Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Hydrogen and methanol activation by some tertiary phosphine ruthenium complexes Hampton, Cashman Roger Stirling Mason 1989

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
[if-you-see-this-DO-NOT-CLICK]
UBC_1989_A1 H35.pdf [ 12.66MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0059788.json
JSON-LD: 1.0059788+ld.json
RDF/XML (Pretty): 1.0059788.xml
RDF/JSON: 1.0059788+rdf.json
Turtle: 1.0059788+rdf-turtle.txt
N-Triples: 1.0059788+rdf-ntriples.txt
Original Record: 1.0059788 +original-record.json
Full Text
1.0059788.txt
Citation
1.0059788.ris

Full Text

HYDROGEN AND METHANOL ACTIVATION BY SOME TERTIARY PHOSPHLNE RUTHENIUM COMPLEXES By CASHMAN ROGER STIRLING MASON HAMPTON B.Sc, University of Victoria, 1967 M.Sc, Simon Fraser University, 1970 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE DEPARTMENT OF CHEMISTRY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September 1989 © Cashman Roger Stirling Mason Hampton, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) ii Abstract The previously known complexes, R^HtCWPRak, have now been correctly reformulated as the Ti2-H2 species (Ti2-H2)(PR3)2Ru(u-Cl)2(u-H)RuH(PR3)2 (R = Ph, p-tol), la and lb, and it is confirmed that in solution they are dimeric and undergo no ligand dissociation. Also, a new analogue of complexes of type 1 is reported: the complex (n2-H2)(isoPFA)Ru(n-Cl)2(u-H)RuH(PPh3)2,4, is formed from the reaction of RuCl2(PPh3)(isoPFA), 3b, with H2 in methanol/benzene, and a crystal structure of 4 shows the ri2-H2 ligand; isoPFA and PPFA (see below) are ferrocene based, chelating P-N ligands, with the structures: R = Pr1 and Ph for isoPFA and PPFA, respectively. Complexes la, lb and 4 all react with 1-hexene to give hexane; the main ruthenium phosphine product in the case of 1 is the corresponding RuHCl(PR3)3 complex, while 4 reacts to give a complex mixture of ruthenium phosphine complexes, including 3b. The amount of hexane formed from the reaction of 4 with hexene is quantified as 2 mol/mol 4. The hydrogenation of 1-hexene catalyzed by la is re-interpreted as occurring via the mechanism: (Ti2-H2)(PPh3)2Ru(^-Cl)2(^-H)RuH(PPh3)2 + hexene —(PPh3)2Ru(u-Cl)2(u-H)RuH(PPh3)2 + hexane (1) (PPh3)2Ru(^-Cl)2(^-H)RuH(PPh3)2 + H2 ^=£~ (Ti2-H2)(PPh3)2Ru(u-Cl)2(u-H)RuH(PPh3)2 (2) Reactions of RuCl2(PPh3)(PPFA), 3a, and RuCl2(PPh3)(isoPFA), 3b, with H2 have been further studied, in connection with earlier mechanistic studies on the iii hydrogenation of organic substrates catalyzed by complex 3a. The complex 3a reacts with 2-8 atm H2 in H-butanol to give ruthenium phosphine products including la. The complex 3b reacts with H2 in methanol/benzene to give 4, as mentioned above, as well as a number of unidentified hydrides; in DMA, the reaction of 3b with H2 gives la, 4, RuHCl(PPh3)(isoPFA) (7), RuHCl(PPh3>3 and other unidentified ruthenium phosphine complexes. The product H2NMe2+Cl" was also isolated from the methanol/benzene reaction mixture, and this product provides evidence that the amine functionality of the P-N ligands is involved in the promotion of the heterolytic cleavage of dihydrogen to give a proton and a hydride (H2 » H+ + H'). Kinetic studies on the hydrogenation of 1-hexene catalyzed by 3a, and by 3b in the present work, are now interpreted according to the mechanism: Ru(n)H(PPh3) Kr Ru(U)H + PPh3 (3) Ru(H)H + H2 Ru(n)H(Tl2-H2) K2 hexene Ru(II)(hexyl) hexene k2 H2 Ru(TJ)H + hexane (4) iv Reactions involving 3b and methanol have also been studied, and 3b is also active for the transfer hydrogenation (from methanol) of ketones and activated olefins. The reaction of 3b with methanol in the absence of base is proposed to occur with the stoichiometry: RuCl2(PPh3)(isoPFA) + 2MeOH H2NMe2+Cl- + H2 + 3b RuHCl(CO)(PPh3)(isoPOF), 5 (5) where the ligand isoPOF is formed from isoPFA by replacement of the NMe2 group on isoPFA by a methoxo group; reaction 6 could occur via the following steps: RuCl2(PPh3)(isoPFA) + MeOH RuHCl(CO)(PPh3)(isoPFA), 6 3b + H2 + HC1 (6) RuHCl(CO)(PPh3)(isoPFA) + MeOH RuHCl(CO)(PPh3)(isoPOF), 5 + HNMe2 (7) HC1 + HNMe2 H2NMe2+Cl- (8A mechanism for reaction 7 is presented and invokes reversible attack by MeOH with replacement of Q", followed by reversible deprotonation of cc<Mdinated MeOH to give successively methoxo, formaldehyde and formyl intermediates, and finally the hydrido-carbonyl, 6. The reaction of 3b with methanol in the presence of KOH is proposed to occur according to the stoichiometry: RuCl2(PPh3)(isoPFA) + KOH + CH3OH RuHCl(CO)(PPh3)(isoPFA) + KCI + H2 + H20 (9) and two pathways have been identified, one base-independent, identical to that proposed for reaction 7, and one showing a second-order dependence on KOH. The latter pathway invokes initial reversible attack on RuQ2(PPh3)(isoPFA), 3b, by MeO", replacing Cl" to give RuCl(OMe)(PPh3)(isoPFA), and subsequent reversible replacement of PPh3 by OH*, followed by concerted loss of OH" and hydride transfer from coordinated OMe- to give a V hydrido-formaldehyde complex RuHCl(T|2-CH20)(isoPFA). A subsequently formed formyl intermediate reacts via intramolecular hydride transfer from the formyl to the metal, H2 loss, and phosphine coordination to give the hydrido-carbonyl 6. vi Table of Contents Abstract List of Tables xiv List of Figures xyi List of Symbols and Abbreviations xxiii Acknowledgements xxviiChapter 1 Introduction Chapter 2 General experimental 5 2.1 General procedures2.2 Instrumentation 8 2.3 NMR studies 11 2.4 Materials 2 SolventsGases 1Phosphines 2 Olefin and ketone substrates 1Other materials 13 Ruthenium complexes(Tl2-H2)(PR3)2Ru(a-Cl)2(u-H)RuH(PR3)2 (R = Ph, p-tol), la, lb 13 vii Ru2CU(PR3)4 (R = Ph, p-tol), 2a, 2b 14 RuCl2(PPh3)(PPFA), 3a 15 RuCl2(PPh3)(isoPFA), 3b 6 (Tl2-H2)(isoPFA)Ru(p-Cl)2(p-H)RuH(PPh3)2,4 17 RuHCl(CO)(PPh3)(isoPOF), 5 18 RuHCl(CO)(PPh3)(isoPFA), 6 9 RuHCl(PPh3)(isoPFA), 7 20 Chapter 3 Characterization and reactivity of some dinuclear ruthenium(H) phosphine complexes containing Ti2-H2 22 3.1 Introduction 2Historical background 22 Factors affecting the stability of T\2-H2 complexes of the transition metals 23 Theoretical considerations 24 Criteria for recognition of dihydrogen complexes 27 Dinuclear, Tj2-^ complexes reported in the present work 32 3.2 Experimental 36 Crystal structuresVisible spectrophotometry 36 NMR spectra 3Reactions of la and lb with 1-hexene 37 Reactions of (Ti2-H2)(isoPFA)Ru0i-Cl)2(p-H)RuH(PPh3)2,4 37 Preparation and observation of dihydrogen adducts of 2a and 2b in situ 33.3 Results 8 Complexes la, lb and 4 3viii Dihydrogen adducts of Ru2Cl4(PR3)4 (R = Ph, p-tol), 2a and 2b 67 3.4 Discussion 6Complexes la, lb and 4 67 Dihydrogen adducts of the complexes, Ru2Cl4(PR3)4 (R = Ph, p-tol), 2a and 2b 79 Chapter 4 Mechanistic studies on hydrogenation catalyzed by tertiary phosphine complexes of ruthenium(H) 81 4.1 IntroductionHydrogenation of 1-hexene catalyzed by la 83 Hydrogenation systems catalyzed by 3a and 3b 86 Hydrogenation systems catalyzed by 3a 87 Hydrogenation systems catalyzed by RuCl2(PPh3)(isoPFA), 3b 91 4.2 ExperimentalReactions of RuCl2(PPh3)(PPFA), 3a, with H2 in toluene-dg, toluene-dg/butanol and butanol 91 Reactions of RuCl.(PPh3)(isoPFA), 3b, with hydrogen in DMA, DMA/C6D6, and benzene/methanol 93 Catalytic hydrogenations of prochiral substrates using RuCl2(PPh3)(PPFA), 3a, and RuCl2(PPh3)(isoPFA), 3b 95 NMR studies of the catalytic hydrogenation of 1-hexene by 3a, 3b and (Ti2-H2)(PPh3)2Ru(n-Cl)2(n-H)RuH(PPh3)2, la 95 Hydrogen-uptake measurements for 1-hexene hydrogenation 96 4.3 Results 9Reactions of RuCl2(PPh3)(PPFA), 3a, with hydrogen in toluene-d«, toluene-dg/butanol and butanol 96 ix Reactions of RuCl2(PPh3)(isoPFA), 3b, with hydrogen in DMA, DMA/C6D6. and benzene/methanol 109 In the presence of added PPh3 10In the absence of added PPh3 109 Catalytic hydrogenations of prochiral substrates catalyzed by (SJi)-RuCl2(PPh3)(PPFA), 3a and (S,rt)-RuCl2(PPh3)(isoPFA), 3b 115 NMR studies of the catalytic hydrogenation of 1-hexene by (TJ2-H2)(PPh3)2Ru(a-Cl)2(u-H)RuH(PPh3)2, la, RuCl2(PPh3)(PPFA), 3a and RuCl2(PPh3)(isoPFA), 3b 115 The la system 11The 3a system 8 The 3b system 119 Kinetic data from hydrogen-uptake measurements for 1-hexene hydrogenation 120 using (Ti2-H2)(PPh3)2Ru(a-Cl)2(^-H)RuH(PPh3)2, la, as catalyst 120 using compound A as catalyst, prepared from RuCl2(PPh3)(PPFA), 3a, and H2 in butanol 120 using RuCl2(PPh3)(isoPFA), 3b as catalyst 121 using (Ti2-H2)(isoPFA)Ru(^i-Cl)2(fX-H)RuH(PPh3)2,4 as catalyst 123 4.4 DiscussionThe (Ti2-H2)(PPh3)2Ru(u-Cl)2ai-H)RuH(PPh3)2, la 1-hexene hydrogenation system 123 Hydrogenations involving RuCl2(PPh3)(PPFA), 3a 127 Reactions of 3a with H2 12Hydrogenation systems catalyzed by 3a 130 Hydrogenations involving RuCl2(PPh3)(isoPFA), 3b 132 Hydrogenation of 1-hexene catalyzed by RuCl2(PPh3)(isoPFA), 3b 132 Hydrogenation of prochiral substrates catalyzed by RuCl2(PPh3)(PPFA), 3a, and RuCl2(PPh3)(isoPFA), 3b 133 er 5 Reactions of RuCl2(PPh3)(isoPFA) and derivatives with methanol 135 5.1 Introduction 13Mechanistic Studies 142 5.2 Experimental 147 Experiments on the effects of various solvents, added PPh3 and air, respectively, on visible spectra of RuCl2(PPh3)(isoPFA), 3b, in solution 147 Experiment on the effect of styrene on 3b in solution 147 Reactions of RuCl2(PPh3)(isoPFA), 3b, and the derivative, RuHCl(CO)(PPh3)(isoPOF), 5, with methanol in the absence of added base 148 Catalytic activity of 5 and 3b 14Reactions of RuCl2(PPh3)(isoPFA), 3b, and derivatives with methanol in the presence of added base 149 With Proton Sponge® as base 14With KOH as base 150 Experiments on the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/methanol, in THF in the presence or absence of styrene 151 Monitoring changes in the ^PpH} NMR spectrum 153 Monitoring changes in the visible spectrum 153 Monitoring changes in the styrene/ethylbenzene ratio 154 xi Monitoring gas evolution 155 Detection of KCI produced in the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/MeOH 158 Formaldehyde detection 15Experiment to detect gas phase CO by GC 159 2D NMR experiments on distillates from reactions of 3b with KOH/MeOH-d4 and MeOH-di 159 An experiment to determine what would happen to CH2O generated in a solution similar to those used for the reactions of 3b with KOH/MeOH 160 Reactions of RuHCl(CO)(PPh3)(isoPFA), 6, with methanol in the absence and presence of base 160 5.3 Results 161 Experiments on the effects of various solvents, added PPh3 and air, respectively, on visible spectra of RuCl2(PPh3)(isoPFA), 3b, in solution 161 Experiment on the effect of styrene on 3b in solution 161 Reactions of RuCl2(PPh3)(isoPFA), 3b, and RuHCl(CO)(PPh3)(isoPOF), 5, with methanol in the absence of base 163 Reactions of RuCl2(PPh3)(isoPFA), 3b, with methanol in the presence of added base 170 With Proton Sponge® as added base 170 With KOH as base 175 Reactions of RuHCl(CO)(PPh3)(isoPFA), 6, with methanol in the absence and presence of base 184 xii Experiments on the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/methanol, in THF in the presence or absence of styrene 185 Experiment to detennine the reactivity of THF with methanolic KOH 185 Monitoring the 31P{1H} NMR spectrum 185 Work-up of THF/MeOH reaction mixtures 188 Gas evolution experiment results 189 Monitoring the THF/MeOH reaction by visible spectrophotometry 191 Detection of KCI 209 Formaldehyde detection 210 Experiment to detect gas phase CO by GC 210 2D NMR experiments on distillates from reactions of 3b with KOH/MeOH-d4 and MeOH-di 210 An experiment to determine what would happen to CH2O generated in a solution similar to those used for the reactions of 3b with KOH/MeOH 213 5.4 Discussion 4 Solution behaviour of RuCl2(PPh3)(isoPFA), 3b 214 Reactions of RuCl2(PPh3)(isoPFA), 3b, and derivatives with methanol in the absence of added base 214 Reactions of RuCl2(PPh3)(isoPFA), 3b, with methanol in the presence of added base 218 With Proton Sponge® as added base 218 With KOH as added base 222 Experiments on the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/methanol, in THF in the presence or absence of styrene 223 Organic and gaseous products from high [KOH] reactions 233 The base-independent mechanism and the reaction of RuCl2(PPh3)(isoPFA), 3b with methanol in the presence of Proton Sponge® 233 The base-independent mechanism and the reaction of RuCl2(PPh3)(isoPFA), 3b, with methanol in the absence of base 234 Chapter 6 Summary and suggestions for further work 235 Bibliography 241 xiv List of Tables Table 3.1 Transition metals which have been found in molecular hydrogen complexes 23 Table 3.2 Electronic configurations and oxidation states of examples of ri2-H2 complexes 4 Table 3.3 Coordination numbers of rj2-H2 complexes, according to electron configuration 2Table 3.4 Vibrational frequencies, VHHCCDI"1), for dihydrogen complexes 26 Table 3.5 Some !H NMR HD coupling constants (Hz) for T|2-H2 resonances 29 Table 3.6 Selected Tj (minimum) values for various hydride and dihydrogen complexes 31 Table 3.7 Selected bond angles and distances for crystal structures of la and lb 39 Table 3.8 Selected bond angles and distances for the crystal structure of 4 1 Table 3.9 Spectroscopic data for la 72 Table 4.1 Ti (300 MHz) values for upfield *H NMR resonances observed for the reaction of RuCl2(PPh3)(PPFA), 3a, with H2 in toluene-ds/butanol at ambient temperature 101 Table 4.2 Hydrogenation results for various substrates using (SJt)-RuCl2(PPh3)(PPFA), 3a and (5^)-RuCl2(PPh3)(isoPFA), 3b as catalysts 116 Table 5.1 Selected bond distances for complexes RuHCl(CO)(PPh3)(isoPOF), 5, RuHCl(CO)(PPh3)(isoPFA), 6, and RuHCl(PPh3)(isoPFA), 7 165 Table 5.2 Selected bond angles associated with the bond distances shown in Table 5.1 166 Table 5.3 Catalytic hydrogen transfer from methanol, catalyzed by RuCl2(PPh3)(isoPFA), 3b, in neat methanol at 105 °C in the absence of base 170 Table 5.4 Colour changes of reaction mixtures and rates of hydrogen transfer from methanol to styrene to give ethylbenzene, as a function of time and initial [KOH] 177 Table 5.5 Selected molecular structure parameters for RuH2(CO)Q?Ph3)3 182 Table 5.6 Some kinetic data for the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/MeOH in the absence of styrene 197 Table 5.7 Some kinetic data for the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/MeOH showing dependence on [styrene] 198 XV Table 5.8 Summary of kinetic data from the GC experiments used to follow styrene hydrogenation during the reaction of 3b with KOH/MeOH in the presence of styrene 204 Table 5.9 Comparison of structural parameters of RuHCl(PPh3)(isoPFA), 7, and RuHCl(PPh3)3 221 List of Figures Figure 1.1 Structure of FA, PPFA and isoPFA 3 Figure 2.1 Apparatus for filling tubes with pressures up to 10 atm 6 Figure 2.2 Glass apparatus used for manipulating materials under elevated hydrogen pressures, up to 10 atm 7 Figure 2.3 Visible spectrophotometer cell (5 cm) 9 Figure 2.4 Visible spectrophotometer cell (1 mm)Figure 3.1 Orbital diagrams showing the contributions to the bonding for the M-Cn2-H2) structure 26 Figure 3.2 The skeletal structure of la 40 Figure 3.3 The skeletal structure of lbFigure 3.4 The molecular structure of 4 showing the positions of hydrogen ligands 42 Figure 3.5 The visible spectrum of la 43 Figure 3.6 Observed and calculated, variable temperature 31P{ !H) NMR of la 44 Figure 3.7 Eyring plot for rate constants for phosphine exchange in la 45 Figure 3.8 Observed and calculated, variable temperature lH NMR spectra of la 46 Figure 3.9 General hydrogen ligand exchange model used to simulate hydrogen ligand exchange within la, lb and 4 7 Figure 3.10 Plot of !H NMR chemical shifts versus temperature for terminal and bridging hydride ligands in la, with extrapolation to 20 °C 47 Figure 3.11 Plot of lH NMR T2 values versus temperature for the Tj2-H2 ligand of la 4Figure 3.12 Eyring plot for rate constants for hydrogen ligand exchange in la 48 Figure 3.13 Plot of Ti values versus temperature for hydrogen ligands of la 49 Figure 3.14 Observed and calculated variable temperature *H NMR spectra for hydrogen ligands of lb 50 Figure 3.15 Plot of lH NMR chemical shifts versus temperature for the TI2-H2 and terminal hydride ligands of lb, including extrapolation to 20 °C 51 Figure 3.16 Plot of JH NMR T2 values versus temperature for the ri2-H2 ligand of lb, including extrapolation to 20 °C 52 Figure 3.17 Eyring plot for rate constants for hydrogen ligand exchange in lb 53 Figure 3.18 Inversion-recovery method plot for lH NMR Ti measurements for the hydrogen ligands of lb 54 Figure 3.19 Plot of lH NMR Ti values versus temperature for the hydrogen ligands of lbFigure 3.20 The 3lP{lR) NMR spectrum of 4. 56 Figure 3.21 The XH NMR spectrum of 4 7 Figure 3.22 Observed and calculated variable temperature !H NMR data for 4 58 Figure 3.23 *H NMR T2 values versus temperature for the rt2-H2 ligand of 4, with extrapolation to higher temperatures 59 Figure 3.24 Dependence of the *H NMR chemical shift on temperature for the |i-H ligand of 4 .60 Figure 3.25 Eyring plot for the ki values (rp-Hi/iL-H exchange) for 4 61 Figure 3.26 Inversion-recovery plot for lH NMR Ti measurements for the hydrogen ligands of 4. 6Figure 3.27 Plot of *H NMR Ti values versus temperature for the hydrogen ligands of 4 2 Figure 3.28 !H NMR spectrum of lb, after two weeks under vacuum at ambient temperature 64 Figure 3.29 31P{1H) NMR spectrum of lb, after two weeks under vacuum at ambient temperature 5 Figure 3.30 Slow exchange structure showing proposed hydrogen ligand positions for lb, reported as a classical terrahydride 69 Figure 3.31 31P{1H} NMR coupling pattern for complex la 71 Figure 3.32 The 31P{1H) NMR coupling pattern for complex 4 78 Figure 4.1 Uptake plots for the la-catalyzed hydrogenation of 1-hexene 84 Figure 4.2 Dependence of maximum rate of hydrogenation on total Ru concentration (la) in DMA at 30 °C 8Figure 4.3 Dependence of maximum rate of la-catalyzed hydrogenation on 1-hexene concentration 84 Figure 4.4 Dependence of maximum rate of la-catalyzed hydrogenation on hexene concentration, plotted according to equation 4.18 84 Figure 4.5 Dependence of maximum rate of la-catalyzed hydrogenation on hydrogen concentration 85 Figure 4.6 Dependence of maximum rate of la-catalyzed hydrogenation on hydrogen concentration, plotted according to equation 4.19 85 Figure 4.7 Dependence of maximum rate of la-catalyzed hydrogenation on added triphenylphosphine concentration 8Figure 4.8 Dependence of the maximum rate of la-catalyzed hydrogenation on [PPI13]-1 85 Figure 4.9 Dependence of maximum rate of la-catalyzed hydrogenation on added lithium chloride concentration 86 Figure 4.10 Dependence of maximum rate of 3a-catalyzed hydrogenation on total Ru concentration... 89 Figure 4.11 Dependence of maximum rate of 3a-catalyzed hydrogenation on 1-hexene concentrationFigure 4.12 Dependence of maximum rate of 3a-catalyzed hydrogenation on partial pressure of H2 89 Figure 4.13 Inverse dependence of maximum rate of 3a-catalyzed hydrogenation -on added phosphine concentration 8Figure 4.14 31P{ 1H} NMR spectrum of the in situ products of the reaction of 3a withH2 '.. 98 Figure 4.15 3 *P {lH} NMR spectrum of the reaction mixture formed from the reaction of 3a with H2 9 Figure 4.16 The upfield !H NMR spectrum of in situ products from the reaction of 3a withH2 100 Figure 4.17 31P{ NMR spectrum of the yellow precipitate from the reaction of 3a with H2 in n-butanol 102 Figure 4.18 !H NMR spectra of the yellow precipitate from the reaction of 3a with H2 in n-butanol 103 Figure 4.19 *H NMR spectra of the yellow precipitate from the reaction of 3a with H2 in H-butanol, within a) 1 h, and b) 6 h of making up the solution 105 Figure 4.20 Variable temperature NMR spectra of the yellow deposit from the reaction of 3a with hydrogen in benzene/hexane/methanol 108 Figure 4.21 The XH NMR spectrum of 3b 110 xix Figure 4.22 The 31P{ NMR spectrum of a solution obtained after ~6 h from mixing 3b in C6D6/DMA Ill Figure 4.23 The *H NMR spectrum of the residue obtained by removal of solvent from the supernatant of the reaction of 3b with H2 112 Figure 4.24 The EI mass spectrum of the white crystals of dimethylammonium chloride isolated from the reaction of 3b with H2 113 Figure 4.25 The 31P{ 1H} NMR spectrum of the blackish red crystals of complex Y isolated from the reaction of 3b with H2. 