. An experiment designed to observed the intermediate was also conducted using a stoichiom etric m ix o f m ethyl iodide and [Ir(C O )2(|i-pz)]2 (3 ) in acetonitrile-d? and monitored by ]H NM R. It was found that while a major fraction o f the 3 reacted to give the alkyl halide adduct, 30% o f the product produced was a previously unobserved dissym m etric species with a methyl resonance at 51.79 ppm. Further isolation and identification was not possible due to the tendency to form the alkyl halide adduct [Ir2(CO)4(H-pz)2(C H3)I] (4) on workup.

Absorbance vs time experiments were also completed on several bridge-substituted an alogues o f [Ir(C O )2(|a -p z) ] 2 (3 ); the rates o f the reaction o f m ethyl iodide with [Ir(CO)2(H-3M epz) ] 2 (3a), [Ir(CO)2( |i-3 5D M epz) ] 2 (3b), and [Ir(CO)2(H-TM cpz)l2 (3c) in the dark were extrem ely slow , so that a com plete (95%) reaction would take up to 2 h at high concentrations o f methyl iodide. An estimate o f the reaction rate constants were made (see Table 8, page 46), and it was determined that the rate o f reaction was inversely related to the extent o f bridge substitution for 3-m ethyl pyrazolyl, 3,5-dim ethylpyrazolyl, and 3,4,5-trim ethylpyrazolyl. The position o f the isosbestic also varied from 395 nm for [Ir(C O )2(M-3M e p z) ] 2 ( 3 a ), 395 nm for [Ir(CO)2([t-3 5D M e p z) ] 2 (3 b ), and 4 1 0 nm [Ir(C O )2(p.-T M epz) ] 2 (3 c ) during tim ed m ultiple scans in the dark. A first-order rate constant k2 was determined in each case, these being 0.0023 s' 1 for the reaction o f methyl iodide with 3a, 0 .0 0 0 1 4 s' 1 for 3b, and 0.00015 s' 1 for 3c. An increase o f [M el] to 2.0 M w as used to set the shortest possib le reaction tim e,39 and temperature dependence

29

experim ents were com pleted for 3 a , and 3 c at local maxima in analogous manner to that described in the unsubstituted pyrazolyl case. The activation parameters were determined by varying the temperatures in 5° C increments; the results are listed in Table 8. The relative rates o f the reactions are reflected in the AG^ values - a difference o f 10 kJ/mol betw een the substituted and the unsv jstituted pyrazolyl diiridium(I) com p lex results in a ten-fold decrease in the rate, and a 5 kJ/mol increase results in a 2-fold decrease in reaction rate. The activation parameters also indicate that AS* for the reactions o f both bridge substituted diiridium com plexes with m ethyl iodide was highly negative, suggesting that the transition state in these system s is ordered and more unfavorable than that for the unsubstituted complex; it is also significant that the enthalpy term is the same within experim ental error as the enthalpy o f the unsubstituted bridge com plex. Therefore the difference in the entropy o f activation is the most significant factor o f the higher free energy terms encountered in the kinetics o f methyl iodide addition to the diiridium tetracarbonyl co m p lex es, suggesting that the transition state o f the reaction o f m ethyl iodide with [Ir(CO )2(lt-pz) ]2 (3) and congeners is sterically inhibited by the 3 and 5 positions o f the pyrazolyl bridge.

The reaction was also sensitive to the amount of light reaching the cuvette during the mixing stages of the experiment; this is shown by a further set o f experiments in which the cuvettes containing the diiridium com plex were exposed to ambient fluorescent light in the time between the w avelength scans during a reaction. W hile the rate v/as still lo w and inversely related to the degree o f substitution, the rate o f reaction increased by an order of m agnitude. N o dependency o f rate on either diiridium concentration or m ethyl iodide concentration was found for the light exposure experiments, i.e. varying the concentration o f methyl iodide from 0 .0 5 M to 4 M did not change the observed rate o f reaction, and zero-order kinetics dominated. In this w ay it was observed that the rate o f the reaction o f m ethyl iodide with [Ir(CO)2(|i-3M epz) ] 2 (3a) increased qualitatively by a factor o f 1 0 (to

reaction rate o f [Ir(CO)2(p.-3 5D M ep z) ] 2 (3b ) and Ir(CO)2(lt-3 4 5T M ep z) ] 2 (3 c) with m ethyl iodide increased by a factor o f 5 0 on exposure to ambient light (to 0.0084 M ' V1 and 0 .0 0 7 9 M ' V1 resp ectively). In addition, it w as noted that there w as a significant change in the w avelength o f the isosbestic point between the irradiated and the non­ irradiated c o m p le x e s - for exam p le, the iso sb estic point o f the dark reaction of [Ir(C O )2( |i -3 4 5T M ep z) ] 2 (3 c ) with methyl iodide was observed at 4 1 0 nm, w hile the isosbestic point o f the equivalent light reaction was observed at 401 nm.

It was observed that the dark reactions departed from psuedo-first order kinetics as the lev el o f contaminants (e.g. the presence o f oxygen) increased. This problem was circum vented by m aintaining a rigorously clean reaction environm ent and the use of Schlenk techniques (whenever possible) to exclude interfering gases.

The light induced reactions were also significantly affected by addition o f radical trap reagents such as hydroquinone and catechol. For exam ple, addition o f 5mg o f either reagent to 3 mL (cuvette volum e) o f [Ir(CO)2(|i-3M epz) ] 2 (3 a ), [Ir(CO>2(|i-3 5DM epz) |2 (3 b ), or [Ir(CO)2(p.-TMepz) ] 2 (3c) in THF reduced the rate constant by a factor o f three; although radical trap effects are typically greater than this,4 0 the effect was reproducible.