114 Figure 4.26 The *H NMR spectrum of YFigure 4.27 The 31P{1H} NMR spectrum of the la-catalyzed hydrogenation solution containing 0.04 M 1-hexene 117 Figure 4.28 The 31P{ NMR spectrum of the la-catalyzed hydrogenation solution containing 1.0 M 1-hexene 11Figure 4.29 The 31P{1H} NMR spectrum acquired during the second hour since the start of the la-catalyzed hydrogenation reaction with 1.0 M 1-hexene 118 Figure 4.30 The 31P{ 1H) NMR spectrum of a 3a-catalyzed hydrogenation solution 119 Figure 4.31 The 31P{!H} NMR spectrum of a 3b-catalyzed hydrogenation solution 11Figure 4.32 H2 uptake plot for 1-hexene hydrogenation catalyzed by 3b 120 Figure 4.33 Dependence of maximum rate of hydrogenation on total Ru concentration (3b) 121 Figure 4.34 Dependence of maximum rate of hydrogenation on the square root of the total Ru concentration (3b) 12Figure 4.35 Hexene dependence of maximum rate of 3b-catalyzed hydrogenation in DMA 122 Figure 4.36 Hydrogen dependence of maximum rate of 3b-catalyzed hydrogenation in DMA 12Figure 4.37 PPhrj dependence of the reciprocal of the maximum rate of 3b-catalyzed hydrogenation of 1-hexene 122 Figure 5.1 Scheme proposed in reference 145 for the dehydrogenation of methanol catalyzed by RuCl(OAc)(PPh3)3 145 Figure 5.2 Glass tube used to monitor changes in the styrene/ethylbenzene ratio during transfer hydrogenation of styrene 152 XX Figure 5.3 a) A 25 mL, modified gas evolution flask for following reactions between RuCl2(PPh3)(isoPFA), 3b, in THF, and a small volume of KOH/methanol solution, b) Similar apparatus, for use with larger volumes of KOH/methanol solution 156 Figure 5.4 The visible spectrum of 3b in CHCI3 solution under nitrogen in the presence and absence of added PPh3 ....; 163 Figure 5.5 The visible spectrum of RuCl2(PPh3)(isoPFA), 3b, in CHC13 solution under nitrogen and on exposure to air, respectively 164 Figure 5.6 The X-ray structure of RuHCl(CO)(PPh3)(isoPOF), 5 164 Figure 5.7 The *H NMR spectrum of 5 167 Figure 5.8 The *H NMR spectrum of a deuteriated sample of 5 168 Figure 5.9 The X-ray structure of RuHCl(CO)(PPh3)(isoPFA), 6 171 Figure 5.10 A stereoscopic view of the structure of 6 17Figure 5.11 The *H NMR spectrum of RuHCl(CO)(PPh3)(isoPFA), 6 172 Figure 5.12 The X-ray structure of RuHCl(PPh3)(isoPFA), 7 174 Figure 5.13 The 31P{1H} NMR spectrum of 6 after ~1 wk in solution under nitrogen in an NMR tube 176 Figure 5.14 The 31P{ !H) NMR spectrum of the supernatant of a reaction mixture formed by heating RuCl2(PPh3)(isoPFA), 3b, in dry methanol in the presence of KOH under nitrogen 178 Figure 5.15 The lH NMR spectrum of the residue obtained from the solution used for the 31P{ 1H) NMR spectrum given in Figure 5.14 179 Figure 5.16 The visible spectrum of the CDCI3 solution used for the lH NMR spectrum given in Figure 5.15 17Figure 5.17 The upfield *H NMR spectrum of RuH2(CO)(PPh3)3 180 Figure 5.18 The X-ray structure of RuH2(CO)(PPh3)3 181 Figure 5.19 The *H NMR spectrum of the residue obtained from the eluted second red band, containing complex X, obtained on chromatography of a worked-up supernatant from a RuCl2(PPh3)(isoPFA), 3b/KOH/methanol reaction rnixture 183 Figure 5.20 Variable time 31P{1H} NMR spectra of a 3b/THF/styrene/KOH/MeOH reaction mixture obtained within 66 min of starting the reaction 186 Figure 5.21 Variable temperature spectra for the same solution described for Figure 5.20 7 Figure 5.22 The 31P{ !H} NMR spectrum of the red reaction solution formed after 2 d for a reaction similar to that used for Figure 5.20 189 Figure 5.23 Gas evolution plot for a reaction between a solution of 3b in THF and a KOH/MeOH solution 190 Figure 5.24 Straight-line semi-log plot for Figure 5.23 indicative of first-order gas evolution, based on total mol gas evolved 190 Figure 5.25 The changes in the visible spectrum during a 20 °C reaction of 3b with basic methanol in THF under nitrogen with styrene present 192 Figure 5.26 Changes in the visible spectrum during the second stage (intermediate [KOH], yellow red) of a reaction between 3b and methanol in THF in the presence of KOH with styrene absent 193 Figure 5.27 Plot of the change in absorbance (605 nm) with time during a reaction of 3b with KOH/MeOH in THF at intermediate [KOH] (reaction 8, Table 5.6). 193 Figure 5.28 Plot of log (At - Ao») vs time (t) for the data of Figure 5.27, showing behaviour that is first-order in [3b] 194 Figure 5.29 Plot of the change in absorbance (605 nm) with time during a reaction of 3b with KOH/MeOH in THF at high [KOH] (reaction 10, Table 5.6) 195 Figure 5.30 Plot of log (At - Aoo) vs time (t) for the data of Figure 5.29, showing behaviour that is first-order in [3b] 196 Figure 5.31 Dependence of kobS, from visible spectrophotometric measurements, on [KOH] for the reaction of 3b with KOH/MeOH in THF (Table 5.6) 199 Figure 5.32 Plot of kobs. from spectrophotometric measurements, versus [styrene] at two different [KOH] values (Table 5.7) 19Figure 5.33 Conventional salt effect plot, showing the effect of LiN03 on the kobs for the first step of the reaction of 3b with KOH/MeOH in THF 200 Figure 5.34 Plot of kobs, from spectrophotometric measurements, and corrected for salt effect, versus [KOH]2, for the reaction of 3b with KOH/MeOH in THF 201 Figure 5.35 Plot of the percentage of styrene, {[styrene]t/[styrene]o} x 100%, present at time (t), during reaction 8, Table 5.8 202 Figure 5.36 Semi-log plot derived from the data of Figure 5.35 203 Figure 5.37 Plot of the percentage of styrene, {[styrene]t/[styrene]o} x 100%, present at time (t), during reaction 6, Table 5.8 205 Figure 5.38 Plot of the percentage of styrene, {[styrene]t/[styrene]o} x 100%, present at time (t), during reaction 7, Table 5.8 205 Figure 5.39 Semi-log plot, showing the first-order behaviour of the primary hydrogenation process, derived from the data of Figure 5.37 207 Figure 5.40 Semi-log plot, showing the first-order behaviour of the primary hydrogenation process, derived from the data of Figure 5.38 207 Figure 5.41 Plot showing [KOH] dependence of for styrene hydrogenation, in the reaction of 3b with methanol in the presence of styrene and KOH in THF (Table 5.8) 208 Figure 5.42 Plot showing [styrene] dependence of kj>t,s for styrene hydrogenation, in the reaction of 3b with methanol in the presence of styrene and KOH in THF (Table 5.8) 20Figure 5.43 2D NMR spectra of deuteriated ethylbenzene obtained by hydrogen transfer from (a) MeOH-dt and (b) MeOH-di to styrene during the reaction of RuCl2(PPh3)(isoPFA), 3b, with MeOH-dj and MeOH-di, respectively, in the presence of KOH in THF 211 Figure 5.44 2D NMR spectrum of a) an acetone/styrene solution, and b) an acetone/ethylbenzene solution 212 Figure 5.45 Plot of [OMe-] versus [OH-] based on the equilibrium constant, K = 18, for the reaction of NaOH with MeOH at 20 °C 227 Figure 5.46 Proposed mechanism for the second-order base-dependent pathway for reaction 5.30 via methoxide attack on 3b 228 Figure 5.47 Base-independent pathway for the reaction of 3b with KOH/MeOH in THF 231 List of Symbols and Abbreviations [ ] molar concentration [ ]o initial concentration [ ]j total concentration [ ] t concentration at time t [ final concentration M D specific rotation at temperature T and at the light wavelength of 589 nm from the sodium D line a adjacent position on a carbon chain A Angstrom(s) Ao initial absorbance Anal. elemental analysis At absorbance at time t atm atmosphere(s) (pressure unit) Aoo final absorbance bipy bipyridyl bq 7,8-benzoquinolinate br s broad singlet Bul tertiary butyl Calcd. calculated Chap. chapter cm centimetre(s) cm-1 wavenumber unit Cp cyclopentadienyl Cy cyclohexyxxiv A difference A heat 8 chemical shift H5 lH NMR chemical shift p5 31P{1H} chemical shift d day(s) d doublet AG* free energy of activation AH enthalpy change AH0 standard enthalpy change AH* enthalpy of activation diop -0-isopropylidene-2,3-dihydroxy-l,4-bis(diphenylphosphino)butane DMA N//-dimethylacetamiddppb l,4-bis(diphenylphosphino)butane dppe l,2-bis(diphenylphosphino)ethandppm bis(diphenylphosphino)methane dppp l,3-bis(diphenylphosphino)propanAS* entropy of activation e extinction coefficient e.e. enantiomeric excess EI electron impact Et ethyl Fc ferrocenyFID flame ionization detector Fig. Figure G C gas chromatography XXV gas chromatography-mass spectrometry hour(s) hapricity of degree n Hertz infrared 1 - [a-A^N-dimethy laminoethyl]-2-diphenylphosphino-ferrocene isoPOF l-[a-methoxyethyl]-2-diisopropylphosphinoferrocene J coupling constant J Joule(s) K equilibrium constant k rate constant kJ kilojoule(s) k0bs observed rate constant A equivalent conductance ^ wavelength lit. literature ^-max wavelength of an absorbance maximum log common logarithm M metal atom M molar fl bridging position between two metal atoms in a molecule m multiplet m/z mass/charge ratio Me methyl GCMS h Hz IR isoPFA meso-tetraphos-1 mg MHz UL mm mmol mol v nm NMR OAc P-N P-P Ph PPFA PR3 Pr1 PS p-tol q R Rw s s (Ph)2P(CH2CH2)PPh(CH2CH2)-PPh(CH2CH2)PPh2 microgram(s) milligram(s) megahertz microlitre(s) millimetre(s) millimole(s) mole(s) vibrational frequency nanometre(s) nuclear magnetic resonance acetyl chelating ditertiary phosphine-arnine chelating ditertiary phosphine phenyl l-[a-ArA-dimemylaminoethyl]-2-diisopropylphosphino-ferrocene tertiary alkyl or aryl phosphine isopropyl Proton Sponge® para-tolyl quartet residual index weighted residual index second(s) singlet sect, str TCD t t Ti tl/2 T2 tetraphos-2 THF TMS UV-vis v br s v/v u) w W1/2 {IH} {phenyl-^} section strong thermal conductivity detector time triplet longitudinal relaxation time halflife transverse relaxation time P(CH2CH2PPh2)3 tetrahydrofuran tetramethylsilane ultraviolet-visible very broad singlet volume-to-volume ratio Larmor frequency weak halfwidth broadband proton-decoupled phenyl proton-decoupled Acknowledgments I am grateful to those who have assisted me during the course of my work and in the preparation of this thesis. In particular, valuable help and guidance has been received from my supervisors Dr. B.R. James and Dr. W.R Cullen. I have appreciated helpful comments from members of my guidance committee: Dr. M. Fryzuk, Dr. P. Legzdins and Dr. A. Merer. Assistance in many practical matters has come from many members of both research groups. Indispensable assistance has been provided by members of the support services, including those working in the analytical, electronic, glass blowing, illustration, mechanical and NMR divisions of the chemistry department. My wife and children have endured my frequent absence and lateness most graciously. Thanks are also due to my parents, as well as to my wife, for their encouragement. 1 Chapter 1 Introduction Homogeneous catalysis holds practical importance for organic synthesis1>lb and also for inorganic reactions.10 The generally mild conditions required for such catalysis, and its amenability to study are other factors which have helped generate the tremendous growth in interest in homogeneous catalytic systems in the past thirty years. Hydrogenation systems for organic substrates have received particular interest,2,3'4 not only for their ease of study, and practical usefulness, but also for the relative simplicity, in selectivity for, organic products. Attractive features notwithstanding, homogeneous hydrogenation systems have received limited commercial interest, because of (a) problems with catalyst/product separation and recovery following the catalytic process and (b) phosphine degradation of phosphine containing catalysts, which occurs under reducing conditions.lb Catalytic hydrogenation of organic substrates usually involves activation of both dihydrogen and the substrate by bonding to a metal atom, so that these molecules therefore undergo reactions which they would not undergo in the uncomplexed state. Activation of dihydrogen involves formation of an active metal hydride species, and this species can transfer a hydride to a coordinated substrate, which has been rendered, by virtue of coordination, susceptible to nucleophilic attack by the hydride ligand. Hydrogen activation has been proposed to occur via either homolytic (eq. 1.1a, b) or heterolytic (eq. 1.2) cleavage of H2. M + H2 MH2 2M + H2 • 2MH (1.1a) (1.1b) 2 MX + H2 • MH + H+ + X- (1.2) Recent reports of rj2-H2 complexes (discussed in Chapter 3) have provided structural precedents for an intermediate or transition state for reaction 1.1. Structural precedents for a transition state or intermediate for reaction 1.2 (via direct heterolytic cleavage, rather than by oxidative addition to give MXH2 and subsequent elimination of HX) are not available, apart from Ozin's claim of matrix-isolated, "ligand-free" Pd^-H^.5 Mechanistic pathways proposed for olefin hydrogenation, by mono- or dihydride species, have usually involved T|2-olefin coordination: C M -— C with subsequent hydride transfer giving an alkyl species; further hydride transfer or, in the case of monohydride catalysts, possibly protonolysis or reaction with another metal hydride gives the alkane. Eisenschmid et ah, however, have recendy reported evidence for "pairwise" hydrogen transfer to olefinic substrates, which implies catalysis via a dihydride intermediate, with hydrogen transfers being fast compared to relaxation of the hydrogen nuclear spins in para hydrogen.6-7 Numerous, other hydrogen sources for hydrogenation have been employed (summarized in sect 5.1), and methanol is one such, little studied, example. Hydrogenation employing dihydrogen and methanol as hydrogen sources is reviewed in more detail in the chapters to follow. Chapter 3 focusses on dihydrogen complexes relevant to the H2 hydrogenation systems of Chapter 4, which involve ruthenium phosphines and ferrocenyl phosphines. Chapter 5 is devoted to the activation of methanol, mainly by RuCl2(PPh3)(isoPFA) (isoPFA = chelated l,2-(a-W,/V-dimethylaniinoethyl)(diisopropylphosphino)ferrocene, Fig. 1.1) and hydrogenation systems are described where hydrogen is transferred from methanol to an unsaturated organic substrate. Most of the present work is related to systems involving ferrocenyl 3 ligands. A wide variety of ferrocenyl Uganda containing phosphine,8,9,10'11 arsine,8,11,12 sulfide8,11-13 and arninoalkyl functions8,10,11,14«15,16 has been prepared. A planar element of chirality is conferred by (^substitution of one of the ferrocene rings. In 1970, Gokel and Ugi reported on a simple optical resolution of a-N^-dimethylammoemylferrocene, FA (Fig. 1.1).16 Phosphine derivatives of FA can be prepared via lithiation of FA and subsequent reaction with the appropriate chlorophosphine.8,10,11,17,18 Fortunately, the lithiation is highly stereoselective, the non-resolved FA giving the (SJl)- and (/?,S)-1,2-(a-NjV-d^methylaniinoethyl)(phosphino)ferrocene in 96% yield. (In, for example, the (SJZ)- enantiomer, S refers to the centre of chirality at the a-carbon atom of the aminoethyl group, and R refers to the planar chirality conferred by the (^substitution of one of the ferrocene rings.) Replacement of the NMe2 group10 by hydroxyl, alkoxy or other functions, provides yet other variations on this type of ligand. The variety of derivatives and the accessibility of chiral derivatives have in large part been the reason for the interest in these ligands. Also, they are reasonably stable in air and orange-coloured and can thus be readily chromatographed; they are also readily recrystallized. Fig. a) b) Figure 1.1a) FA; b) PPFA (R = Ph) and isoPFA (R = Pr») 1.1 shows the main ferrocenyl phosphines used in this work, PPFA and isoPFA. Chapter 2 General experimental procedures 2.1 General procedures Unless otherwise specified, air sensitive reagents and products were manipulated under a nitrogen atmosphere, by using a single or double-manifold vacuum system and Schlenk techniques. Some reactions described in the chapters to follow, and requiring inert or hydrogen atmospheres, were carried out in glass tubes sealed by heating with a torch, or by closing a valve. For sealing glass tubes (mcluding NMR tubes) under a hydrogen or nitrogen atmosphere, a single-manifold vacuum line connected to a hydrogen or nitrogen source, and also to a manometer, was used. The tube to be sealed was fitted with a ground-glass joint, and a Teflon valve, if desired. Any required solids were placed in the tube, which was then evacuated on the vacuum line. Any required liquids (freeze/thaw-degassed) were then vacuum-transferred into the tube, over a drying agent when desired. With the tube partially immersed in liquid nitrogen, the vacuum-line manifold and the tube were filled with hydrogen or nitrogen to the desired pressure, as indicated on the manometer. If the pressure were about 10 cm Hg below atmospheric pressure, then the pressure in the tube on warming, and after sealing, would be 2-3 atm. A torch was used to seal the tube, at a constriction below the joint, or the Teflon valve was closed, if present During sealing with a torch, the tube to be sealed was not isolated from the rest of the gas in the manifold; the larger gas volume made for a smaller pressure increase on heating with the torch, thus avoiding positive pressure in the tabe during sealing. The larger volume was particularly important if the pressure in the tube during sealing was just under atmospheric pressure. 6 When pressures of in excess of 3 atm were desired, the set-up shown in Figure 2.1 was used. The gas pressure was adjusted using the pressure gauge, with the tube to be sealed at room temperature, or partially immersed in liquid nitrogen if higher pressures were desired. The Teflon valve (obligatory in this case) attached to the tube was then closed and the tube allowed to warm up to ambient temperature. Figure 2.1 Apparatus for filling tubes with pressures up to 10 atm. For manipulating materials under elevated hydrogen pressures, a tube such as that shown in Fig 2.2 was used. A solid reagent was introduced through the Teflon valve opening, and shaken down to region "A" in the tube, along with a small magnetic stir bar. Solvent was vacuum-transferred into the tube at region "C". With the solvent frozen, the pressure gauge gas metal valve assembly to tube fitted with a Teflon valve 7 pressure was then adjusted as described above, the valve on the apparatus (Fig 2.2) was Teflon valve Figure 2.2 Glass apparatus used for manipulating materials under elevated hydrogen pressures, up to 10 atm. The diagram is approximately to scale, and the height of this apparatus is -40 cm. closed and the tube allowed to warm to ambient temperature. Solvent and solids were mixed and reaction with hydrogen allowed to occur. Precipitates formed could be filtered under hydrogen by tipping the reaction mixture onto the frit at "D" in the tube and cooling the tube at point "B". The filtered solid could be washed with freeze/thaw-degassed solvent, the wash-solvent, for example, being vacuum-transferred into the tube at point "C" (with filtrate at "B" frozen). In this procedure, filtrates that are unstable in the absence of hydrogen are not visibly affected by the filtration, and solvents for crystallization can also 8 be layered by vacuum transfer onto the filtrate in section "B" of the tube. The filtered solid can also be shaken, with the filtrate frozen, or moved with the magnetic stir bar into section "A" for recrystallization. 2.2 msrrumentation Electrical conductance measurements were carried out using a Serf ass Conductance Bridge, supplied by the Arthur H. Thomas Co., with a 3403 conductivity cell from the Yellow Springs Instrument Co. Inc. Infrared spectra were usually recorded on KBr pellets, and solution spectra were recorded using a KBr cell with path length 0.25 mm. Unless otherwise specified, a Nicolet 5DX FTIR spectrophotometer was used. UV-visible spectra were recorded on a Perkin Elmer 552A spectrophotometer, fitted with either of two thermostatted cell compartments, one for 1.0 cm and 1.0 mm cells, and the other for use with a 5 cm cell. Specially designed cells were used for experiments requiring inert or hydrogen atmospheres (Figs 2.3,2.4). The spectrophotometer was equipped with both a Hewlett-Packard 7015B X-Y recorder, used for measuring complete spectra from 315 to 900 nm, and a Houston Instruments OmniScribe® (continuous feed) recorder, used for following rates for longer periods at a selected wavelength. Nuclear magnetic resonance spectra were obtained by using Bruker WP80, Bruker WH400, Bruker-Nicolet HXS 270 or Varian XL300 spectrometers. *H NMR spectra were recorded as 8 (ppm), with positive shifts being downfield from tetramethylsilane. 31P{1H) NMR spectra are recorded as 8 (ppm), with positive shifts downfield from 85% H3P04(aq), and with triphenylphosphine used as the external reference (at 8 -5.6). Hexane/hexene isomer ratios in catalytic hydrogenation studies were obtained by using a Varian 6000 gas chromatograph equipped with a column packed with 20% tricresyl phosphate on chromosorb and an FID detector. Heptaldehyde and heptanol were detected 9 ground 16 mL volume (B) Figure 23 Visible spectrophotometer cell used for kinetic studies under nitrogen to monitor the spectral changes in the reaction of RuCl2(PPh3)(isoPFA) with KOH/methanol in THF. A Kontes Teflon valve and F-145 septum plugs, supplied by Alltech Associates, were used for all Teflon valves and septa, respectively, shown in Figs. 2.2,2.3, 2.4 and 5.2. ground 20 mL volume Figure 2.4 Visible spectrophotometer cell used for absorption measurements under a controlled atmosphere. on the same instrument using an OV-17 column. Styrene/ethylbenzene ratios for kinetic experiments were determined using a Hewlett-Packard 5890 gas chromatograph equipped with a carbowax capillary column, a 3393A integrator and FID detector. Other gas chromatographs used were a Varian 90-P (for preparative GC purposes), and a Hewlett-Packard 5830A, both with TCD detectors. Helium was used as the carrier gas for all gas chromatography. A constant presure gas-uptake apparatus for measuring stoichiometric and catalytic absorption (or evolution) of dihydrogen was used as described in detail elsewhere.19'20 Optical rotation measurements were recorded on a Perkin Elmer 141 polarimeter at ambient temperature using a one decimetre pathlength cell which could hold lmL of solution. The rotations were measured at the sodium-D line (589 nm). The specific rotation of any optically active product, and the optical purity (e.e.) were calculated in a standard manner using the equation: = « x 100 1 X c where [a]Q = specific rotation, at temperature T, measured at the sodium-D line, a = observed rotation(+) or (-). 1 = path length of the cell in decimetres. c= concentration of solution in g/100 mL. Optical purity was calculated using the equation: [a] . Optical purity = reaction x ioo% [a] P«e (2.2) where [ajreaction and [a]pure are equal to the rotation of the reaction product and optically pure compound, respectively. All mass spectrometry was performed by the UBC mass spectrometry service. 11 2.3 NMR studies The Ti measurements were carried out using a standard inversion-recovery pulse sequence (180°-t-90°) on a Varian XL-300 spectrometer operating at 300 MHz. To simulate chemical exchange NMR spectra, the program, DNMR4, QCPE program no. 466, obtained from the Quantum Chemistry Program Exchange, Department of Chemistry, Indiana University, was used. This program was written in the Fortran IV prograrriming language; the DEC 2060 version we received was adapted for the UBC mainframe (UBC computer centre) with assistance from UBC computer centre staff in changing machine-dependent parameters and changing from single to double precision. Sample calculations on test data, obtained from the QCPE, were reproduced, and the calculated spectra were superimposable on those supplied by the QCPE. The T2 values, required as input for this program, were calculated using the equation: T2 = _L-K X W, ,„ l'2 (2.3) where w1/2 is the half-width of the observed peak. AG* values were obtained from Eyring plots for the exchange rate contants, k, obtained from simulations over a range of temperatures. Spin-transfer studies were carried out at 300 MHz on the Varian XL300 spectrometer. To observe spin-transfer between two signals A and B, the decoupler was set at A and decoupler power increased until maximum decrease in integration of signal B was observed, on recording the spectrum at this power. Areas of peaks A and B were monitored with reference to an unaffected peak elsewhere in the spectrum. 2.4 Materials Solvents Tetrahydrofuran was distilled from sodium/benzophenone under a nitrogen atmosphere, stored under nitrogen and used within 48 hours. Anhydrous diethylether, hexanes, n-hexane and benzene, when referred to as dry, were distilled over CaH2 prior to vacuum distillation, and stored under nitrogen. DMA was stirred over CaH2 for at least 24 h, or refluxed for ~1 h prior to distillation under reduced pressure. On one occasion (see section 3.2) DMA was further stirred over RuHCl(PPh3)3 (500 mg/L) and distilled a second time to obtain extra pure material.21 Dimethylacetamide was used within a day or so of distillation. Alcohols, when referred to as dry, were distilled over Mg/l2. Gases Purified helium, hydrogen, nitrogen, argon, and oxygen, from Union Carbide Canada Ltd, and "syn gas" (H2/ CO, 1:1), from Matheson, were usually used without further purification. Hydrogen used in kinetics experiments was passed through a Matheson Deoxo purifier to remove trace oxygen. Phosphines Triphenylphosphine, and tri(p-tolyl)phosphine were used as supplied for synthesis work. Triphenylphosphine was recrystallized from methanol for use in hydrogenation kinetics studies. Racemic and optically active PPFA22 and isoPFA8 were synthesized as described elsewhere from FA (Fig. 1.1) and the appropriate CIPR2. Olefin and ketone substrates The substrates, a-A^-acetamidoacrylic acid, a-/V-acetarrudocinnamic acid and itaconic acid were used as supplied. Styrene, cyclohexene, 1-hexene, vinylmethylketone, ethylmethylketone and acetophenone were passed through 3-5 cm of neutral AI2O3 before use, to remove peroxide impurities. Other materials Proton Sponge®, l,8-bis(<urathylammo)naphthalene was purifed as described elsewhere,21'23 by chrorratographing a hexane solution on alumina. Ruthenium complexes Hydrated ruthenium(IH) trichloride was obtained on loan from Johnson Matthey Ltd. Di-u.-chloro-u^hydrido-Tj2-dihydrogenhy diruthenium(n), Ru2(r|2-H2)H2Cl2(PR3)4 (aryl = R = phenyl, p-tolyl), la, lb. The DMA-free complexes were prepared as described elsewhere,21-23 from [RuCl2(PPh3)2]2 and 1 atm H2 in toluene in the presence of Proton Sponge®, and from RuCl3(P(p-tol)3)2(DMA) and 1 atm H2 in benzene in the presence of Proton Sponge®, respectively. la: Anal. Calcd. for C72H64CI2P4RU2: C, 65.21; H, 4.86. Found: C, 64.95; H, 4.93. 31P{1H} NMR (121 MHz, toluene-ds, ambient temperature, hydrogen atmosphere) 8 70.5,46.0 (lit.21-23: 8c6D6(amWent temperature) 68,43). The lH NMR spectrum, and further information on the 31P{1H} NMR spectrum, are presented in section 3.3, as well as the structure, Table 3.7, Fig 3.2. lb: <5% phosphoras-containing impurity present, based on 31P{1H) NMR (121 MHz, CD2Cl2, ambient temperature, 2-3 atm hydrogen atmosphere) 8 68.8,44.7 (lit.21-23 8toiuene-d8(amDlcnt temperature) 69.2,44.0). Further information on the 3^{^H} NMR spectrum, and the *H NMR spectrum, is presented in section 3.3. The preparation of the bis-DMA solvate, (t|2-H2)(PPh3)2Ru(u.-H)(u-Cl)2RuH(PPh3)2-2DMA, from Rua3(PPh3)2(DMA) and 1 atm H2 in DMA in the presence of Proton Sponge®, has been reported in the literature 21>23 but in the present work this complex was prepared by a different route, from RuCl2Q?Ph3)(isoPFA), as follows: RuCl2(PPh3)(isoPFA) (0.45 g, 0.56 mmol), prepared as described below, was stirred in DMA (3 mL) under hydrogen (10 atm) for 4 d at ambient temperature, in which time an orange precipitate was deposited. This was filtered under hydrogen and washed twice with methanol and once with hexane. The yield was estimated to be ~70 (±20) mg (~40%). Anal. Calcd. for C80H82O2N2CI2P4RU2: C, 64.04; H, 5.20; N, 1.87. Found: C, 63.75; H, 5.43; N, 1.64. The 31P{1H) and *H NMR spectra were the same as for the DMA-free complex. IR vKBr 2129,1970,1903,1822 cnr1, all broad, weak. Di-|i-cMoroaUcMorotetralds(triarylphosphm Ru2Cl4(PR3)4 (R = aryl = phenyl, p-tolyl), 2a, 2b The triphenylphosphine complex was prepared as described elsewhere,21124 from RuCl3(PPh3)2(DMA) DMA and 1 atm H2 in DMA, or by stirring RuCl3(PPh3)2(DMA) DMA21-27 (0.5 g, 0.55 mmol) in n-butanol (2 mL) under hydrogen (8 atm) for 2 d (further discussion of this preparation is given in sect 3.4). In the latter procedure, Ru2Cl2(PPh3)4,2a was deposited as a light brown solid, which was filtered and washed with methanol. Yield: 0.3 g (76%). Anal. Calcd. for C72H60CI4P4RU2: C, 62.07; H, 4.31. Found: C, 61.44; H, 4.50 (discussed in sect 3.4). 31P{1H) NMR (121 MHz, toluene-dg, ambient temperature, under vacuum) 8 60.3 (br s), 52.9 (br s); 31P{ !H} NMR(121 MHz, toluene-dg, ambient temperature, under 2-3 atm hydrogen) 8 60.0 (br s), 52.7 (br s); 8(-58°C) 60.9 (d), 53.3 (d), 2JPP = 42 Hz (lit21 8toluene-d8 (-40°C) 60.0 (d), 52.4 (d), 2JPP = 42 Hz; also, Ut24 oWne^ (-70 °C) 58.8 (d), 53.0 (d), JPP = 41.5 Hz). The tri(p-tolyl)phosphine analogue, Ru2Cl4(P(p-tol)3)4,2b was prepared by stirring 0.25 g (0.16 mmol) RuCl3(P(p-tol)3)2DMA • DMA21*27 in n-butanol (3 mL) under hydrogen (7 atm) overnight. To the resulting brown solution, diethyl ether (-10 mL), hexanes (-15 mL) and methanol (-10 mL) were added. Additions of solvent were interspersed by periods of keeping the solution for several hours in the freezer, at ~-30 °C, to induce crystallization. Some crystals did form during these periods in the freezer. Finally, the volume was reduced to 10 mL, and the brown deposit, including the crystals, was filtered and washed with methanol. X-ray photographs of crystals from this sample indicated poor diffraction. Total yield: 0.046 g (23%). AnaL Calcd-for C84H84CI4P4RU2: C, 64.62; H, 5.38. Found: C, 64.12; H, 5.88. 31P{1H) NMR (121 MHz, toluene-ds, ambient temperature, hydrogen atmosphere) 8 58.8 0>r s), 52.0 (br s) (lit.2i.23 (.40 °Q 60.1 (d), 52.5 (d), 2JpP = 42 Hz). Dichloro{ l-[a-N,N-din ethylarnmc>emyl]-2-diphenylphosphino-ferrocene)(triphenylphosphine)ruthenium(Ii), RuCl2(PPh3)(PPFA), 3a This was prepared us described elsewhere,25 except that an improved reagent ratio (RuCl2(PPh3)3 : PPFA) w^s used (1.24 g, 1.29 mmol: 0.73 g, 1.64 mmol, respectively), thus giving an excess of the soluble reagent, PPFA, in this reaction in which the starting complex, RuCl2(PPh3)3, and product, RuCl2(PPh3)(PPFA), 3a, are suspended solids. Anal. Calcd. for GwH^NC^FeRu: C, 60.36; H, 4.95; N, 1.60. Found: C, 60.54; H, 5.15; N, 1.70. 31P{!H} NMR (121 MHz, C&>6, ambient temperature, under vacuum) 8 77.5(d), 44.3(d), 2JPP = 39 Hz. The previously reported25 values, also in C^D6, of 8 170.5,137.0, "downfield from P(OMe)3", which, in turn, is 141 ppm downfield from 85% H3PC>4, are very different from the values given here and must have been mis-reported The literature difference in chemical shifts for the two resonances is the same as that given here, as is the coupling constant. A 31P{ ^H} NMR peak at 8 23.9 was typically found for solutions of this complex and is identified as OPPha;21-26 this peak was found even when the QD6 had been stirred for ~3 h under vacuum over calcium hydride before vacuum transfer into the NMR tube containing the RuCl2CPPh3)0?PFA)- The optically 16 active complex was prepared using (SJl)- PPFA (optical purity = 84%), which itself was synthesized as described, from FA of the same optical purity22 (see also Chapter 1). DicMorof l-[a-iV,N-dimethylamm ferrocene} (triphenylphosphine)ruthenium(II), RuCl2(PPh3)(isoPFA), 3b RuCl2(PPh3)3 (0.8 g, 0.8 mmol), prepared from RuCl3xH20 and PPh3 in methanol as described,27 and racemic isoPFA (0.56 g, 1.5 mmol) prepared as described,8 from FA prepared from ferrocene carboxaldehyde, were refluxed overnight under nitrogen in a solvent mixture of hexane or hexanes (50 mL) and CHC13 (5 mL). The suspended RuCl2(PPh3)3 gradually reacted to give a green, sparingly soluble product, which was filtered, washed with hexane(s) nnd dissolved in hot benzene. This solution was filtered and the volume reduced to 10 mL. Hexane(s) (50 mL) was then layered onto this solution under nitrogen and crystallization allowed to occur slowly at -30 °C over several days. The dark green product was isolated by filtration and washed with hexane(s). Yield: 0.43 g 56.66; H, 5.99; N, 1.60. Alternatively, recrystallization of the initially precipitated crude product from CH2Cl2/hexane(s) gave a product that was CH2Cl2/hexane(s) to give a product that appeared pure by 31P{ lH} NMR but which was 0.63% low in carbon. Recrystallization of this latter product from THF resulted in a product with acceptable elemental analysis (C, 56.30; H, 6.06; N, 1.68). Celite® was used for the filtrations for recrystallizations in CH2C12 and THF to remove a very fine black, insoluble material that appeared to be produced in the preparation. 31P{1H} NMR (121 MHz, CDC13, ambient temperature, nitrogen atmosphere) 8 96.2(d), 36.6(d), 2Jpp = 32 Hz. *H NMR (300 MHz, CDC13, ambient temperature, nitrogen atmosphere) 8 0.72 (m, 6H, PCHC//3), 1.18 (dd, 3H, 3JPH = 15 Hz, 3JHH = 7 Hz, PCHC//3), 1.43 (d, 3H, 3JHH = 7 Hz, FcCHO/3), 1.75 (dd, 3H, 3JPH = 13 Hz, 3JHH = 7 Hz, PCHC//3), 2.24, 2,82 (2m, 2H, PC//CH3), 2.35, 2.44 (2s, 6H, N(C//3)2), 4.22 (s, 5H, FeC^s), 4.4 (t, IH, JHH = 3 Hz, FeCs^a), 4.58 (bs, 2H, FeC^s), 6.2 (q, IH, 3JHH = 6 Hz, FcC#CH3), 7.37 (m, 9H, Ph meta, para), 7.92 (t, 6H, J = 8 Hz, Ph ortho). The optically active complex was prepared as described above, but using (SJt)-isdPFA of optical purity = 84%. This ligand, in turn, was prepared as described,8 from FA of the same optical purity. Crystals of RuCl2(PPh3)(isoPFA), 3b (with racemic isoPFA), grown by I.R. Butler, were subjected to an X-ray analysis. The data28 reveal a pseudo-octahedral structure with approximately trans chlorides, cis phosphorus donor atoms and an ortho phenyl hydrogen atom in the vicinity of the sixth position. Di-p-chloro-p-hydrido-rj2-dihydrogerihydridobis(triphenylphosphine) {1 -[a-Nfl-dimethy laminoethyl] -2-dii s o p ro p y Iphosphino-ferrocene} diruthenium(n), (n2-H2)(isoPFA)Ru(p-Cl)2(p-H)RuH(PPh3)2,4 All manipulations were carried out under a hydrogen atmosphere. A suspension of the starting complex, 3b (0.334 g, 0.41 mmol) was reacted with H2 (1-4 atm) at room temperature in a benzene/methanol (10:1) solvent mixture (4 mL) in an apparatus (Fig. 2.2) designed to minimize chances of atmospheric contamination. Reaction occurred in 4 -10 days to give a brown solution and an orange precipitate. Because of the solubility limitation, larger amounts required longer reaction times. The rate also appeared to depend on methanol concentration. The precipitate was filtered and washed with methanol and benzene. A more detailed description of the experimental procedure is described in sect 4.2. Yield: 0.094 g (39%). Anal. Calcd. for C56H66NCl2P3FeRu2: C, 57.29; H, 5.62; N, 1.19. Found: C, 57.37; H, 6.02; N, 1.39. ^Pf^H} NMR (121 MHz, CD2CI2, ambient temperature, 2-3 atm hydrogen atmosphere) 8 -78.5 (two overlapping doublets with 2Jpp = 25 or 40 Hz, PPh3 trans to CI), 77.2 (br s~possibly two br s overlapping, isoPFA), 64.6 (two overlapping doublets with 2JpP = 32 or 40 Hz, PPh3 trans to u\-H). lH NMR (300 MHz, CD2CI2, ambient temperature, 2 - 3 atm hydrogen atmosphere) 8 -18.6 (dd, 2JPH = 28, 32 Hz, Ru-H), -12 to -17 (bs, Tl2- H2 and u-H exchanging), 0.3, 0.39, 1.5, 1.8 (4dd, 12H, 3JPH = 15 Hz, %H = 7 HZ, PCHCH 3), 1.48 (d, 3H, 3JHH = 7 Hz, FcCHC// 3), 2.35, 3.1 (2s, 6H, N(C// 3)2), 4.1 (s, 5H, C5H 5), 4.08,4.32, 4.42 (3s, 3H, C$H 3), 4.50 (q, 1H, 3JHH = 7 Hz, Pc-CH CH3) (Signal at 4.50 resolved in CDCI3), 6.9-7.2 (m, 18H, Phmeta, para), 7.4, 7.55 (2t, 12H, JHH = 9 Hz, Ph ortho). Variable temperature experiments (Fig 3.25)' and NMR discussion (sects. 3.3, 3.4) are presented later. About 14% of the product appears to comprise two other isomers, with 31P{1H} signals comparable to those given above, but slightly shifted: 79(two overlapping doublets), 77.6(s, possibly two s IR v*8' 2109, 2025 cm-*. Crystals of (n2-H2)(isoPFA)Ru(u-Cl)2(u.-H)RuH(PPh3)2*C6H6 suitable for an X-ray structure determination were obtained by recrystallization from CH2Cl2/hexane/benzene (sect. 4.2). The crystal structure (Table 3.8, Fig. 3.4) is described later. Carbonylchlorohydrido{ 1 Ia-methoxyethyl]-2-dhsopropylphosphinofe moce ne} (triphenylphosphine)ruthenium(II), RuHCl(CO)(PPh3)(isoPOF), 5 A suspension of the starting complex, 3b (0.245 g, 0.3 mmol), was refluxed under nitrogen in methanol (5 mL) for 7 d. During this time, the green starting complex, itself only sparingly soluble, reacted to give a yellow precipitate. The air-stable precipitate was isolated by filtration and washed with acetone. Yield: 0.16 g (67%). The product was recrystallized from mcWoromethane/methanol. Anal. Calcd. for Css^sChClPiFeRu: C, 57.93; H, 5.72; N, 0.0; O, 4.07; Cl, 4.5. Found: C, 57.27; H, 5.59; N, 0.0; O, 3.90; Cl, 4.5. 31P(1H) NMR (121 MHz, CDCI3, ambient temperature, nitrogen atmosphere) 8 46.0 (d), 46.4 (d), 2JPP = 188 Hz. lH NMR (300 MHz, CDC13, ambient temperature, nitrogen atmosphere) 8 -13.9 (t, 1H, 2JPH = 19 Hz, Ru-//), 0.83 (dd, 3H, 3JPH = 16 Hz, 3JHH = 7 Hz, PCHC//3), 1.14 (d, 3H, 3JHH = 8 Hz, FcCHC//3), 1.81 (dd, 3H, 3JPH = 18 Hz, 3JHH = 8 HZ, PCHC//3), 2.36 (m, 1H, PC//CH3), 2.38 (s, 3H, OC//3), 3.3 (m, 1H, PC//CH3), 4.24, 4.28, 4.30, 4.32 (4s, 8H, FeCs/Zs, FeC^), 6.52 (q, 1H, 3JHH = 7 Hz, FcC//CH3), 7.38 (m, 9H, Ph meta, para), 7.71 (m, 6H, Ph ortho). IR vKBr(Ru-H): 2007 cm-!(w); v^CO) 1923 cm-Kstr). Crystals suitable for an X-ray structure determination were obtained and the crystallographic data are given later (Tables 5.1,5.2, Fig. 5.6). Carbon ylcMorohydrido {1 - [a-N,N-dime&ylam phosphmofemx:ene}(triphenylphosphme)ramenium(ir), RuHCl(CO)(PPh3)(isoPFA), 6 A suspension of the starring complex, 3b (0.06 g, 0.07 mmol) was refluxed with 25 uL of a 3 M methanolic solution of KOH (~0.07 mmol) in 2 mL of methanol, till no more solid, sparingly soluble, green starting material was visible and a reddish-yellow solid had been deposited (-30 min). The precipitate was filtered from the red supernatant under an inert atmosphere and washed with methanol. Yield of crude product: 0.018 g (25%). The efficiency of this preparation, in terms of product selectivity, is enhanced if the RuCl2(PPh3)(isoPFA), 3b is very finely ground. Larger crystals are slow to dissolve and take longer to react Secondary reactions, fortunately 10 to 100 times slower than the formation of 6, do occur, hence, ihe reaction must be stopped before significant amounts of the secondary products form, yet after the starting material has disappeared. A sample of the product was recrystallized from metiianoVdichloromethane. Anal. Calcd. for C39H460NClP2FeRu: C, 58.64; H 5.76; N 1.75; Cl, 4.4. Found: C, 57.52; H, 5.86; N, 1.70; Cl, 5.5. 31P{1H} (121 MHz, CDC13, ambient temperature, nitrogen atmosphere) 8 50.0 (d), 31.9 (d), 2JPP = 290 Hz. This coupling is confirmed by both 300 MHz and 400 MHz spectra. The 31P{ 1H} NMR spectrum showed another set of doublets, with varying intensity (from ~10% relative to the above, up to almost equal intensity), and with chemical shifts differing by 8 -0.15, and with the same 2Jpp value. *H NMR (300 MHz, CDCI3, ambient temperature, nitrogen atmosphere) 8 -15.3 (dd, 2JPH = 30,20 Hz, Ru-H), 0.82 (dd, 3H, 3JPH = 12 Hz, 3JHH = 6 Hz, PCHO/3), 122 (dd, 3H, 3JPH =12 Hz, 3JHH = 6 Hz, PCHC//3), 1.26 (d, 3H, 3JHH = 6 Hz, Fc-CHCi/3), 1.68 (m, 6H, PCHC//3), 2.3 (br s, 6H, N(C//3)2, resolved from 0 °C down to -50 °C into two separate singlets at 2.21, 2.38 ppm), 2.5 (m, IH, PC//CH3), 3.28 (m, IH, PC#CH3), 4.2 (s, 5H, FeC5//5), 4.28, 4.35, 4.41 (3s, 3H, C5//3), 5.48 (q, IH, 3JHH = 5 Hz, Fc-C//CH3), 7.35 (s, 9H, Ph meta, para), 7.72 (m, 6H, Ph ortho). Another rninor (-10%) high-field *H NMR hydride resonance always accompanied that described above: -15.8(f) with 2JPH = -23 Hz. IR v^'tRu-H) 2079 cm-i(w); v^CO) 1879 cm-l(str). A single crystal X-ray structure determination (Tables 5.1, 5.2, Figs. 5.9,5.10) was carried out on crystals grown as described below. Chlorohydrido {1 - [a-N//-dimethylaminoethyl] -2-diisopropyl-phosphmoferrocene}(triphenylphosphme)rathenium(II), RuHa(PPh3)(isoPFA), 7 The starting complex, 3b (0.08 g, 0.1 mmol), and Proton Sponge® (0.023 g, 0.1 mmol) were placed under nitrogen in methanol (1-2 mL) for 5 months. Large black-red crystals (-20 mg) of 7 grew in this time, as well as yellow crystals of 6. The amount of 7 was difficult to measure as some red crystals had grown around the crystals of starting material. This mixture of crystals of 3b, 6, and 7 was separated from the solution by filtration or decanting and washed with methanol and dried. The large black-red crystals of 7 were then isolated by handpicking them out of the mixture. Anal. Calcd. for C38H48NClP2FeRu: C, 59.06; H, 6.22; N, 1.81. Found: C, 58.38; H, 6.00; N, 1.80. 31P{ !H} NMR (CDC13,121 MHz, ambient temperature, under vacuum) 8 98.0 (d), 61.5 (d), 2JPP = 40 Hz. lU NMR (CDCI3, 300 MHz, ambient temperature, nitrogen atmosphere) 8 -23.1 (dd, 2JPH = 40,27 Hz, RuH). IR vKB'tRuH) 2028 cm-i(w); a peak at 1897 cm_1(w) was probably due to the presence of some 6. A single crystal X-ray structure determination was carried out and is discussed later (Tables 5.1,5.2, Fig. 5.12). Chapter 3 Characterization and reactivity of seme diniaclear rethemurn(II) phosphine complexes. containing T|2-H2 3.1 Introduction Historical background The possibility of a transition state, involving dihydrogen molecules interacting with a metal centre, has been considered over the past thirty years.29 In 1976, Ashworth and Singleton30 cited indirect evidence for stable H2 binding. These workers observed that all the reactions of RuH4(PPh3)3 and RuH3(dppe)2+ proceeded by initial loss of H2, and by analogy with the similar behaviour of reversibly bonded dioxygen compounds, they proposed that these complexes could be regarded as dihydrogen adducts of ruthenium(EI). This appears to be a weak analogy, however. Direct proof for such binding in these and other reported systems was lacking until about 1983.29 In 1980, Kubas reported31 that the infrared spectra of the complexes, ''M(H2)(CO)3(PR3)2 (M = Mo, W; R = Pr', Cy), as well as the lability of dihydrogen in these complexes, suggested that the H2 bonding might be unusual. Eventually, Kubas' group obtained neutron diffraction and X-ray crystal structures of W(H2)(CO)3(PCy3)2, and the neutron structure showed a dihydrogen moiety. The !H NMR spectrum of the HD complex showed that the HD coupling constant was large, and similar to that for gaseous HD, thus providing the most convincing evidence for the presence of bound dihydrogen. These NMR and diffraction data were first reported in 1983.32-33 The next stable H2 complex to be reported was [IrH(H2)(bq)(PPh3)2]+, by Crabtree etal.34 Their !HNMR Ti measurements indicated the binding of molecular hydrogen. This criterion is discussed below. Subsequently, Morris et al. reported35 that the complexes, [MH(H2)(dppe)2]+ (M = Fe, Ru) contained bound dihydrogen; the X-ray crystal structure showed a short H-H distance for the H2 moiety, and a large HD coupling constant for the HD complex confirmed the presence of bound dihydrogen (further discussion of this criterion appears below), consistent with Ashworth and Singleton's speculation about the Ru complex.30 Recent articles, including reviews, cite numerous other examples of dihydrogen complexes.2936,37 while some such species are stable only at low temperature, others exist at ambient temperatures. A number of previously known polyhydrides are now recognized as containing the T|2-H2 ligand; among these is RuH2(H2)(PPh3)3 38 again, as previously suggested.30 Factors effecting the stability ofrf-Hi complexes of the transition metals Molecular hydrogen complexes have thus far been reported for the transition metals shown in Table 3.1,29,37 and include examples for the electron configurations shown in Table 3.2. Complexes stable at room temperature, if necessary under 1 atm hydrogen, have been found only for d2, d4, d6 and d8 configurations. The coordination numbers29,36-37-^ (counting T)2-H2 as one) for complexes along with their electron Table 3.1 Transition metals which have been found in molecular hydrogen complexes. Cr Fe Co Ni Cu Mo Ru Rh Pd W Re Os IT configurations are shown in Table 3.3. All known dihydrogen complexes (the count being over 50, to date) are 18 electron species, considering T)2-H2 as a 2-electron donor. Other points to note are:29 all known complexes are neutral or cationic, and none are anionic; all contain rc-acceptor ligands, except for the matrix-isolated, "ligand-free" PdH2 and the small cluster complexes, Cu2H2(H2)x and Cu3(H2); and most examples are octahedral. Table 3,2 Electronic configurations and oxidation states of examples of T|2-H2 complexes. electron oxidation group example(s) configuration state number d° VI 6 [MoH4(Ti2-H2)(dppe)2]2+ 39 d2 V 7 ReH5(Ti2-H2)(PPh3)2, ReH5(ri2-H2)(dppe)36 d4 H 6 M(H)(Ti2-H2)Cp(CO)2 (M = Mo, W)40 d4 IV 8 [OsH3(Tt2-H2)(PPh3)3]+ d6 n 8 many examples,29*36,37 e.g., [MH(T|2-H2)(dppe)]+ (M = Ru, Fe)35 d8 I 9 [Rh(ri2-H2){P(CH2CH2PPh2)3}]+,41 Co(Ti2-H2)R(CO)3 (R = H, CH3)42 d10 0 10 Ni(ri2-H2)(CO)343 Table 3.3 Coordination numbers of rj2-H2 complexes, according to electron configuration coordination electron number configuration 9 0 (only one d° complex proposed to date) 8 2 7 4 (taking ri5-C5H5-, in M(H)(Ti2-H2)Cp(CO)2 (M = Mo, W), as occupying three coordination sites) 6 6 5 8 4 10 (exceptions mclude only the niatrix-isolated "hgand-free" Pd^^H^ and small clusters, such as Cu3(ri2-H2)) Theoretical considerations Kubas' group29 has considered the question of whether steric or electronic factors are more important in determining whether H2 binds as rj2-^ or as the dihydride. Within the series of complexes, MoH2(CO)3(L-L), where L-L is a chelating, ditertiary phosphine ligand, the basicity of the phosphine (Le. electronic factors) appears to be more important in determining the mode of H2-binding. The more basic, alkyl ditertiary phosphines promoted dihydride formation, whereas dppe (with phenyl groups) permits r|2-H2 coordination. It had been observed that stable dihydrogen complexes contained bulky ligands, and it was earlier thought that steric factors were more important; Kubas, however, concludes from the data on the Mo series that electronic factors are more important. But the fact remains that all room-temperature-stable dihydrogen complexes do contain bulky ligands. It seems reasonable that bulky ligands, in some systems, might promote dihydrogen coordination over dihydride formation if such a reduction in coordination number would better accommodate the bulky ligands. There are insufficient experimental data to discuss the factors meaningfully. A study44 on protonation of the series of complexes, [RuH(Cp)(Ph2P(CH2)nPPh2)]+ (n = 1-3) has demonstrated the importance of chelate ring size in determining whether dihydride or r|2-H2 ligands are produced in this system. The dppm ligand, n = 1, promoted the formation of the dihydrogen complex exclusively, while dppp, n = 3, promoted only dihydride formation. The dppe complex, on protonation, yielded a mixture of the two types. An insight into how electronic factors might influence the mode of H2 bonding is provided by theoretical treatments.29-45'46'47 The favoured model for the bonding is shown in Fig. 3.1 Molecular orbital calculations46 indicated that donation of o electrons to the metal atom provides the main contribution to the bonding, whereas the contribution of back-bonding from a metal to the O* orbital of the T|2-H2 moiety is small. In the crystal structure of W(rj2-H2)(CO)3(P(Pri)3)2, the H2 is oriented so that such n interaction is possible.46 Calculations also confirm47 that rc-acceptor ligands enhance the M— H2 interaction, whatever the metal, consistent, for example, with the experimental observation for the MoH2(CO)3(L-L) complexes. Burdett et al.41 note, in particular, that a 7t-acceptor ligand trans to the r|2-H2 ligand stabilizes the bonding mode, by reducing the extent of back-bonding into C*(H2). However, complexes such as M(r|2-H2)(CO)5 (M = Cr, Mo) are unstable with respect to loss of H2, inspite of the number of 7t-acceptor (CO) ligands; the back-bonding to c*(H2) may, in these cases, be too small.29 It seems that some back-bonding is necessary for stability.29 Thus, d° complexes are expected to be unstable.36 Figure 3.1 Orbital diagrams showing the contributions to the bonding for the M-(T|2-H2) structure. The following summary may be made of factors determining whether or not a ri2-H2 complex may be formed: a) The metal There is not yet sufficient reason to exclude other transition metals (e.g., group IV, V or even XI and XII metals) as candidates for ri2-H2 complexes, but generalizations are possible. Classical hydrides are more likely for me early transition metals; for example, WH6(PR3>3 is classical, but ReH7(PR3)2 contains a dihydrogen ligand. Also, classical hydrides are more likely for third-row metals; for example, OsH4(PR3)3 is classical, while the Fe and Ru analogues contain dihydrogen. b) The 18 electron rule It may not be accidental that all T|2-H2 complexes are 18-electron species. If there were, say, a vacant site on a transient 16-electron dihydrogen complex, a second dihydrogen molecule (if available) would be likely to coordinate, H because any metal centre that can stabilize one rj2-H2 would probably be able to stabilize another. Although not all the ri2-H2 complexes are formed by addition of free H2 to unsaturated precursors (some are formed by protonation of hydrides), free dihydrogen is always generated in situ in these systems. That is, coordinated dihydrogen is labile and the coordination reversible.29 On the other hand, there may have been insufficient attempts to protoriate 16-electron hydrides. c) Electron configuration So far it appears that an even-numbered dn configuration is required, and d6 is preferable. d) Ligand type The presence of 7t-acceptor ligands helps to stabilize these complexes, and bulky ligands may favour a dihydrogen complex over a dihydride. f) Structure Octahedral geometry seems to be preferred. Criteria for recognition of dihydrogen complexes The VHH mode in the infrared spectrum has attracted more interest than other IR vibrational modes for the purpose of characterization, but only a few complexes have exhibited the assigned broad weak band in the range 2400 to 3100 cm-1 (Table 3.4). Often the band is either obscured by other bands or is too weak to be seen. In a few cases29 other frequencies have been assigned as due to vibrations of the MH2 unit (e.g., va(MH2) 1345 - 1568 cm-1, vs(MH2) 868 - 960 cm-», 8(MH2) 400 - 471 cm-*). The usefulness of infrared and Raman spectra has been quite limited. Prior to the publication48 of the structure determined in the present work and shown later in Figure 3.4, neutron and X-ray structure determinations had been successful in locating the TJ2-H2 ligand in W(Ti2-H2)(CO)3(PPrS)2 and [FeH(ri2-H2)(dppe)2]+. In W(Ti2-H2)(CO)3(PPri3)2, the H-H distance is given as 0.75 A49 by X-ray, 0.84 A49 and 0.82(1) A50-5* by neutron diffraction. In [FeH(rj2-H2)(dppe)2]BF4, the H-H distance is 0.89(11) A (X-ray); a neutron diffraction study of a crystal of the BPI14* salt yields 0.815(20) A.52.53 The X-ray structure48 (Fig. 3.4) of (isoPFA)(ri2-H2)Ru(u-H)(u-28 Table 3.4 Vibrational Frequencies, v^cnr1), for Dihydrogen Complexes Complex VHHCcm-1) reference W(r|2-H2)(CO)3(PCy3)2a 2690 29 Mo(r|2-H2)(CO)3(PCy3)2a 2950D 29 Mo(ri2-H2)(CO)(dppe)2a 2650 29 W(Tl2-H2)(CO)5b 2711 29 Co(ri2-H2)(CO)2(NO)C 3100 29 FeH2(Tl2-H.2)(PEtPh2)3a 2380 36,54 FeH2(ri2-H2)(PBuPh2)3a 2400 58 aNujol mull. bLiquid xenon solution at -70 °C. cXenon matrix at -261 °C. Calculated, based on the observed frequency for the D2 complex. Cl)2RuH(PPh3)2,4 has revealed the rt2-H2 ligand, with an H-H distance of 0.80(6) A. There are only three crystal structures to date showing the T|2-H2 ligand; disorder and other problems have frustrated attempts to locate the T|2-H2 unit in other crystal Structures. 23,55,56,57,58 Data from NMR studies, particularly JHD coupling constants and Ti measurements, have been used extensively for the study of T|2-H2 complexes. Formation of the corresponding TJ2-HD species using HD instead of H2 in the synthesis, or using deuteriation of an rj2-H2 complex, allows for observation of the HD coupling constant (Table 3.5). The value of JHD is an order of magnitude larger than that for compounds containing classical, non-linked hydride and deuteride ligands, and is comparable to that for HD gas (43.2 Hz). The value of JHD has been also used as a measure of the interaction of the HD molecule with the metal centre, lower values of JHD being indicative of stronger interaction. Sometimes, however, in fluxional complexes, this coupling is obscured, and Crabtree's Table 3.5 Some *H NMR HD Coupling Constants (Hz) for T|2-H2 Resonances3 Complex JHD reference W(HD)(CO)3(PPri3)2 33.5 49 trans -[OsH(HD)(meso-tetraphos-l)]BPh4 26.4 37 trans -[FeH(HD)(meso-tetraphos-l)]BF4 32.3 37 [Rh(HD)(tetraphos-2)](02CCF3) 18 54 Mo(HD)(CO)(dppe)2 34 57 Ru(HD)Cp(dmpe) 22 59 [Ru(HD)Cp(PPh3)(CNBut)]PF6 28.6 60 [Ru(HD)Cp(dppe)]PF6 24.9 44 [Ru(HD)Cp(dmpe)]PF6 21.9 44 [RuD(HD)(dppe)]+ 32.9 61 [RuD(HD)(depe)]+ 32 61 [FeD(HD)(dppe)]+ 30 61 [FeD(HD)(depe)]+ 28 61 [Ru(HD)Cp(CO)(PR3)]+ (PR3 = PPh3, PMe3) PMe2Ph, PCy3) 28-30 59 [IrH(HD)(bq)(PPh3)2]+ 29.5 62 [IrH(HD)(bq)(PCy3)2]+ 29 62 rFeD(HD)(Ph2PCH2CH2PEt2)]+ 32 63 '"of. HD gas, 43.2 Hz49. The JHD values for deuteride-hydride complexes are an order of magnitude smaller (e.g., JHD < 1 Hz for [RuHDCp(dppe)]+ 44). coupling is obscured, and Crabtree's group36-62 introduced the use of another NMR criterion, the *H NMR longitudinal relaxation time, Tx, for the ri2-H2 resonance. Because the H-H distance, rim, for the r|2-H2 ligand is relatively short, the NMR relaxation time, Ti, is also short (<100 ms.) The relaxation mechanism is usually "dipole-dipole" for protons less than 2 A apart36, and for this mechanism, the rate of relaxation, ROOD), is related to the internuclear distance generally by the equation:36 R(DD) = PiCDD)]-1 = 0.3Y4h2r6(2n)-2J(co) R(DD)=[T(DD>r1= 0.3Y4hVW2r "^TT + ^"TT" 1 1 • L(i + co2t2) (1 + 4(o2x2) J c c (3.1) where J(co) = spectral density, y = gyromagnetic ratio, h = Planck's constant, r = internuclear distance, Tc = rotational correlation time and co = Larmor frequency. Thus, Ti is proportional to runr6 and the range of Ti values for T)2-H2 atoms is roughly an order of magnitude less than the range for coordinated hydride ligands. Equation 3.1 predicts a Ti minimum when the Brownian motion is best matched ' with the Larmor frequency, co; ie., when: tc = 0.63/co (3.2) where xc is a measure of the Brownian motion. Thus, if a Tj minimum can be observed, by a series of Ti measurements at various temperatures, xc can be calculated by using equation 3.2, and hence, I"HH can be calculated, the other factors being known.36 Table 3.6 lists some Ti and THH values36 estimated by this method, which assumes dipole-dipole relaxation to be the only mechanism Thus, the Ti value is both diagnostic for showing the presence of the ri2-H2 ligand and useful for estimating I*HH for it. Morris et al.52 have discussed the need for a refinement of the minimum Ti method, because of the effect on the Ti by contribution of another mechanism in addition to that related to the tumbling of the complex. The refinement is an attempt to include the effect, via dipole-dipole relaxation, of the rotation of the Trlj molecule about the M-H2 bond axis, and defines two correlation times: TH. for the tumbling of the complex as a whole, and XH2« fOT me additional motion of the spinning H2. The H-H (H2) distances for the two complexes, [FeH(H2)(dppe)2]+ and W(H2)(CO)3(PPrS)2, for which solid state H-H Table 3,6 Selected Ti (minimum) values8 for various hydride and dihydrogen complexes complex temperature Ti (ms) rHH(A) fHH(A) (K) (minimum) (from TO (from solid state) classical hydrides: WH6(PMe2Ph)3 235 181 1.74Cb WH5(PMePh2) 240 179 1.74C [frH2(CO)2(PPh3)2]+ 210 245 1.81C [ReH8(PPh3)]- 200 245 1.83C dihydrogen complexes: [FeH(Ti2-H2)(dppe)2]+ 210 <8.5 < 1.05C 0.89(1 l)c, 0.815(20)d W(T]2-H2)(CO)3(PPri3)2 193 <5 < 0.96C 0.75(16)c, 0.82(l)d RuH2(T|2-H2)(PPh3)3 266 30 1.16C [IrH(H2)(bq)(PPh3)2]+ 200 8 1.04C 0.92e ReH5(ri2-H2)(PPh3)2 200 110 1.20C aFrom reference 36; rnH calculated from equation 3.1. bHamilton and Crabtree36 have introduced the empirical constant C to bring calculated THH values into line with known solid state values; they have suggested the value 0.9 for dihydrogen complexes and 1.0 for classical hydrides. cFrom X-ray diffraction data. dFrom neutron diffraction data. eFrom solid state NMR data. distances are known (vide supra), were calculated by using the following modified expression for the spectral density, J(co), at a particular frequency, cu (= 27rv):52 H J(co) = 0.25 H 1 + art 2„2 + 0.75 (1 + 2a) H 1 + coz0 2/ H__)2 1 + 2a where a = TH/*H2-(3.3) (3.4) Reasonable H-H distances (run) for the H2 moiety are obtained when the correlation time for the spinning motion of the H2-ligand is assumed to be much faster than that for the complex as a whole; i.e.: TH2 « *H. and a ~ 0. (3.5) For this case, the spectral density function becomes: 0.25T J(co) = H 1 + G)V (3.6) The value of rnH, obtained using this expression for J(co) in equation 3.1, differed from the result using Crabtree's method, by a factor of (0.25)1/6, or ~0.8. Independent evidence for this spinning motion has been reported,51'52 and the spin rate was measured for the complex, W(D2)(CO)3(PPri3)2, as being >130 kHz.51 Dinuclear, rf-fy complexes reported in the present work Interest in Ru complexes containing monodentate tertiary phosphine ligand systems has grown out of ongoing efforts to characterize possible intermediate hydrides in catalytic olefin hydrogenation systems. Interest in the dinuclear, ri2-H2 complexes to be described arose, in particular, out of efforts by others in this department25'64-65 to characterize hydrides formed in hydrogenation systems employing RuCl2(PPh3)(PPFA), 3a, and RuCl2(PPh3)(isoPFA), 3b, as catalyst precursors. Initial efforts by Rodgers et al.25 had failed to detect hydrides in the reaction of 3a with hydrogen at pressures of <1 atm. This thesis reports the formation of three dihydrogen complexes in various systems, including the complex (Ti2-H2)(PPh3)2Ru(p-H)(p-Cl)2RuH(PPh3)2, la, previously formulated with all hydrogen ligands as classical hydrides. The characterization and reactivity of the complexes la and lb have been reinvestigated in the present work. Prior to the discovery and initial characterization of complexes 1, by Dekleva, Thorbum and James,23 a monohydride bisphosphine intermediate was believed to be an active species in the hydrogenation of olefins catalyzed by the Ru(II) trisphosphine complex, RuHCKPPl^2-4'66'67'68'69 (further discussion appears in chapter 4). Related to these studies is the interest in possible hydride intermediates important in the activation of hydrogen by DMA solutions of the Ru(HT) complexes, R11X3L2 (X = Cl, Br, L = PPI13, ASPI13).69 A series of hydrides, initially characterized as [HRuXL^h (X = Cl, Br, L = PPh3, AsPh3), were isolated69-70 and their catalytic behaviour investigated70-71 (again, discussed further in chapter 4). The preparative reactions were initially69 written as equation 3.7, the measured H2 uptake corresponding to 1.5 moles H2 per Ru: RuX3L2 + 1.5H2 • i/2[HRuXL2]2 + 2HX (3.7) A few years later, Dekleva et al.21^3 discovered that a subsequent slower H2 absorption occurred, and that, finally, a total of 2 moles H2 was absorbed per Ru. The products were consequently formulated, correct in terms of stoichiometry, as [RuH2XL2]2, and the reactions written as: RuXsI^ + 2H2 i/2[RuH2XL2]2 + 2HX (3.8) (X = Cl, Br, L = PPh3, P(p-tol)3, AsPh3) A series of chemical reactions supported this formulation and, in addition, a crystal structure of the complex with X = Cl and L = P(p-tol)3 was obtained,21-23'71 although the hydrogen ligands were not located. Dekleva et al.23 also rationalized the !H and 31P{ JH} NMR spectra, and the visible spectra, of two of the systems (X = Cl; L = PPI13, P(p-tol)3), based on the structure of the p-tolyl system (see Fig. 3.3, later). Although in his thesis work, Dekleva21 alluded to the possibility that lb might be an T|2-H2 complex, no experimental data to verify this were obtained. The present studies, to be described in this chapter, show that this type of complex contains one terminal and one bridged hydride, and a r|2-H2 moiety. The characterization and reactivity of (isoPFA)(r|2-H2)Ru(u.-H)(u-Cl)2RuH(PPh3)2, 4, are also presented in this chapter. The discovery of this complex arose out of a search for possible hydride intermediates in the catalytic hydrogenation system: 1 hexene R"Cl2(PPh3)(isoPFA) 1-hexene DMA, 30 °C hexane Chapter 4 contains a detailed account of this search, the background of the work, and the relevance of the findings to the catalytic hydrogenation. In this chapter, the focus is on the complex itself, and its place in the series of dinuclear dihydrogen complexes examined in the present work. Also presented in this chapter are two newly discovered adducts of dihydrogen and the complexes Ru2Cl4(PR3)4 (R = Ph, p-tol), 2a and 2b, respectively, formed in situ in solution. The history of the dimeric complexes 2 is closely linked with that of RuCl2(PPh3)3, which has received considerable attention2"4 as a catalyst for a wide variety of organic reactions, and as a precursor for RuHCl(PPh3)3. Early reports27-72-73 of a solvated, monomelic bisphosphine species, resulting from the dissociation of PPh3 from RuCl2(PPh3>3, in various solvents, were later disputed.74-75 On the basis of 31P{1H} NMR evidence (OCH2C12(-97 °C) 58.8 (d), 53.0 (d), 2JPP = 41.5 Hz), Hoffman and Caulton described the bisphosphine species as the dimer; in the non-coordinating solvent, CH2C12, a monomer would be expected to be square planar or tetrahedral, thus giving a 31P{1H} NMR singlet26 These workers also suggested the possibility of a polymeric species in the solid state,26 and other workers76-77 have discussed the possibility of polymeric species in concentrated solution. James and coworkers24-70-78 subsequently reported the isolation of this species, from the reaction of RuCl3(PPh3)2(DMA)-DMA with hydrogen in DMA; the DMA-free complex 2a was precipitated by addition of methanol. A molecular weight measurement in toluene (Calcd,: 1390; Found: 1200) confirmed the dimeric formulation. In DMA at 25 °C, 2a exhibited a singlet at 8 66.7 in the 31P{ iH} NMR spectrum, and this singlet was initially interpreted24'70 as indicating a monomelic species, RuCl2(DMA)2Q?Ph3)2; Dekleva21 later showed that at low temperature an ABCD spin pattern was the major feature of the 31P{1H} NMR spectrum (8DMA/toiuene-dg(-70 °C) 52.88 (d, PA), 48.34 (d, PB), 42.67 (d, Pc), 42.22 (d, PD), JAB = 29.9 Hz; JCD = 36.8 Hz), and he attributed this pattern to the complex, Ru2Cl4(DMA)(PPh3)4, thought to be a triply chloro-bridged dinuclear molecule, with DMA coordinated and both Ru atoms being six-coordinate. In the 2a preparation described above, the DMA evidently dissociates during the methanol precipitation. Studies in this department have been carried out on the reactions of RuCl2(PPh3)3 with hydrogen to give RuHCl(PPh3)3 and, because of the complicated solution behaviour of RuCl2Q?Pn3)3, studies were also done in the presence of excess phosphine and/or chloride. In the presence of excess chloride, solutions of RuCl2(PPh3)3 contain the triply chloro-bridged Ru2Cl5(PPh3)4- as the main species.21*23 To "probe the intermediate stages" of the reaction: Ru2Cl5(PPh3)4- + H2 + 2PPh3 • 2RuHCl(PPh3)3 + 2HC1 + Or (3.10) Dekleva21 studied the reactivity of 2a toward hydrogen in the absence of PPI13, and he reported this complex, in the absence of added base and in a non-basic solvent, to be inert toward hydrogen; but in the presence of Proton Sponge® in DMA or toluene-dg, the reaction proceeded to give [RuH2Cl(PPh3)2]2. now known to be a T|2-H2 complex (vide supra): [RuCl2(PPh3)2]2 + 3H2 base »[RuH2Cl(PPh3)2]2 + 2HC1 (3.11) This reaction in toluene was used in the present work to synthesize the DMA-free complex. In the course of the present work, it was discovered that the complexes 2a and 2b do show some reactivity toward H2 at 2 - 3 atm in toluene solution. 3.2 Experimental Crystal structures (Ti2-H2)(PPh3)2Ru(^-H)(u-a)2RuH(PPh3)2toluene-d8DMA: Dark red crystals of this complex grew in a NMR tube containing a toluene-dg solution of the bis-DMA solvate of the complex under 2-3 atm H2. Anal. Calcd. for a toluene-dg-DMA solvate, C83H81ONCI2P4RU2: C, 65.89; H, 5.36; N, 0.93. Found: C, 65.73; H, 5.47; N, 0.00. The bis-DMA solvate was prepared from RuCl2(PPh3)(isoPFA), 3b, as described in section 2.4. A single crystal X-ray diffraction study was carried out by J. Simpson, University of Otago, New Zealand, in the hope of fmding the T|2-H2 ligand.55 (Ti2-H2)(isoPFA)Ru(u-H)(u-Cl)2RuH(PPh3)2C6H6: Crystals of the benzene solvate of 4, suitable for X-ray analysis, were obtained by means of the procedure described in section 4.2. The crystal structure determination was performed by J.-P. Charland at N.R.C., in Ottawa, and the structure has been published.48 Visible spectrophotometry To confirm the reported visible spectrum of la,23 specially purified DMA (distilled from RuHCl(PPh3)3,21 see section 2.4) was used for extinction coefficient measurements. The DMA-free complex, la, was prepared as described (section 2.4). NMR spectra Complex la: For variable temperature lVL NMR and 31P{ NMR spectra for simulation purposes, the complex, la, was generated in situ under hydrogen from 2a and Proton Sponge® (cf. section 2.4). Low temperature Ti measurements for the hydrogen ligands were carried out on the bis-DMA solvate in toluene-dg solution under hydrogen using the inversion-recovery method. The Tx measurements above 0 °C were carried out on the in ttta-species generated both from 3a in butanol/toluene under hydrogen (section 4.2) and from 2a with Proton Sponge® in toluene-dg under hydrogen (section 2.4). Complex lb: The complex, lb, was prepared as described in section 2.4, and isolated prior to recording the NMR spectra. Reactions of la and lb with 1-hexene The solvate-free complex, la, (~lf>3 M) was mixed with a ten to twenty-fold excess of hexene in toluene-dg in a NMR tube under vacuum at ambient temperature. The iH NMR spectrum was recorded both after 3 h from time of mixing and after 9 & The vacuum distillate of the solution was then analyzed by gas chromatography. For the analogous reaction of 1-hexene with lb, the complex (~10"2 M) was mixed with an approximately ten-fold excess of 1-hexene in CD2CI2 in a NMR tube under vacuum at ambient temperature. The *H NMR spectrum was recorded both after 7 h from time of mixing and after 9 d. The vacuum distillate of the solution was then analyzed by gas chromatography. Reactions of (rf-H2XisoPFA)Ru(^-Cl)2(ii-H)RuH(PPh3)2,4 To test the stability of 4 in CD2CI2 solution under vacuum, the complex, (~10 mg, 0.01 mmol) was dissolved in CD2CI2 in a 5 mm NMR tube, and the tube was sealed under vacuum. The !H NMR spectrum was recorded within six hours and after three weeks in the NMR tube, kept mostly at -30 °C but also for several days at ambient temperature. Reaction with 1-hexene was studied by mixing 4 (5 mg, 0.0043 mmol) with 1-hexene (0.040 mmol) under vacuum in CD2CI2 (1 g) at ambient temperature. The lH NMR spectrum was recorded after 8 h and both *H NMR and 31P{1H} NMR spectra recorded after two weeks. The vacuum distillate of this solution was then analyzed by gas chromatography. Preparation and observation of dihydrogen adducts of 2a and 2b in situ The complexes, 2a, prepared by stirring RuCl3(PPh3)2(DMA)-DMA under hydrogen (several atm) in n-butanol (sect 2.4), and 2b, prepared as described in section 2.4, were each dissolved in toluene-ds under 2-3 atm H2 and the NMR spectra recorded of the respective solutions containing the resulting H2 adducts. 3.3 Results Complexes la, lb and 4 An elemental analysis of a sample of the crystals of la, (T|2-H2)(PPh3)2Ru(u.-Cl)2(|i-H)RuH(PPh3)2toluene-d8-DMA, submitted for X-ray analysis, shows a C and H content consistent with a DMAtoluene-dg solvate; the nitrogen result of 0.0 % is perhaps in error as only a small sample was available (0.29 mg). Although toluene was found in the crystal structure, there appeared to be other solvent molecules present, possibly disordered, which could not be refined. Consequently, the hydrogen ligands could not be located. The complex was nevertheless shown to be a chloro-bridged dimer, with an arrangement of Ru, Cl and P atoms almost identical to that of the analogous tri(p-tolyl)phosphine structure obtained earlier (see Figs. 3.2, 3.3; Table 3.7). For complex 4, the crystallographic data are of excellent quality and refinement proceeded well to reveal the hydrido and dichloro-bridged dinuclear complex pictured in Fig. 3.4. Structural parameters are summarized in Table 3.8 and may be compared with the data for the two other analogous structures of la and lb (Table 3.7). The skeletal atoms of all three structures are numbered in an analogous manner to highlight their similarities, except that the nitrogen atom of the isoPFA ligand in 4 replaces P(4) of la and lb. It should be noted that the skeleton structure of 4 is not superimposable on its mirror image, and thus has this element of chirality, in addition to the chirality of the isoPFA ligand, which may be R,S or SJi. The dihydrogen ligand bond length, H(l)-H(2), is 0.80(6) A, and this ligand is symmetrically coordinated, "side-on", to the Ru atom bearing the isoPFA ligand Table 3.7 Selected bond angles and distances for crystal structures of (Ti2-H2)(PR3)2Ru(u-Cl)2(u-H)RuH(PR3)2 (R = Ph, p-tol), la and lb laa lbb Distances (A) Ru(l)-Ru(2) 2.83 2.80(0) Rud)-P(l) 2.22 2.23(1) Ru(l)-P(3) 2.25 2.28(1) Ru(2)-P(2) 2.27 2.26(1) Ru(2)-P(4) 2.39 2.39(1) Ru(l)-Cl(l) 2.48 2.46(1) Ru(l)-Cl(2) 2.56 2.57(1) Ru(2)-Cl(l) 2.44 2.47(1) Ru(2)-Cl(2) 2.47 2.48(1) Angles (°) P(l)-Ru(l)-P(3) 97.6(5) P(l)-Ru(l)-Cl(l) 166.8(5) P(l)-Ru(l)-Cl(2) 94.4(5) P(l)-Ru(l)-Ru(2) 111.9(4) P(3)-Ru(l)-Cl(l) 95.6(5) P(3)-Ru(l)-Cl(2) 113.8(4) P(3)-Ru(l)-Ru(2) 148.4(4) P(2)-Ru(2)-P(4) 104.1(5) P(2)-Ru(2)-Cl(l) 92.1(5) P(2)-Ru(2)-Cl(2) 167.6(5) P(2)-Ru(2)-Ru(l) 110.0(4) P(4)-Ru(2)-Cl(l) 102.9(4) P(4)-Ru(2)-Cl(2) 88.0(4) P(4)-Ru(2)-Ru(l) 139.3(3) Ru(l)-Cl(l)-Ru(2) 70.2 69.3(4) Ru(l)-Cl(2)-Ru(2) 68.4 67.3(3) aFrom the present work.55 bFrom reference 23. Figure 3.2 The skeletal structure of la, determined by X-ray diffraction. c P(30 Figure 3.3 The skeletal structure of lb, determined by X-ray diffraction. From reference 23. 41 Table 3.8 Selected bond angles and distances for the crystal structure of (Ti2-H2)(isoPFA)2Ru(u-Cl)2(p-H)RuH(PPh3)2,4 Distances (A) Ru(l)-Ru(2) 2.811(2) Ru(l)-P(l) 2.229(2) Ru(l)-P(3) 2.249(2) Ru(2)-P(2) 2.240(2) Ru(2)-N 2.300(4) Ru(l)-Cl(l) 2.453(2) Ru(l)-Cl(2) 2.620(2) Ru(2)-Cl(l) 2.435(2) Ru(2)-Cl(2) 2.528(2) Angles (°) P(l)-Ru(l)-P(3) 96.35(7) P(l)-Ru(l)-Cl(l) 168.32(4) P(l)-Ru(l)-Cl(2) 98.02(8) P(l)-Ru(l)-Ru(2) 114.33(6) P(3)-Ru(l)-Cl(l) 95.19(7) P(3)-Ru(l)-Cl(2) 114.36(7) P(3)-Ru(l)-Ru(2) 148.16(4) P(2)-Ru(2)-P(4) 96.5(1) P(2)-Ru(2)-Cl(l) 101.22(8) P(2)-Ru(2)-Cl(2) 168.54(4) P(2)-Ru(2)-Ru(l) 113.40(6) P(4)-Ru(2)-Cl(l) 92.4(1) P(4)-Ru(2)-Cl(2) 94.7(1) P(4)-Ru(2)-Ru(l) 138.46(9) Ru(l)-Cl(l)-Ru(2) 70.20(6) Ru(l)-Cl(2)-Ru(2) 66.16(6) Parameters for hydrogen ligands: Distances (A) H(l)-H(2) 0.80(6) Ru(2)-H(l) 1.50(4) Ru(2)-H(2) 1.47(4) Ru(2)-H(12) 1.71(4) Ru(l)-H(12) 1.49(4) Ru(l)-H(13) 1.50(4) Angles (°) Ru(l)-H(12)-Ru(2) 122(3) Ru(2)-H(l)-H(2) 72(4) Ru(2)-H(2)-H(l) 76(4) H(12)-Ru(l)-H(13) 89(2) H(12)-Ru(2)-H(2) 108(2H(12)-Ru(2)-H(l) 78(2) Ru(2)-Ru(l)-H(13) 108(1Ru(l)-Ru(2)-H(l) 96(2) Ru(l)-Ru(2)-H(2) 127(2Figure 3.4 The molecular structure of 4, from X-ray diffraction data, showing the positions of hydrogen ligands. The bridging hydride is unsymmetrically cc<>rdinated: Ru(l)-H(12) 1.49(4) A, Ru(2)-H(12) 1.71(4) A, and the Ru-(H2) distances, Ru(2)-H(l) 1.50(4) A, Ru(2)-H(2) 1.47(4) A, are not significantly different from the terminal, Ru(l)-H(13) distance of 1.50(4) A. The H-H bond is parallel to the Ru(l)-Ru(2) direction, and the plane defined by Ru(2), H(l) andH(2) roughly bisects the P(2)-Ru(l)-N and H(12)-Ru(l)-Cl(l) angles. Visible spectra of DMA solutions of la (e.g., Fig. 3.5) show a maximum at 495 nm (e = 1580 ± 10 M^cnr1) over a range of concentrations (1.01 x 10-3 M < [RU]T < 6.2 x 10~3 M) that roughly correspond to those used in the hexene hydrogenation study by Thorbum.71 Thorburn's catalysis data will be considered in sect. 4.4. 1 1 1 1 1 ' 300 400 500 600 700 800 900 Wavelength (nm) Figure 3.S The visible spectrum of la in DMA under hydrogen. The 31P{ JH} NMR spectra of the bis-DMA solvate of la exhibited the same features as the solvate-free complex generated in situ under H2 from 2a and Proton Sponge® (cf. sect 2.4). Peaks at 8 70.5 and 46.0 ppm at ambient temperature broaden on cooling to give four resolved resonances, at -58 °C, whose chemical shifts (79.5 (s), 63.5 (s), 62 (s), 29.5 (s)) remained temperature invariant down to -88 °C; there is thus rapid exchange of phosphorus environments at ambient temperature. The calculated (sect 2.3) and observed 31P{!H} NMR spectra are shown in Fig. 3.6. No couplings are resolved at 121 MHz, but the variation in broadness of these peaks (at -68 °C, W1/2 = 78,47,94,100 Hz, respectively) is consistent with the values of coupling constants reported by Dekleva et al."23 for the 32.4 MHz spectrum (see discussion, sect 3.4). The model used for the simulation was an ABCD - BADC exchange system, defined by a first-order rate constant, k. The "T2" values were calculated from the above halfwidths (wi/2), and T2 values and chemical shifts were kept constant over the temperature range of the simulation. It is noted that these "T2" values were merely input to a computer program for simulation purposes and probably reflect values of coupling constants more that actual T2 values. The obs'd c&lc'd 20 °C 2*C -18<C -38°C -58 «C -68 °C k=8x lOV k=8x 104!"1 k=l x 10*s-l k=l x 103 t-l JUL_ _J_JL k=40 S'1 1 s-90 ' K r 70 60 " 30 40 30 ppm 90 W ' 7cT 60 SO 40 30 ppm Figure 3.6 Observed and calculated, variable temperature 31P{'H} NMR data (121 MHz, toluene-dg) for la measured under H2. exchange constants, also shown in Fig. 3.6, indicate rapid exchange at 20 °C (k = 8 x 105 s-1). An Eyring™ plot (Fig. 3.7) yields values for AH* (70±12 kJ/mol), AS* (110±50 J/mol/K) and AG* (38±3 kJ/mol at 20 °C) for the phosphine exchange. Errors in k values from these line-shape analyses are greater for very fast and very slow exchange, since line shapes change less with temperature in these regimes than they do in intermediate exchange-rate regimes. These errors are estimated to vary from -10% for intermediate values of k, up to -50% for highest and lowest values. These errors were propagated by estimating their effect on the slope and intercept of the Eyring plot to give the errors in the activation parameters. The !H NMR spectrum of la at 20 °C (Fig. 3.8) shows a signal at -12.9 ppm (wi/2 = 75 Hz) which broadened on cooling to -18 °C, while further cooling yielded the slow exchange spectra: eg., at -69 °C: 8 -8.6 (d, 1 u-H, 2JPH = 66 Hz), -12.6 (br s, w^ ~ 500 Hz, 2 T|2-H2 protons), -17.2 (t, 1 terminal H, 2JPH = 30 Hz). Exchange spectra simulations employed an A2 - ^ BX « *"• CXY exchange model (cf. Fig. 3.9) where one proton, B, with a long T2 exchanges with both two protons, A2, with a short T2 and one other proton, C, with a long T2. The observed and calculated upfield !H NMR spectra -3 -j 1 i i . 0.003 0.004 0.005 l/roc1) Figure 3.7 Eyring plot for rale constants for phosphine exchange in la. of la are shown in Fig. 3.8. The exchange constant for the A2 •« BX exchange is ki, and for the BX •« CXY process is k2. In order to include the observed temperature dependence of the chemical shifts in the calculation, values for these shifts for the slow exchange spectra were plotted (Fig. 3.10) and extrapolation made to higher temperatures so as to be consistent with fast-exchange-averaged peaks. The chemical shift of the T|2-H2 resonance was held invariant with temperature; this was done rather arbitrarily, as broadening above and below -53 °C made observation of any change of chemical shift with temperature difficult. The T2 values used for the bridging and terminal hydrides were 8 and 12 ms, respectively, for all simulations except for the one for the spectrum at -78 °C, where slightly lower values (6 and 8 ms, respectively) were used in order to mimic broadening possibly caused by limitations in solubility at this temperature. Figure 3.11 shows how "T2" for the ri2-H2 resonance was allowed to decrease with decreasing ot»'d calc'd 20 °C 8°C -63 °C -78 °C mu-H| Ru-H: kl =^2=2 * 105 s-1 k]=k2=8 x 104 s-1 ki=k2=7.25 x 103 t k2=7x 102 s-1 kj=80 s-] ki=70 s"1 kj=k2=5 s") =k,= l s:l 8 -10 -12 -14 -16 -18ppm -8 -10 -12 -14 -16 -I8ppm Figure 3.8 Observed and calculated, variable temperature 'H NMR (300 MHz, toluene-dg) spectra of la measured under H2. temperature; values of "T2" were obtained from the half-widths for this resonance in the slow exchange regime and extrapolated to higher temperatures by using a "log T2 versus T*1" plot (Such a plot normally gives a straight line with slope of -1, since T2"1 a T1 in liquids and gases.7915 However, the broadness of dihydrogen resonances in general has not been fully explained, and the extrapolation used in Figure 3.11, and Figures 3.16 and 323 which are discussed later, is really empirically based, using the variation of the T2" of the dihydrogen resonance with temperature in the slow exchange regime.) The resonance for T|2-H2 at -78 °C was so broad as to almost disappear, but narrowed on return to higher temperatures; thus, it was difficult to estimate a "T2" from the half-width at this A2XYZ — BXYZ CXYZ H2 — H — H \/\/ Ru Ru Figure 3.9 General hydrogen ligand exchange model used to simulate hydrogen ligand exchange within la, lb and 4. bridging -25201 -2540+3 2 -2*0-tn "o -25OT S -1600-U -2620-• a • a -SO -60 -40 -20 0 20 40 temperature (°C) terminal -31801 N -S200-1 -S220-§ -S240--526C • a • a -80 -60 -40 -20 0 20 40 temperature (°C) Figure 3.10 Plot of 'H NMR chemical shifts versus temperature for terminal and bridging hydride ligands in la, with extrapolation to 20 °C; data measured under H2. Points marked 'a' are estimated as described in the text. 0.0030 • measured • extrapolated 0.0050 Figure 3.11 Plot of JH NMR (300 MHz, toluene-dg) T2 values versus temperature for the T|2-H2 ligand of la, measured under H2, including extrapolation to 20 °C. temperature. The exchange constants are shown in Fig. 3.8, and quantify the rate of exchange at various temperatures. This rate is rapid at ambient temperature (at 20 °C, k values = 2 x 105 S'1) and slows as the temperature is lowered (at -78 °C, k values = 1 s_1). The constants, ki and k2, were the same within experimental error. The fit of the simulation to the observed spectra is good;80 however, careful comparison of observed and calculated spectra recorded near the coalescence temperature reveals a difficulty experienced in matching the line shapes of the broad peaks in these spectra by simple adjustment of ki and k2. Figure 3.12 shows the Eyring plot for this system which yields 0.003 0.004 0.005 0.006 i/r (Kl) Figure 3.12 Eyring plot for rate constants (k s (kt + k^P) for hydrogen ligand exchange in la. the values for AH*, AS* and AG*(20 °C) of 60±10 kJ/mol, 60±40 J/mol/K and 41±2 kJ/moL respectively. The quality of the spectra used for low temperature Ti measurements was poor because of the low concentrations used. Errors in Ti values may be as high as 20%. Figure 3.13 shows me variation of Ti values with temperature. Because of the exchange processes, minima could not be located. The Ti values completely average at temperatures 49 1200-T -40 -20 temperature (' C) 40 • U-H • terminal-H • exchange-averaged Figure 3.13 Plot of Ti (300 MHz, toluene-dj) values versus temperature for hydrogen ligands of la. Values above 0 °C were obtained using a sample generated in situ in a NMR tube containing 2a and Proton Sponge® under 2 - 3 atm H2. above —40 °C. However, the values at -73 °C (Ti in ms, 5: 300, -8.6; 77, -12.6; 900, -17.2) indicate the presence of a T|2-H2 resonance at -12.6 ppm From -40 °C up to room temperature, the averaged Ti shows little change, with values of 40 ms at -40 °C, and 20 ms at both 6 °C and 20 °C Complex lb: Corresponding *H NMR measurements were made on the tri(p-tolyl)phosphine analogue of la and the high-field spectra of these two complexes found to be very similar at both high and low temperature. The upfield *H NMR spectra of lb were recorded from 20 0 down to-94 °C. The 20 ° and -94 °C spectra are virtually the same as 400 MHz spectra reported earlier by Dekleva et al.21^3 Figure 3.14 shows the observed and calculated spectra over the range, 20 0 to -83 °C The exchange model used for the simulation is similar to that employed in the simulation for hydrogen ligand exchange in the PPI13 obi'd calc'd 20 °C 0°C -22 °C -44 °C -64 "C -84 °C k=k]=k2 k=3 x I05 s"1 4 -1 k=9 x 10 J k=7 x 103 t 1 2 -1 k=8 x 10 s k=55 s k=5 s 8 -10 -12 -14 -16 -18 -20ppm-8 -10 -12 -14 -16 -18 -20ppiT Figure 3.14 Observed and calculated variable temperature 'H NMR spectra (300 MHz, CD2Q2. under hydrogen) for hydrogen ligands of lb. analogue, except for incorporation of additional coupling information available from the spectra of the p-tol analogue, resolution being better due to higher solubility. The model used, then, is A2 - *"•• BXYZ - ""• CXY, with k = ki ~ k2. In the model, a special case of that shown in Fig. 3.9, with indicated coupling constants set to zero, the exchange constant, ki, applies to the A2 - ^ BXYZ exchange, while the constant k2 applies to the BXYZ x ^ CXY exchange. The proton B, with a long T2, exchanges with both the A protons, with short T2 values, and also with the proton C, with a long T2. The lowfield resonance is due to B, the high-field resonance due to C and the intermediate resonance due to A2. The coupling constants used in the simulation are: JBX. JBY and JBZ (78, 8,2 Hz, respectively) and Jcx. JCY (both 27 Hz). Figure 3.15 shows how the chemical shifts of the T|2-H2 and terminal hydride resonances vary with temperature. Values of shifts obtained from the slow exchange spectra were plotted against temperature; the extrapolation to higher temperatures is rather arbitrary but consistent with the fast-exchange-averaged peak values, as in the case of the PPI13 analogue. However, this procedure leads to exchange -2000 -3000 --4000-13 13 13 B rf-Hj • terminal H; * estimated as described in the text. -5000-• • -6000 -1 00 -50 0 SO temperature ("C) Figure 3.1S Plot of lH NMR chemical shifts versus temperature for the n2-H2 and terminal hydride ligands of lb, including extrapolation to 20 °C. constants that arc in line with the low temperature exchange rates (see the Eyring plot later, Fig. 3.17). The T2 values used for the terminal and bridging hydride, "C" and "B", respectively, in the exchange model above, were obtained from the half-widths of resonances in the low-temperature spectra. These T2 values were held constant, at 20 ms, over the temperature range of the simulation; but the T2 of the dihydrogen ligand was varied as shown in Fig. 3.16, with the T2 values from the slow-motion-regime spectra extrapolated to higher temperatures in a manner similar to that used above for the la !H NMR spectra. Activation parameters obtained from an Eyring plot (Fig. 3.17) of these exchange constants were: AH* = 49±7 kJ/mol AS* = 3Q±30 J/mol/K AG*(20 °C) = 41±2 kJ/mol 0.003 0.004 • measured -- extrapolated 0.005 0.006 1/T (K1) Figure 3.16 Plot of !H NMR T2 values versus temperature for the TI2-H2 ligand of lb, including extrapolation to 20 °C. -2-J • —i • 1 0.003 0.004 0.005 0.006 l/roc1) Figure 3.17 Eyring plot for rate constants for hydrogen ligand exchange in lb. A typical inversion-recovery plot for the !H NMR Ti measurements on lb is shown in Fig. 3.18. The broad T|2-H2 resonance can be seen to recover more rapidly than the hydride resonances. Values of Ti measured over the temperature range -94 ° to -64 °C are displayed in Fig. 3.19. At -94 °C, Ti(ms) values are 44 (T)2-H2), 120 (u-H) and 350 (terminal H); the value for the vp-Hi resonance is within the range36 (4- 100 ms) expected for an T|2-H2 ligand. 54 Figure 3.18 Inversion-recovery method plot for 'H NMR Tt measurements (300 MHz, CD2C12, -83 °C, under hydrogen) for the hydrogen ligands of lb. Delay times (ms) are given at the left. The T|2-H2 ligand resonance recovers faster than that of the hydrides. 400 300 Vi V- 200 -100 -a a O O D • • 8 -100 -90 -80 -70 temperature (°C) O bridging H B terminal H -60 Figure 3.19 Plot of *H NMR Ti values (300 MHz,CD2Cl2, under hydrogen) versus temperature for the hydrogen ligands of lb. The 31P{ iH} NMR spectrum of 4 is shown is Fig. 3.20 and is invariant with temperature. A major set of resonances implies two major isomers or diastereomers: 8 78.5 (two overlapping doublets, P(l) trans to CI), 77.2 (br s, possibly two are overlapping, P(2) of isoPFA), 64.6 (two overlapping doublets, P(3) trans to p-H). The presence of a second set of 31P{1H} NMR resonances (Fig. 3.20), of lesser intensity, is probably due to two other rninor isomers, perhaps diastereomers. The region downfield of 8 0 in the 300 MHz iH NMR spectrum of 4 in CD2CI2 at 20 0 under hydrogen is shown in Fig. 3.21, and the detailed assignments are given in section 2.4. Signals from 8 6.9 to 7.55 are due to the PPh3 ligands, while those from 8 0.3 to 4.45 are due to the isoPFA ligand. Two distinct N(C/73)2 peaks at 8 2.35 and 3.1 show the coordination of the nitrogen atom in solution; the free ligand in solution exhibits only a single N(C//3)2 resonance at 8 2.1. Further indication of coordination of the N atom of isoPFA is the position of the FC-C//CH3 quartet at 8 4.45, which is shifted downfield from the free ligand value of 8 4.0.8 In the upfield region (Fig. 3.22), the main features, at 20 °C, are a broad resonance, and a doublet of doublets at 8-18.6 (see below) which is always accompanied by another minor doublet of doublets some 16 Hz farther downfield This rninor signal is probably results from another less abundant diastereomer, as noted in the 31P{ iH} NMR spectrum. An exchange process occurs mainly at the isoPFA end of the dinuclear molecule between the dihydrogen and bridging hydride ligands; the signals for the dihydrogen and bridging hydride ligands are resolved, though exchange-broadened, at -9 °C, and sharpen on cooling to -48 °C. The upfield region at -61 °C shows 8 -18.5 (t, IH, 2jPH . 30 Hz), -17.3 (br s, IH, p-H), -12.8 (br s, 2H, rj2-H2); see also sect 2.4. The observed and calculated (sect. 2.3) upfield iH NMR spectrum is shown in Fig. 3.22. The exchange simulation was carried out using basically the same model (Fig. 3.9) as for the !H NMR simulations described earlier, except that in this case, for 4, ki » k2; at 20 °C, ki = 4000 s"1, k2 = 40 s"1. The ki constant applies to the exchange process involving the T|2-H2 and p-H ligands, while the k2 constant applies to i | i i II | i i i i | i IT i | i i i i p TTT[ rrvr| i rrr] r i i i \i i i i | 82 80 78 76 74 • ITTTI r ii i |i i i rri i i i i i i i i I i r i 11 i nrnniTmnrnTi m i | i ( i v| 72 70 . | , , ,-| i i i i I 11 11 | i n i-i i i ri I i m-|-n i i | i i i i | i i i n 68 66 64 62 60 58 PPM Figure 3.20 31P{1H} NMR spectrum (121 MHz, CD2C12, ambient temperature, under hydrogen) of 4. Resonances marked with an 'a' are possibly due to two minor isomers of 4. Ut ON 1 I 11 11 I ' I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I ' ' I I II I I I I I I I (I I I I I I I I I I I I I I I I I I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 9 8 7 6 5 f> 3 2 1 0 PPM Figure 3.21 The JH NMR spectrum (300 MHz, CT^CLj, ambient temperature, under hydrogen) of 4. The resonance marked 'a' is due to CD2CI2 ; that marked 'b' is due to an unidentified impurity; the resonance marked 'c' is due to water, that marked'd' is due to grease 58 obs'd calc'd 20°C -9°C -48°C PPM Figure 3.22 Observed and calculated variable temperature *H NMR data (300 MHz, CD2C12, under hydrogen) for 4. the exchange process involving the bridging and tenxdnal hydride ligands. The model consists of a BXYZ spin system exchanging rapidly with an A2XYZ spin system and slowly with a CXYZ system, but only Jcx and JCY were allowed to be non-zero. The 'T2" values for the simulation of the Tp-Hi hgand resonance were permitted tp increase with increasing temperature by means of extrapolation from the low temperature values (see Fig. 3.23). The input T2" values for the -9 0 and 20 °C simulations of the other hydride resonances were given the value of 90 ms, obtained from the halfwidth of the terminal hydride resonance, at 5-183, from the -9 °C spectrum; for the -48 °C simulation, values of 59 • measured — extrapolated 0.002 0.003 0.004 0.005 0.006 1/T (K-1) Figure 3.23 'H NMR T2 values versus temperature for the TI2-H2 ligand of 4, with extrapolation to higher temperatures. 25 ms were used, to simulate hydride peak broadening in the rather poor quality spectrum recorded at -48 °C. The chemical shift of the p-H resonance was permitted to shift slighdy upfield with decreasing temperature (Fig. 3.24). The activation parameters obtained from an Eyring plot (Fig. 3.25) of the ki values, for the ri2-H2/p-H exchange, are: AH* = 51 ±8 kJ/mol, AS* = -3±30 J/mol/K and AG* = 52±1 kJ/mol at 20°G The inversion-recovery method was used to obtain Ti measurements (sect 2.3) for the ^ NMR upfield resonances; these measurements were carried out at 300 MHz, under hydrogen and over a range of temperatures. The inversion recovery plot for the -70 °C measurement, shown in Fig. 3.26, and a plot of the measured Ti values, over the range, -5150 -5350 I • i i i i i i i i i i | i i -100 -80 -60 -40 -20 0 temperature (°C) Figure 3.24 Dependence of the 'H NMR chemical shift on temperature for the U.-H ligand of 4. The shifts for the T)2-H2 and terminal H ligands were taken as constant at -3898 and -5572 Hz, respectively. -88 to 20 °C, is shown in Fig. 3.27. The Ti values show the presence of one T|2-H2 ligand (e.g., at -60 °C, for the resonance at 8 -12.8, Ti = 13.8(3) ms) and two hydrides (at -60 °C, for the resonances at 8 -17.3 0i-H) and -18.5 (terminal), the Ti values are 182(6) ms and 251(5) ms, respectively). In general for exchanging systems, the Ti values begin to average when the exchange time becomes less than the Ti of a given state;62 in the present case the Ti's of the TJ2-H2 and u,-H ligands begin to average at about -50 °C, where the exchange time = l/k\ = 1/(5 s-1) = 200 ms, and the T\ of the ligand at this temperature is about 200 ms (Fig. 3.27). The terminal Ru-H Ti begins to average with the Ti's of the other hydrogen ligands at about -10 °C, and since the terminal Ru-H Ti at this temperature is 230 ms, an estimate of k2 from the Tj measurements would be about 4 s-1; this is consistent with the estimate, from the line shape simulation, of k2 = 5 s1 at -9 °C. The Ti values of the resonances for inc p.-H and T)2-^ ligands are completely averaged at -9 °C. 1 0 -60 O -1 -"i 1 1 1 • I i 1 r-0.0034 0.0036 0.0038 0.0040 0.0042 0.0044 0.0046 1/TCK-1) Figure 3.25 Eyring plot for the kj values (T^-H^-H exchange) for 4. ri2-H2 2 > Figure 326 Inversion-recovery plot for *H NMR T, measurements (300 MHz, CD2C12, -70 °C, under hydrogen) for the hydrogen ligands of 4. Delay times (ms) are shown at the left • Tf-H2 ° p-H • terminal H -100 -80 -60 -40 -20 0 20 40 temperature (°Q Figure 3.27 Plot of 'H NMR T,. values (300 MHz, CD2CI2, under hydrogen) versus temperature for the hydrogen ligands of 4. Spin-saturation transfer studies at 300 MHz, under hydrogen, revealed no spin transfer between the three upfield resonances at -48 °C; however, at -9 °C, irradiation of the r)2-H2 . resonance caused collapse of the |i-H resonance, although irradiation of the p-H resonance decreased the Tl2-H2 resonance by only 60%. When the p-H resonance was irradiated, there was also a slight reduction in the terminal Ru-H peak area by < 20%, and when the terminal Ru-H resonance was irradiated, there was approximately 50% reduction in the p-H resonance. Reactions of la and lb with 1-hexene: Both complexes reacted with 1-hexene under vacuum to give hexane, isomerized hexenes and the respective RuHClQ?R3)3 complexes. Complex la: Three hours after mixing, the only hydride signal observed in the *H NMR spectrum was that of RuHCl(PPh3>3 (5(ambient temperature) -17.4 (q, 2JpH = 26 Hz), in agreement with the hterature.81 After 9 d, the *H NMR spectrum still exhibited only the 3UU - 400 E 300 200 -100 -0 B O H B B O O • B B -r same hydride resonance. The gas chromatogram of the distillate of this solution showed the presence of hexane, 1-hexene and 2-hexene (cis and trans) in a 10:1:100 ratio. Thus, only 1% of the original hexene remained, most being isomerized. Complex lb: The *H NMR spectrum taken after 7 h shows unreacted lb (8 -13.28(br s), cf. sect. 2.4) and RuHCl(P(p-tol)3)3 (8 -18.0 (q, 2JPH = 26 Hz), cf. lit82) and a barely visible triplet at 8 -15.8. After 9 d, the only hydride signal observed in the lH NMR spectrum was the quartet due to RuHCltT^tol)^. The gas chromatogram of the distillate of this solution showed the presence of hexane and 2-hexene (cis and trans) in a 1:3 ratio; less than 1% of the original 1-hexene remained Reactivity of la and lb under vacuum and under nitrogen Complex la: The NMR spectra of the complex in toluene-ds under vacuum in an NMR tube showed partial decomposition had occurred after 9 d. The 31P{1H} NMR spectrum was of poor quality because of low concentration, and peaks at 8 70.5 and 46.0, due to the original complex, dominated the spectrum. In the *H NMR spectrum, two hydride signals at 8 -12.9(br s) and -17.4 (q, 2JPH = 26 Hz) in a ratio of 6:1 indicated a mole ratio of la : RuHCl(PPh3)3 of 1.5:1. The lH NMR spectrum of a similar solution, but left under nitrogen for 9 d shows the same two hydrides to be present, suggesting that the same decomposition process occurred. Complex lb: A solution of this complex in CD2CI2 under vacuum in an NMR tube was largely unchanged after 10 h, as indicated by the lH NMR spectrum. But after 2 weeks, the integration of the lH NMR upfield signals (8 -13.28 br s, -18.0 q, 2JPH = 26 Hz) indicated a lb : RuHCltPtp-tol)^ ratio of 1:1. No other major hydride signals were observed (Fig. 3.28). The 31P{1H} NMR spectrum (Fig. 3.29) showed three main species to be Figure 3.28 *H NMR spectrum (300 MHz, CD2C12, ambient temperature) of lb, after two weeks under vacuum at ambient temperature. The resonance marked 'a' is due to lb; resonance 'b' is due to RuHCl(P(p-tol)3)3. An expansion of the upfield region is shown in the inset I—i—i—i—i—i—i—i—il i i—i—i—i—i—i—i—i—i—i—r UN 1 i—i i I |—i—i i I |—i—i—i—I—|—i—i—i—i—|—r—i—i—i—|—i—i—I—i—|—i—i—i—i—|—i—i—i—i—i—i—i—i—i—j—i—i—i—i—I—i—i BD 60 40 20 0 -20 PPM Figure 3.29 The 31P{ *H} spectrum (300 MHz, CD2C12, ambient temperature) of lb, after 2 wk under vacuum at ambient temperature. Resonances are marked as follows: 'a', lb; 'b', RuHCl(P(p-tol)3)3; 'c', the A'B'CD' species described by Dekleva;21 'd\ Ph3PO; unmarked resonances are unidentified. present: lb (8(20 °C) 68.7 br s, 44.7 br s, cf. sect 2.4), RuHa(P(p-tol)3)3 (8 54.2 br s, cf. lit.81) and another complex with a four-spin ABCD system (8(20 °C) 59.8 (d, JPP = 23 Hz), 51.7 (d, JpP = 30 Hz), 40.8 (dd, JPP = 42, 30 Hz) and 20.2 (dd, JPP = 42, 23 Hz), hereinafter referred to as the A'B'CD'83 complex, to distinguish it from the four-spin, ABCD systems described above. The same products were formed under nitrogen, indicating little, if any, reaction of lb with nitrogen. Reactions of 4 Stability in CD2CI2 solution under vacuum: The solution !H NMR spectrum showed that 4 was stable as such for hours at -20 °C. However, after three weeks in the NMR tube, mostly at -30 °C, but also for several days at ambient temperature, the solution exhibited NMR spectra that indicated -30% rearrangement to la (^Pf^H} NMR 8 (ambient temperature) 47 (br s), 71 (br s) and 8 (-68 °C) 79.5,64.5, 61.5 and 35.5 (4 br s); *H NMR 8 (ambient temperature) -13.1 (br s), and 8 (-68 °C) -8.9 (d), -12.4 (br s), -17.2 (br s)), -30% residual 4 (3ip{iH) NMR 8 (ambient temperature) 79 (dd), 77.5 (s), 65 (dd); lU NMR 8 (ambient temperature) -14 (very broad resonance), -18.6 (dd), and 8 (-68 °C) -12.8 (br s), -18 (br s), -18.6 (br s)), and other phosphorus containing products. Reaction with 1-hexene The lU NMR spectrum of the reaction mixture, after 8 h from mixing, indicated partial reaction to give new hydride signals, 8 (ambient temperature) -9.1 (d), -9.8 (bs). After two weeks no high-field (hydride) signals were found and the 31P{1H} NMR spectrum confirmed that the dihydrogen complex had completely reacted. The NMR spectra indicated the presence of Rua2(PPh3)(isoPFA), 3b (^P^H) NMR: 8 (ambient temperature) 94 (d), 35.5 (d), 2JPP = 32 Hz; !H NMR 8 (ambient temperature) as listed in sect 2.4, within 0.1 ppm) and other unidentified species. The gas chromatogram of the vacuum distillate showed that 0.010 mmol hexane and 0.030 (cis and trans) 2-hexene were produced from 0.0043 mmol 4 and 0.040 mmol 1-hexene. A dihydrogen adduct of [RuCHPPhdrii, 2a The dimer 2a prepared by the method of Wang24-70 and in n-butanol as described earlier (p. 14), was in neither case obtained analytically pure, and the 31P{ !H} NMR spectrum of the product from the latter reaction indicated ~10% phosphoms-containing impurity. Also, the identity of the product from the butanol preparation is implied by its fairly clean in situ reaction with hydrogen in the presence of Proton Sponge® to give la. Figs. 3.6 and 3.8 show the NMR spectra obtained for la produced from the [RuCl2(PPh3)2k from a butanol preparation; the 31P{ lH) NMR spectrum of this la solution shows only trace signals at 43.5 and 24.0, the latter being OPPI13 21«26 The NMR data for 2a under vacuum are given in sect 2.4. Under hydrogen (2-3 atm), the NMR spectra indicate formation of an adduct with dihydrogen: lH NMR (300 MHz, toluene-dg, ambient temperature) 8 -10.3 (br s), Ti = 17 ms, integrated phenyl: T]2-^ ratio = 60:1; 8 (-58 °C) -10.1 (br s), and a weak signal(s) at 8 —9, with -20% of the area of the 8 -10.1 signal; 31P{1H} NMR (121 MHz, toluene-dg, ambient temperature) 8 60.0 (bs), 52.7 (bs); 8 (-58 °C) 60.9 (d), 53.3 (d), 2JPP = 42 Hz. Other weak signals in the 3iP{*H} NMR spectrum indicate minor amounts of phosphoms-containing impurities. A dihydrogen adduct of [RuChiPip-tol^^i, 2b Again, the NMR spectra of 2b, under hydrogen, indicate the presence of a dihydrogen adduct (!H NMR (toluene-dg, 300 MHz, ambient temperature) 8-10.1 (br s), Ti = 17 ms, phenyl: rj2-H2 ratio = 80:1; 31P{1H} NMR (toluene-dg, 121 MHz, ambient temperature) 8 59 (br s), 52 (br s). Weak 31P{1H} NMR signals in the 8 40 - 50 range again indicate minor amounts of phosphorus containing impurities. 3.4 Discussion Complexes 1 and 4 On the basis of the low temperature Ti measurements, the complexes la and lb should be reformulated as containing a dihydrogen moiety. These complexes were reported23 along with other similar ones ([RuHaXL^h*. X = CI, L = ASPI13; X = Br, L = PPh3, AsPh3); all the species showed similar NMR spectra at ambient temperature, and the latter group probably also contain the v^Hi ligand. The crystal structures of the two complexes la and lb are almost identical (see Table 3.7 and Figs. 3.2 and 3.3). Corresponding bond lengths in many cases are the same within experimental error. This similarity for skeletal bond lengths is not surprising in view of the sirnilarity of the phosphine ligands in the two complexes. The Ru-Ru distance of 2.83 A in la is almost the same as (but significantly different from) that in the structure of the P(p-tol)3 analogue (2.80 A), and is close to the middle of the range generally found for a Ru-Ru single bond, 2.632 to 3.034 A.84-85 (A range of 2.28 to 2.95 A quoted elsewhere21'23 is in error, the value, 2.28 A, being the length of a multiple bond.86'87) Acute bridging angles (M-B-M, M = metal, B = bridging atom) of about 70° are regarded as being caused by the presence of a metal-metal bond in M2B2 systems;88 thus, there is further support in this case for Ru-Ru bonding in la from the Ru-Cl-Ru angles: ZRu(l)-Cl(l)-Ru(2) = 70.2°, ZRu(l)-Cl(2)-Ru(2) = 68.4°, while corresponding values for the P(p-tol)3 analogue, lb, are 69.3 and 67.3°. These may be compared with the "remarkably acute" ZFe-S-Fe = 68° in a Fe2S2 system,88 presented as evidence for an Fe-Fe bond. Dekleva et al.73 attributed deviations from "idealized" (C2) symmetry in the structure of lb to the presence of hydride ligands, although, as in the present PPh3 structure hydride ligands were not located. The probable hydride positions in lb (numbered in order of increasing chemical shift) were assigned as shown in Fig. 3.30; the P atoms also were numbered in order of increasing chemical shift in the low temperature 31P{ iH} NMR spectrum. Assignment of the hydride ligand positions was based on the low temperature 400 MHz iH NMR spectrum (cf. the 300 MHz spectrum, Fig. 3.14), the low temperature 31P{iH} and 31P{Ph-iH} spectra (at 32.4 MHz), as well as vacant site positions in the solid state structure. Lengthening of bonds, Ru(l)-Cl(2) and Ru(2)-P(4), compared to other corresponding bond lengths in the lb molecule, seemed to confirm the positions assigned to H(3) and H(l);23 no such lengthening of Ru(2)-Cl(l) was observed "trans" to an "H(2)" atom. The two "H(2)" atoms are, however, now regarded as the atoms of a symmetrically coordinated TJ2-^ ligand. The Ti values at -94 °C: 120 ms (H(l)), 44 ms (H(2), or more correctly, H2(2)), 350 ms (H(3)) indicate a hydride, a dihydrogen and a hydride ligand, respectively (Figs. 3.18,3.19). The Ru(2)-Cl(l) bond in lb, presumably trans to the T|2-H2 ligand, is about the same length as the Ru-Cl bonds trans to a phosphine. This suggests, assurning the placing of the T|2-H2 ligand is correct, Figure 3.30 Slow exchange structure showing proposed hydrogen ligand positions for lb, reported as a classical tetrahydride.23 that the rra/tf-influence of an Tj2-H2 ligand is weak relative to that of a hydride. The low temperature !H NMR spectrum of lb has been published, but this is the first report of the low temperature lH NMR spectrum of the PPh3 analogue. Dekleva, using CD2CI2. was not able to observe the low temperature *H NMR spectrum because of limitations of solubility.21 However, the solubility in toluene-dg is sufficient for the purpose. The low temperature JH NMR spectrum contains three upfield resonances at -69 °C, Otoiuene-dg -8.6 (d), -12.6 (br s), -17.2 (t), which average at 20 °C to one broad singlet (8 -12.9). The simulation (Fig. 3.8) confirms that the ratio of areas under these three resonances is 1:2:1, respectively. The Ti measurements (sect 3.3) for these resonances indicate a hydride, a dihydrogen ligand and a hydride, respectively. The Ti values (Fig. 3.13) are almost completely averaged, on warming to -43 °C, because of hydride/dihydrogen exchange. The crystal structures and NMR spectra for the two complexes, la and lb, are very similar, and the positions of the hydrides of the Ph analogue can be assigned as in the p-tol complex (Fig. 3.30) by similar arguments.23 Thus, the H(l) resonance at 8 -8.6, due to a bridging hydride, is consistent with its position trans to two phosphine ligands, P(3) and P(4), as there is a tendency for resonances of hydrides trans to ligands of high rraw-influence to be generally at lower field than those of hydrides trans to ligands of low fra/ts-infiuence, other things being equalP The terminal hydride, H(3), which is trans to one ligand of low frans-influence, Cl(2), resonates at higher field (8 -17.2). The Ru(2) atom is a more likely site for the r|2-H2 ligand because of the close structural correspondences between la and 4 (discussed later), where the rj2-H2 ligand is crystallographically located on Ru(2); and the length of the Ru(l)-Cl(2) bond (2.56 A), longer than the other Ru-Cl distances by -0.1 A, is consistent with it being trans to the H(3) ligand of high rra/tf-influence.89 The 31P NMR resonances of phosphines trans to hydrides are typically at higher field90-91, and the assignment of the P(3) and P(4) resonances is consistent with their positions being trans to the hydride, H(l). The reason for the lengthening of the Ru(2)-P(4) bond relative to other Ru-P bonds in the molecule is not clear, a similar lengthening occurs in the p-tol analogue. Assignment of these positions in a solid state structure, using solution NMR data, seems valid because, as for lb, the low temperature 31P{ lH} NMR spectrum of the PPh3 complex is consistent with retention, in solution, of the solid state structure. The low temperature 32.4 MHz spectrum reported by Dekleva21 (32.4 MHz, CD2CI2): 73.9 (d, 2JPH = 40 Hz, P(l)), 59.4 (dd, 2JPH = 20, 5 Hz, P(2)), 55.4 (td, 2JPH = 40, 5 Hz, P(3)) and 28.0 (dd, 2JPH = 40,20 Hz, P(4)) exhibits the coupling pattern diagrammed in Fig. 3.31. These 31P{1H} NMR data are similar to those obtained for la in the present work: (121 MHz, toluene-dg, -58 °C down to -88 °C) 8 79.5 P(l), 63.5 P(2), 62.5 P(3), 29.5 P(4). The 31P{1H} NMR spectrum for the bis-DMA solvate and for the DMA-free complex are the same in toluene-dg; thus, the DMA is not coordinated in solution. Dekleva et al.21^3 also found the visible spectra in toluene and in DMA to be similar and that la obeyed Beer's Law in DMA, confirming the original work of Wang,70 and conformity to Beer's Law is confirmed in the present work. Thus Thorburn's observations (Table 3.9) Figure 3.31 31P{!H} NMR coupling pattern for complex la.21 indicating non-conformity to Beer's Law for la in DMA (perhaps because of impurities in the DMA) are surely in error and there is no evidence for the monomer-dimer71 equilibrium he suggested. Wang's molecular weight determination70 of the bis-DMA solvate in benzene, on a sample analyzing correctly for (RuHCl(PPh3)2DMA]2 (now, more correctly formulated as (Ti2-H2)(PPh3)2Ru0i-a)2(^-H)RuH(PPh3)2-2DMA) had given an effective value of 500, and was interpreted by invoking dissociation of solvated DMA molecules from the dimer in benzene solution. Thus, the actual value determined experimentally for (Tl2-H2)(PPh3)2Ru(u-a)20t-H)RuH(PPh3)2-2DMA) was 1500 (calcd,: 1499). Thus, this molecular weight determination is further confirmation of dinuclearity in benzene solution, and in the light of spectroscopic similarities in DMA and in toluene, implies Table 3,9 Spectroscopic data for la •31P{1H) NMR spectrum solvent, S values toluene, 69.5, 45.3 C6D6, 68.0, 43.1 CD2C12, 73.9 (d, 40 Hz), 59.4 (dd, 20, 5 Hz), 55.4 (td, 40, 5 Hz), 28.0 (dd, 40, 20 Hz) (see also section 3.3) !H NMR spectrum solvent, 8 values C6D6, 13.8 C6D6) 12.81 (see also section 3.4) temperature (°Q -60 ambient -80 atmosphere not specified temperature (°Q 30 ambient atmosphere not specified Visible spectrophotometric data (DMA solution) ATOM (nm) 500 500 495 495 e (M^cm-i), concentration (Mx 103) 1530b(±200) (Beer's Law obeyed) 700c, 0.4 750°, 0.9 90O, 1.3 1000°, 2.0 1100*=, 4.0 1200=, 7.5 1500b, 0.18 -12 1580, 6.20 1570, 3.12 1580, .1.01 atmosphere not specified argon not specified 25 °C reference 70a 21 21 reference 70 21 temp.(°C) reference 70 71 hydrogen (not appreciably different in the absence of hydrogen) hydrogen 25 °C not specified 21 present work aThe 31P{ lH} NMR data reported by Wang70 were in error, his "-60°C" spectrum actually looks very much like the room temperature spectrum. The identity of the complex Wang studied is, however, confirmed as la by its visible spectrum, as indicated in the table. bPresumably, e was calculated based on the concentration of ruthenium, rather than concentration of dimer, this was not made clear by the author. cCalculation based on concentration of ruthenium. dinuclearity in DMA as welL Thus, there is NMR, spectrophotometry and molecular weight evidence for retention of the solid state structure of complexes 1 in solution. The model (P(l) exchanging with P(2), P(3) exchanging with P(4)) that successfully simulates the exchange process observed in the 31P spectrum of the PPh3 complex, probably reflects the change in phosphorus environments caused by the hydrogen ligand exchange processes. Thus, for example, P(l) on Ru(l), and P(2) on Ru(2), exchange environments as the TJ2-H2 ligand resides on alternate ruthenium centres. This rationalizing of the variable temperature 31P behaviour requires the exchange rate constants from the 31P and lH NMR simulations to be equal within experimental error, and this is probably true: the rate constant (k = 8 x 105 s*1) for the phosphorus exchange at 20 °C and that Qc = 2 x 105 s'1) of the hydrogen exchange at 20 °C are similar, considering the error estimate of ±50% given above for k values in the fast exchange region. The !H NMR exchange rate constants for hydrogen ligand exchange for the two analogues, la and lb, are similar. However, comparison of the *H NMR spectra of the PPh3 complex (-63 °C, Fig. 3.8) and the P(p-tol)3 complex (-64 °C, Fig. 3.14) reveals slightly more broadening of signals for the p-tol complex, even though, for the latter complex, resolution is better because of higher solubility of this complex. The mechanism of exchange may involve coordinated r|3-H3 units, bridging dihydrogen, or both. Involvement of an T|3-H3 unit in hydride/dihydrogen exchange in [Ir(H)(rt2-H2)(bq)(L)2]+ (bq = 7,8-benzoquinoline; L = PPh3) has been suggested,62 and the trihydrogen ligand has been shown to exist in the complexes, [Ir(H3)(r|5-C5H5)(L)]+ (L = PMe3, PPh3, AsPh3)92 MfoCp? (where M = Nb, Ta; Cp_ = C^Me, C5H4(SiMe3) or C5H3(SiMe3)2) and RuH3(C5Me5)(PR3) 93»94 There is no documented precedent for bridging H2, although within the complex, CpRu(p-H)4RuCp,95 the (p-H)-(p-H) distances in the solid state (X-ray) perhaps indicate two bridging H2 molecules; the Ti measurements in solution indicated classical hydrides, but this, of course, does not preclude the existence of p-H2 groups in the solid state.95b Kubas et a/.51 have shown that hydrogen ligands can readily shift back and forth from classical hydrides to a nonclassical dihydrogen ligand in solution. The suggestion has been made61'96 that an N2 ligand should be replaceable by an ri2-H2 ligand when the dinitrogen stretching frequency falls in the range 2060 - 2150 cm1; examples of replacement of coordinated dihydrogen by dinitrogen have also been observed.36,61162-96 However, no evidence of reaction with nitrogen by the complexes, la and lb, was observed; the decomposition of the complexes, la and lb, in the absence of hydrogen, was not appreciably affected by the presence of dinitrogen. The unidentified A'B'CD' complex, referred to in section 3.3 as a decomposition product of lb has been discussed by Dekleva,21 who also examined the 31P{1H}, 31P{phenyl-1H) and JH NMR of the analogous PPh3, A'B'CD' complex. The similarity of the 31P NMR spectra of these two A'B'CD' complexes imply very similar structures. Also, he pointed out that the similarity of the 31P{ lH] chemical shifts and coupling patterns for the two complexes, lb and the p-tol A'B'CD' complex, imply a similar - dinuclear structure and phosphine-ligand disposition for these two molecules. The chemical shifts of the two upfield 31P NMR resonances, Pc and PD (in order of increasing chemical shift), would allow for two phosphine ligands trans to a bridging hydride. Also, a 31P{phenyl-1H} NMR spectrum of the PPh3 A'B'CD' complex showed that: the PD resonance was coupled to only a single hydride with 2JPH = 63 Hz, Pc appeared to be coupled to at least one hydride with 2Jpn » 30 Hz, PB was broadened only slightly, and PA was possibly coupled to more than one hydride, one of which had 2JPH « 20 Hz 21 Dekleva was unable to determine the number of hydrides by NMR; however, from the reaction of the PPh3 A'B'CD' complex with PPh3, he inferred the following reaction: A'B'CD' + 2PPh3 +• RuHCltPPh^ (3.9) and it was concluded that this A'B'CD' species had at least one hydride per ruthenium. The formulations Ru2H2Cl2(PPh3)4 and Ru2H3Cl3(PPh3)4 were considered, but no definite conclusions were drawn. The decomposition of lb, under nitrogen or vacuum, observed in this work, probably proceeds by loss of dihydrogen, followed by disproportionation of Ru2H2Cl2(P(p-tol)3)4 to give RuHCl(P(p-tol)3)3 and other unidentified products. Further, it appears from the reaction of lb with hexene, under vacuum, that either, or both of lb and Ru2H2Cl2(P(p-tol)3)4 react with hexene, because the only phosphme-containing product in significant concentration is RuHCl(P(p-tol)3)3. Whether or not the T|2-H2 ligand atoms are actually transferred to hexene in this latter reaction is not yet resolved. The reaction of RuH4(PPh3)3 with 1-pentene was reported by Cole-Harnilton and Wilkinson,91 prior to the complex being shown to be a dihydrogen dihydride species.38 The reaction gave an undetennined amount of dihydrogen, 2 moles of pentane and an apparent pentadiene complex, with no Ru-hydride products being detected. Although the mechanism proposed involved initial loss of hydrogen, followed by reaction of an Ru(H>2 moiety with pentene, the data do not preclude 1-pentene as the source of the observed dihydrogen, with the r|2-H2 and hydride ligand atoms involved in formation of the pentane. In the present work, however, since RuHCl(PR3)3 remains after reaction, with hexene, of both species 1 (R = Ph, p-tol), it seems more likely that the dihydrogen is transferred to form the alkane. The question of the fate of the hydrogen ligands in the reaction of la and lb with hexene could prove difficult to sort out even by deuteriation studies, because isomerization of alkene, and H ligand exchange are both occurring, as well as hydrogen-atom exchange with ortho-phenyl hydrogens.71 Other reactions of the complexes, [(TI2-H2)L2RU(U-X)2(U-H)RUHL2] (X = Cl, Br; L = PPh3, P(p-tol)3, AsPh3), with L, CO and bipy, giving RuHXL3, RuHCl(CO)2L2 and RuHCl(bipy)L2, respectively, proceed with evolution of 1 mole of dihydrogen per mole of starting complex.23 Similar evolution of hydrogen has also been observed in related reactions of RuH4(PPh3)3.30'38 Pi ^-dissociation of the molecular hydrogen ligand has been suggested for all these reactions The kinetics of the reaction of la and lb with the respective phosphine have been studied by Thorburn71 and Dekleva et al.21'23 The reaction rate showed a direct inverse first-order dependence on dihydrogen, consistent with a pre-dissociation equilibrium involving loss of dihydrogen. This dihydrogen dependence and first-order dependences on both phosphine and ruthenium were interpreted in terms of the following mechanism: K _ [RuH2Cl(PR3)2]2 [RuHClfPRshh + H2 (3.12) [RuHCl(PR3)2]2 + PR3 products (3.13) While a value for K2K could be obtained from the kinetic data at 30 °C, individual k2 and K values could only be estimated. The estimate for K (£ 2 x lfr5 M for both systems) is close to that obtained from a re-examination of the kinetic data obtained by Thorburn70 (sect 4.4) for the hydrogenation of 1-hexene catalyzed by la. With the new formulation of the tetrahydride dimers as 1, K becomes the dissociation equihbrium constant for the loss of the dihydrogen ligand from the dimer (ri2-H2)(PR3)2Ru(u-Cl)2(n-H)RuH(PR3)2 H2 + (PR3)2Ru(u-Cl)2(u-H)RuH(PR3)2 (3.14) which is followed by: (PR3)2Ru(n-Cl)2(u-H)RuH(PR3)2 + 2PR3 » 2RuHCl(PPh3)3 (3.15) The crystal structure of 4 (Rg. 3.4) is the third reported structure to reveal a Ti2-H2 ligand; no other crystal structures of a dinuclear dihydrogen complex have been reported, in which the dihydrogen ligand has been located. The H-H distance, 0.80(6) A, is similar to those found in the mononuclear complexes (sect 3.1). The presence of a T)2-^ ligand is supported by the *H NMR Ti measurements at low temperature, the value of 12 ms (5 -12.8) indicating a very short H-H distance. The exchange exhibited by the hydrogen ligands is unique in that it mostly involves the U-H ligand and r)2-H2 ligand, which is at the isoPFA end of the molecule. Exchange is slow between the p-H and the terminal hydride, at the other end of the molecule. The linear Eyring plot for ki (for the fast process) shows the self-consistency of the results. Except for the nitrogen atom, the mirror image of the skeleton of this isoPFA structure is approximately superimposable on those of the analogous structures of la and lb, even to the point that the bond distances and angles are similar (cf. Tables 3.7 and 3.8, and Figs. 3.2, 3.3 and 3.4). For example, distances for Ru(l)-Ru(2) are all between 2.80 and 2.83 A; distances for Ru(l)-P(l) are 2.22 to 2.23 A; and for Ru(2)-P(2), 2.24 to 2.27 A. The Ru(2)-P(4) distances are the longest Ru-P distances in Table 3.7, while the Ru(2)-N distance is longer than all the Ru-P distances in the isoPFA structure (Table 3.8). Such close correlations occur all down the list of bond distances and angles in the Tables 3.7 and 3.8. Furthermore, the effects of trans influences, cited in discussion of Table 3.7, are also reflected in the parameters in Table 3.8 for the isoPFA structure. The longest Ru-Cl distance in each of the three structures is the Ru(l)-Cl(2) bond. This is attributed to its position trans to a hydride. In the case of the isoPFA structure, the Ru(l)-Cl(2) bond (2.620(2) A), which is lengthened relative to other Ru-Cl distances in this molecule (2.453(2), 2.435(2) and 2.528(2) A) is confirmed by the crystal structure to be trans to the terrninal hydride (H(13)). The *H NMR chemical shift of this terminal hydride (8 -18.6) is consistent with its position trans to a ligand (Q) of weak trans influence. The magnitudes of the coupling constants for the 8-18.5 resonance are in the range typical of a hydride cis to a phosphine ligand,89-91 The rro/u-influence of the il2-^ ligand (trans to Ru(2)-Cl(l), 2.435(2) A) is weak, and this is the shortest Ru-Cl distance in this molecule. As implied earlier in this section, the Ru(2)-N distance, which corresponds to Ru(2)-P(4) in each of la and lb, is anomalously large. This could be attributed to the fra/is-influence of the bridging hydride; but, as in la and lb, the Ru(l)-P(3) distance (2.239(2) A) is comparable to the other Ru-P distances (2.229(2) and 2.240(2) A) in this molecule, and is evidently only slighdy affected by its trans relationship to the bridging hydride (H(12)). This is particularly surprising because in 4 the unsymmetrically bridging hydride, H(12), is bonded more closely to Ru(l) (Ru(l)-H(12): 1.49(4) A) than to Ru(2) (Ru(2)-H(12): 1.71(4) A). There is, however, evidence in the 3 ^{^H} NMR spectrum of the trans-influence of the bridging hydride on P(3). The two doublet-of-doublets must be due to P(l) and P(3), which are to each other, and have differing structural relationships to P(2) (cf. Fig. 3.4). The difference in chemical shifts is undoubtedly due to the trans-influence of bridging ligands. The upfield resonance at 8 64.6 is therefore assigned to the P(3) atom, which is trans to a bridging hydride, H(12), and the resonance at 78.5 is assigned to the P(l) atom, which is trans to a bridging chloride. The relations of the observed couplings to the structure are shown in Figure 3.32, and the 31P{1H} NMR data for 4 are thus consistent with retention of solid state structure in solution. Although the crystallographically determined H(12) position seems inconsistent with other skeletal bond lengths in the molecule, it is possible to rationalize this using electron counting. The Ru(2)-H(12) bond distance is long, while the Ru(l)-H(12) distance is about the same as other Ru-H distances in the molecule. Thus, if this bridging hydride atom is considered to Figure 3.32 The 31P{!H) NMR coupling pattern for complex 4. The p.-Cl atoms are the only bridging atoms shown in the structure diagram to facilitate the comparison with complexes 1 (See Fig. 3.31). The crystallographically located U.-H atom is approximately trans to P(3). be coordinated solely to Ru(l), and assuming the existence of a metal-metal bond, the electron count is 18 at each Ru atom. The reaction of 4 with 1-hexene to give 2 moles of hexane is again reminiscent of the work of Cole-Hamilton and Wilkinson91 (vide supra) on the similar reaction of RuH4(PPh3>3 with pentene to give 2 moles of pentane. As in the discussion above, it is not known whether the T(2-H2 ligand is the source of hydrogen for one of the moles of hexane, or whether, in a mechanism similar to the proposal of those workers, the hexene is the source of hydrogen for one mole of hexane, with hexadiene, perhaps coordinated, also produced. The partial disproportionation of 4 to la under vacuum, in CD2CI2, occurred slowly, over a period of weeks. Under hydrogen, in CD2CI2, however, there was no change in the :H NMR spectrum far several weeks after mixing the components in a sealed NMR tube. It is not clear whether this disproportionation is the pathway for the formation of la in the reaction of RuCl2(PPh3)(isoPFA), 3b, with hydrogen in DMA (see sect 4.3). The three dinuclear dihydrogen complexes la, lb and 4 described above are analogous to the complex, (T|2-H2)(PCy3)2Ru(p-Cl)2(p-H)RuH(PCy3)2, reported in the hterature.96 The only other reported example of a dinuclear dihydrogen complex is the poorly characterized, "Cu2H2(H2)x", which was reported at a conference 97 and is referred to by Kubas29 in his review. Dihydrogen adducts of the complexes, Ru2CU(PR3U (R = Ph, p-tol), 2a and 2b Dekleva's attempts21 to prepare the complexes 2a and 2b in DMA by the method of Wang70, produced samples with with unsatisfactory elemental analyses, attributed to a "DMA impurity" (for 2a, Calc'd for (C36H30a2P2Ru)2: C, 62.07; H, 4.34. Found: C, 62.81; H, 5.42; N, 0.87). The carbon analysis for the product obtained from butanol solution in the present work show an improvement in elemental analysis, but is still not entirely satisfactory (sect 2.4). Both complexes 2a and 2b largely retain their structure under hydrogen in toluene-dg; however, the broad !H NMR resonances at about 8 -10, exhibited by both complexes under hydrogen, show Ti values that indicate the binding of an rj2-H2 ligand. The integrations of these resonances indicate only one r|2-H2 for every two Ru2Cl4(PR3)4 moieties. The suggestion by other workers76 of the possibility of polymeric species in concentrated solution was mentioned in section 3.1; oligomeric species, such as Ru(rj2-H2)a(PR3)2{Rua2(PR3)2}6(Tl2-H2)Cl would explain the JH NMR integrations in the present case under H2. However, the solution 31P{1H} NMR resonances for 2a under hydrogen were almost identical to those for 2a under vacuum (see sects. 3.3 and 2.2), which might perhaps be due to very loose binding of the T|2-H2 ligand. Joshi in this laboratory has recendy discovered that [RuCl2(dppb)]2 also binds an T|2-H2;98 the adduct of this latter complex exhibits a *H NMR upfield resonance similar to those described above (8 —10, Ti - 13 ms). As discussed earlier, solutions of RuCl2(PPh3)3 contain, among other species, 2a, and the discovery of the binding of dihydrogen to this dimer under slightly elevated pressures thus has important implications for the mechanism of the reaction of RuCl2(PPh3)3 with hydrogen, and the widespread use of RuCl2(PPh3)3 for catalytic hydrogenation of olefins. Chapter 4 Mechanistic studies on hydrogenation catalyzed by tertiary phosphine complexes of ruthenium(II) 4.1 Introduction Interest in homogeneous hydrogenation grew slowly after early reports by Calvin in 193899 on the hydrogenation of benzoquinone catalyzed by cupric acetate, and by Roelen,100 in 1944, on hydrogenations catalyzed by cobalt carbonyl complexes. Then after foundational work101 by Halpern and coworkers in 1952 - 62, rapid expansion of the field began to take place, mainly in the area of platinum metal catalysts. The history of ruthenium(II) catalysts stems from when Halpern, Harrod and James102 reported in 1961 that aqueous solutions of chlororuthenate (H) were active catalysts for the hydrogenation of several olefinic compounds, including maleic, fumaric and acrylic acids. In 1965, the complexes, RuCl2(PPh3)3> and RuHCl(PPh3)3> were reported, by Wilkinson's group,66 as extremely effective catalysts for the hydrogenation of terminal olefins. These Ru complexes, as well as the related RhCl(PPh3)3 complex, are widely used now as catalysts in organic syntheses.3 Because of experimental difficulties, Wilkinson's group had not obtained detailed kinetics for the hydrogenations catalyzed by RuHCl(PPh3)3. James' group found that kinetic studies were possible in DMA103, and a series of studies213,4'21'23,69"71'73'78 was initiated in this solvent on this and related complexes. Addition of PPh3 was found to inhibit hydrogenation, yet there was no observable dissociation of phosphine from RuHCl(PPh3)3 in solution.26'72-73 To explain this and other kinetic data, James2 proposed the following mechanism as one of the pathways for certain olefinic substrates: RuHCl(PPh3)3 + olefin ^== RuHCl(PPh3)2(olefin) + PPh3 (4.1) RuHa(PPh3)2(olefin) ^= PPha RuCl(PPh3)2(alkyl) ^= RuCl(PPh3)3(alkyl) Rua(PPh3)2(alkyl) (4.2) (4.3) Hi Rua(PPh3)3(alkyl) • ** RuHa(PPh3)3 + alkane (4.4) However, Eisenberg et a/.6,7 have recently demonstrated a "pair-wise" addition of hydrogen to olefins, catalyzed by RuHCl(PPh3)3, Ru(H2)H2(PPh3)3 and Ru(N2)H2(PPh3)3 (as well as some Rh and Pd catalysts); this discovery of pair-wise addition perhaps contradicts the mechanism suggested above, involving alkyl formation, followed by oxidative-addition of H2 and subsequent hydride migration to give the alkane with regeneration of the monohydride catalyst The findings of Eisenberg's data imply catalysis via a dihydride intermediate, with hydrogen transfers being fast compared to relaxation of the hydrogen nuclear spins in para hydrogen. Thus, a reinterpretation of some unpublished data obtained by Hui10^ and Markham72 on hydrogenation catalyzed by RuHCl(PPh3)3 should be carried out at some time. Because of its suggested involvement in reactions 4.1 - 4.3, a bis phosphine species was later isolated by James et al.69 The species is now known, from the present work, to be (Tj2-H2)(PPh3)2Ru(p-Cl)2(p-H)RuH(PPh3)2, la, (sect. 3.3) and not, as previously thought,69 (RuHCl(PPh3)2]2. Related to the RuHCl(PPh3)3 hydrogenation system, it is noteworthy that the "bis phosphine species", la, appears to be accessible in solutions of RuHCl(PPh3)3. Several workers56'104-105 have observed broadening of the *H NMR hydride signal of RuHCl(PPh3)3 under a hydrogen atmosphere, and have attributed the broadening to various exchange processes. Dekleva et al.21'56 also found that, when solutions of RuHCl(P(p-tol)3)3 were warmed under 1 atm hydrogen in toluene-dg, the high-field hydride quartet at 5 -16.8 broadened and a new weak broad signal at 5 -12.8 appeared in the *H NMR spectrum; this latter signal corresponds to the averaged, ambient temperature, high-field signal of the (Ti2-H2)(P(p-tol)3)2Ru(p-Cl)2(p-H)RuH(P(p-101)3)2 species, lb. As the "bis phosphine species" was a possible catalytic mtermediate,70'71'78 James and Wang carried out kinetic studies on the hydrogenation of acrylaxnide catalyzed by "[RuHa(PPh3)2]2" in DMA,78 and Thorburn71 followed this up with a similar study but using 1-hexene as substrate instead of acrylamide. The present work includes a reconsideration of Thorburn's data71 on the 1-hexene hydrogenation system using the catalyst la, as it is now reformulated Also in the present work, catalytic systems based on two analogues of RuCl2(PPh3)3 are considered: (a) the 1-hexene hydrogenation system employing RuCl2(PPh3)(PPFA), 3a, initially investigated by Rodgers et al.;25 (b) the 1-hexene hydrogenation system using RuCl2(PPh3)(isoPFA), 3b, as a catalyst Hydrogenation of 1-hexene catalyzed by la The re-interpretation of Thorburn's kinetic data71 is worthwhile because his analysis was based on a mistaken formulation of the ruthenium complex and on the assumption that the dimer catalyst dissociates in solution to monomers (see sect 3.4). The kinetic data presented below are Thorburn's,71 and the account given rests heavily on Thorburn's description of the work. Data on solution behaviour of la and its reactivity toward hexene come mainly from the present work. Representative uptake plots are given in Figure 4.1, and maximum rates were readily measured Complete reduction of 1-hexene occurred and the final solution contained the initial Ru complex, now known to be la; isomerization of 1- to 2-hexene was negligible. The kinetic dependences, together with some reciprocal analysis plots are shown in the appropriate Figures: Ru (Fig. 4.2), 1-hexene (Figs. 4.3,4.4), H2 (Fig. 4.5, 4.6), PPh3 (Figs. 4.7, 4.8) and LiQ (Fig. 4.9). 7 0 VO 20 time (s x l(r3) Figure 4.1 Uptake plots for the hydrogenation of 1-hexene in DMA at 30 °C catalyzed by la.71 16 0 0 40 . 60 12-0 [hexene] (Mx 10) Figure 4.3 Dependence of maximum rate of hydrogenation on 1-hexene concentration in DMA at 30 °C with PH2 = 1 atm, [Ruh-, as la, - 2 x 10"3 M.71 16-0 12 0 0 10 20 30 [Ru]T (M x 103) Figure 42 Dependence of maximum rate of hydrogenation on total Ru concentration (la) in DMA at 30 °C with PH = 1 atm, [hexene] = 0.2 M.71 3 0 [hexene]1 (M1) Figure 4.4 Dependence of maximum rate of hydrogenation on 1-hexene concentration as plotted according to equation 4.18 (sect 4.4).71 4 0 50 12-0 PHJ (mm Hg) 200 400 600 600 0 05 10 [HJOIxlO3) Figure 4.5 Dependence of maximum rate of hydrogenation on hydrogen concentration in DMA at 30 °C with [hexene] = 0.2 M, 0.7 M and [Ru]T, as la, = 2 x 10"3 M.71 8 0 4 0 8 0 [PPhsJCMxlO2) Figure 4.7 Dependence of maximum rate of hydrogenation on added trhphenylphosphine concentration in DMA at 30 °C with [hexene] = 0.2 M, 0.7 M and [Rub, as la,= 2 x 10"3 M.71 120 10 2-0 30 [HJ1 (M"» x lO"3) Figure 4.6 Dependence of maximum rate on hydrogen concentration as plotted according to equation 4.19 (sect 4.4).71 [PPhs]-1 (M1 x 10"1) Figure 4.8 Dependence of the maximum rate on [PPh3]-171 30 0 0 4 0 8 0 12 0 [LiCl] (M x 102) Figure 4.9 Dependence of maximum rate of hydrogenation on added lithium chloride concentration, in DMA at 30 °C with [hexene] = 02 M and [Ru]T, as la,= 2 x 10"3 M.71 Hydrogenation systems catalyzed by 3a and 3b A few comments on the choice of ligands in these studies are in order. As mentioned in Chapter 1, PPFA and isoPFA (Fig. 1.1) are ferrocene based ligands, possessing both phosphine and amine functionalities. The ligands are thus chelating, mixed soft/hard PN donor systems, with both planar and central elements of chirality. For practical reasons, catalysts which are selective for substrate type, e.g. olefin versus ketone, are more desirable in organic synthesis than ^discriminate (non-selective) catalysts. Thus, the original interest25 was to compare the known terminal olefin selectivity of, for example, RuCl2(PPh3)3 with the selectivity of complex 3 systems. It had been noted that rhodium complexes, such as Rh(bipy)2+ can selectively catalyze hydrogenation of a ketone in a ketone/olefin mixture,106 while similar systems using diphosphines selectively catalyzed the reduction of olefins.107 Another reason for the choice of a PN ligand was the possibility that the coordinated nitrogen donor might accept a proton and assist in the heterolytic activation of dihydrogen.25 There was also a possibility that mechanistic studies on the 3a and 3b systems might shed light on the analogous RuCl2(PPh3)3 systems, which are still not fully understood. Finally, because of the chirality of PPFA and isoPFA, the systems offered the prospect of catalytic asymmetric hydrogenation. Hydrogenation systems catalyzed by 3a: Previously, Rodgers et al.25 reported a study of the catalytic hydrogenation system: 1-hexene —RuCl2(PPh3)(PPFA) ^ hexane (4.5) n-butanol 40 °C In an attempt to identify possible intermediates involved, reactions of 3a with H2 were carried out in the absence of substrate. Rodgers et al. reported that n-butanol solutions of 3a absorbed slighdy more than 0.5 mole H2 per mole Ru at 40 °C at several pressures, the solution colour changing from green to red; these reactions at constant hydrogen pressure followed pseudo first-order behaviour, and visible spectral changes (loss of intensity of a band at 628 nm and increasing absorption at 350 - 450 nm, replacing an initial 457 band) were said to reveal "isosbestic points". Hydrides were not detected in the butanol solutions, or in the solution with benzene as solvent, where reaction with H2 was slow and incomplete. Attempts to isolate solids from the butanol solutions by solvent evaporation, even under H2, led to slow regeneration of the green colour of 3a. No 31P{ 1H} NMR signals could be detected from the hydrogenated butanol solutions. The system described by equation 4.5 effected efficient catalytic hydrogenation under mild conditions, with an activity comparable to that of RuHCl(PPh3)3, one of the most active catalysts known for the hydrogenation of terminal olefins. The hydrogenation rate was negligible for the internal alkene, 2-hexene, and for the substituted olefins, styrene and a-N-acetamidoacrylic acid. Rodgers et al. found no evidence for isomerization of 1-hexene during the hydrogenation. The kinetics of the hexene system described in equation 4.5 were followed using a gas uptake apparatus as described in reference 7. The S-shaped reuptake plots showed an initial autocatalytic type region before reaching a maximum rate at 200-800 s. The 'induction period' did not appear to be affected by 1-hexene concentration. The dependence on ruthenium in the absence of added PPI13 (Fig. 4.10), as measured by maximum uptake rates, was first order at low concentrations, but became less than one at higher [Ru]. The dependence on 1-hexene (Fig. 4.11) and on H2 pressure (Fig. 4.12) was first order, and the data also showed a direct inverse dependence on added PPI13 (Fig. 4.13). Because a hydride was not detected in the reaction of 3a with hydrogen in the absence of substrate, a mechanism for the hexene hydrogenation was proposed that involved initial quantitative formation of a Ru(I) species by hydrogen reduction: The lack of 31P{1H} NMR signals from hydrogenated ruthenium solutions was taken as support for proposing a paramagnetic Ru1 species. A pre-equilibrium loss of phosphine was invoked to account for the inverse dependence on phosphine and the apparent approach to half-order dependence on ruthenium, as well as a direct first-order dependence on Ru in the presence of added PPh3 (Fig. 4.10). The mechanism proposed to account for all these observations was: Run + V2H2 Ru1 + H+ (4.6) Ru(I)(PPh3) Ru(T) + PPh3 (4.7) Ru(I) + H2 Ru(ni)H2 hexene hexene Ru(I)Oiexene) Ru(I) + hexane (4.8) eo-i [Ru]T (MxlO4) Figure 4.10 A: Dependence of maximum rate on total Ru concentration (3a) in n-butanol at 40 °C; 0.04 M 1-hexene and 1 atm total pressure. B: As for A, but in the presence of 11.9 x 104 M added P^.25 PHJ (ami) Figure 4.12 Dependence of maximum hydrogenation rate on partial pres sure of H2 in n-butanol at 40 °C, 0.04 M 1-hexene and [Ruh-, as 3a, = 2.47 x If/4 M. Pressure reading corrected for vapour pressure of n-butanol.25 0 2.0 40 60 [1-hexene] (MxlO2) Figure 4.11 Dependence of maximum hydrogenation rate on 1-hexene con centration in n-butanol at 40 °C, 1 atm total pressure, and [Ru]T, as 3a, = 2.47 x 10-4 M.25 0 4.0 10 120 [PPhJ (MxlO4) Figure 4.13 Inverse dependence of maximum hydrogenation rate on added phosphine concentration in n- butanol at 40 °C, 0.04 M hexene, 1 atm total pressure and [Ru]T, as 3a, = 2.47 x ICT4 M. with a rate-law of the form: -d[H2] _ (k!Ki +k2K2)Kp[RuT][H23[hexene] dt ~ Kp + [PPh3] Thorbum65 followed up this work with gas uptake experiments in other solvents, and he found the stoichiometry in DMA at 10 °C to be 0.56 mole H2 per mole Ru, this decreasing with increasing temperature: 0.39 at 30°C, 0.19 at 50 °C. The uptake in DMA at 30 °C in the presence of Proton Sponge® was 0.57 mol H^/Ru. Attempted visible spectrophotometric experiments to monitor the kinetics of the reaction in 1:1 toluene/methanol gave no clear indication of reaction order, and no clean isosbestics, but did give spectral changes similar to those reported by Rodgers et al. Also, after storage of the hydrogenated solution for one day under Ar, the spectrum exhibited changes representing possible partial reversion to the starting spectrum Butler64 was later able to isolate a red solid from the H2 reaction with 3a in toluene/methanol, and found a "paramagnetic hydride" with a very broad NMR signal at 8 -17, that shifted "from sample to sample". Butler also found that there was no reaction between 3a and H2 in toluene, and Thorburn65 correspondingly reported zero uptake at 1 atm over 2 h in toluene, even in the presence of Proton Sponge®. Thus, at the time this present work was initiated, it appeared that a hydride was formed, at least in toluene/methanol; and mere were suggestions of reversibility in the reactions in DMA and in butanol. It was certainly puzzling that there should be no reaction in toluene, and yet a slow reaction in benzene. It was clear, however, that the presence of a polar solvent such as DMA or an alcohol had a pronounced effect on the reactivity toward H2. 90 (4.9) Experiments were therefore undertaken in attempts to isolate and/or identify hydride(s), formed by the reaction of 3a with dihydrogen, which might be active mtermediates in the catalytic hydrogenation of 1-hexene. Hydrogenation systems catalyzed by RuCl2(PPh3)(isoPFA), 3b: Ian Butler, from this department, first prepared 3b, and obtamed crystals for subsequent X-ray analysis.108 The solid state structure of this molecule is analogous to that of RuCl2(PPh3)3,26>109 and is best described as square pyramidal, with approximately trans chlorides, and the PPh3 ligand being approximately trans to the NMe2 moiety of the isoPFA ligand To complement work on the 3a-catalyzed 1-hexene hydrogenation system of Rodgers et al.,25 the study of 1-hexene hydrogenation catalyzed by 3b was considered worthwhile, and out of mechanistic studies the need arose to search for hydrides or other intermediates which might be involved in the catalytic cycle. Unlike for the 3a system, little work had been done at the outset of this present work on the reactions of 3b with H2. Butler had studied64 the reaction of a CH2CI2 solution of this complex with H2, but as no hydride was detected, the use of a more polar solvent was tried in the present work, because such solvents are known to promote monohydride formation via heterolytic cleavage of H2.2-4 4.2 Experimental Reactions oSRuCh(PPh)(PPFA), 3a, with H2 in toluene-dg, toluene-dglbutanol and butanol Some of complex 3a, typically 0.10 g (0.11 mmol), was placed in a 5 mm NMR tube and undried toluene-dg was freeze/thaw-degassed and vacuum-transferred into the NMR tube. Hydrogen was then admitted at below 1 arm into the NMR tube which was sealed while partially immersed in liquid nitrogen; the hydrogen pressure on wanning to ambient temperature was estimated to be 2 - 3 atm. The !H NMR and 31P{ NMR spectra were recorded after leaving the solution for 12 h at ambient temperature in the dark. In exactly corresponding experiments, some 2-3 drops of n-butanol were also added in an attempt to promote the reaction which occurred in neat butanol (described in the next paragraph); the small quantity of butanol was used to avoid swamping any possible 1H NMR hydride signal(s). The amount of hydrogen available in these NMR tube experiments would be about 0.2 mmol (2 mL H2 at 2 - 3 atm and 20 °C). For the reaction in neat butanol at ambient temperature, the procedures described in section 2. lwere used. The complex 3a (0.5 g, 0.6 mmol) was reacted with 2-8 atm H2 overnight in dry n-butanol (3 mL); complex 3a was largely undissolved initially, but the suspension gradually changed to give a brown solution and yellow precipitate, A. When the mixture was stirred, the final product had a very small particle size and was difficult to filter, this product with small particle size is hereinafter referred to as the yellow precipitate of A. Golden yellow crystals formed in the mixture if it was not stirred; this more crystalline product will be referred to as the yellow crystals of A. The yellow precipitate/crystals of A were washed with butanol, and then very small quantities of toluene (e.g., 3 x 0.2 mL). It should be noted that the volume of a vessel used for this reaction would hold 10 to 30 mmol H2 at the pressures used. The volume of gas evolved on dissolving the yellow precipitate of A was measured using a gas uptake/evolution apparatus under a 16.6 cm Hg nitrogen atmosphere, and with the solution at 28 °C. Also, elemental analyses, conductivity measurements, and further experiments to study gas evolution on dissolving in toluene-dg were performed on the yellow crystals of A. Conductivity measurements were performed on a 1:1 methanol/dichloromethane solution (-4 x 104 M). The identity of the gas evolved was determined by GC, by sampling with a syringe the gas present in a small headspace (~0.5 mL) above a dry toluene solution of the yellow crystals of A under nitrogen at ambient pressure and temperature. Reactions ofRuCh(PPh3)(isoPFA), 3b, with hydrogen in DMA, DMAICfPe, and benzenelmethanol The reaction of 3b with -10 atm H2 in DMA to give (Ti2-H2)(PPh3)2Ru(p-Cl)2(p-H)RuH(PPh3>2, la, is described in section 2.4 using the apparatus shown in Figure 2.1 In standard Schlenk apparatus, complex 3b (0.2 g, 0.25 mmol) and PPI13 (1.3 g, 5 mmol) were stirred in 5 mL DMA under 1 ami H2 for 2 days at ambient temperature. The resulting violet precipitate was washed three times with hexanes, dried, and identified by *H NMR spectroscopy. The complex 3b (0.032 g, 0.040 mmol) was stirred under H2 (7 atm) in a 1:1 solvent mixture of DMA/C6D6 (2 mL) at ambient temperature for 6 h, and the solution then transferred under H2 to an NMR tube which was subsequently sealed. No precipitate was observed. The 31P{ !H} NMR spectrum at 20 °C was recorded within a further 2 h. The complex 3b (0.24 g, 0.3 mmol) was placed in a glass tube with benzene (3 mL) and methanol (0.2 mL) under H2 (0.9 atm), and the tube sealed while partially immersed in liquid nitrogen. When the tube was warmed to room temperature, the H2 pressure would have been 2-3 atm Although not very soluble initially, the green solid, 3b, had reacted after -3 days at ambient temperature to give a reddish-brown solution and a few reddish particles. When the "solution" was heated to 40-45 °C, the solution turned brown and -0.1 g of green solid precipitated from the solution. The system was cooled to ambient temperature, and the green solid reacted slowly, over a period of 2 days, to give an orange solid that was isolated by decanting and removing the solvent under vacuum; the orange solid was analyzed by *H NMR spectroscopy. In a similar experiment, the supernatant solution was separated from the orange precipitate by decanting, and the solvent removed from the supernatant under vacuum A CDCI3 solution of the residue was analyzed by lH NMR spectroscopy. Crystals for X-ray analysis were obtained from this methanol/benzene system. Experiments were carried out, using the apparatus and procedure described in Figure 2.2 and section 2.1. In one case, 3b (0.060 g, 0.074 mmol) was placed in the tube described, into which benzene (7.5 mL) and methanol (0.5 mL) were vacuum transferred. The tube was charged with 2 atm H2 and the Teflon valve closed. The contents of the tube were then mixed and left for ~3 weeks at ambient temperature. During this time, the green 3b complex gradually disappeared, yielding a brown solution and an orange precipitate. The mixture was filtered under hydrogen, as described in section 2.1, and the solid then transferred with the aid of a small stir bar, guided by an external magnet, back into section A of the tube. Benzene (9 mL) and dichloromethane (5 mL) were then vacuum-transferred into section C of the tube, keeping part B (containing the filtrate) immersed in liquid nitrogen. It was hoped that by exposing the components to vacuum only when the solvent was frozen, decomposition might be avoided. The orange precipitate was dissolved in the benzene/dichloromethane mixture, and hexane (20 mL) then vacuum-transferred into the tube with parts A, B and C immersed in liquid nitrogen. (While most of the hexane, dichloromethane and benzene was transferred to sections A or C, less than 10% may have been transferred through the frit to section B.) When the hexane and benzene/dichloromethane solutions were mixed in section A, no precipitate appeared, and some solvent (~2 mL) was then removed from the solution by pumping under vacuum, during which time section B was immersed in liquid nitrogen. The tube was then recharged with hydrogen (~1 atm), and the Teflon valve closed; crystals slowly formed on the walls of section A. When the crystals appeared to have ceased growing, the solution in section A was transferred, under hydrogen, to another tube through the opened Teflon valve by inverting the tube while the solution in section B was kept frozen. The pressure in the tube, containing the crystals and the frozen solution (section B), was then reduced to much less than 1 arm, the Teflon valve closed, and the section A arm sealed off using a torch. The crystals could thus be removed without having air contaminate the atmosphere present in section B. The crystalline compound was identified by *H and 31P{1H} NMR, and by X-ray crystallography, as the benzene solvate of (rj2-H2)(isoPFA)Ru(|J.-Cl)2(|i-H)RuH(PPh3)2,4, (see sects. 2.4 and 3.3). Other crystals (some white, some black-red, the latter being diamond shaped) gradually grew in the benzene/methanol/ dicUoromethane/hexane solution, under hydrogen in section B. These crystals were filtered off, at the frit above section B, by inverting the tube and applying a pressure differential across the frit, and then washed with hexane. The white crystals, which proved to be dimethylammonium chloride, and the unidentified black-red crystals of Y were analyzed by spectroscopy, and the spectra are reported in section 4.3. Catalytic hydrogenations ofprochiral substrates using RuCh(PPh3)(PPFA), 3a, and RuCl2(PPh3)(isoPFA), 3b Catalytic hydrogenations of prochiral substrates were carried out using the optically active complexes (SJl)-3, at ambient temperature and 1000 psi H2, in 30 - 50 mL steel bombs equipped with pressure guages and glass liners. Work-up of the hydrogenation products from prochiral olefins was done according to the procedure described in reference 20, and yields were determined by lH NMR spectroscopy. Gas chromatography of distillates of hydrogenated solutions was used to determine yields for hydrogenation of prochiral ketones. NMR studies of the catalytic hydrogenation of 1-hexene by 3a, 3b and (rj2-H2)(PPh3)2Ru(n-Cl)2(H-H)RuH(PPh3)2, la The following general procedure was used to obtain NMR spectra of catalytically hydrogenating solutions: hydrogen gas uptake was monitored as indicated in section 2.2 in a hydrogenation flask fitted with a F-145 septum plug (from Alltech Associates) for withdrawing samples. When the rate of hydrogen uptake appeared to be reaching a maximum, a -0.8 mL sample was withdrawn using a syringe and injected into a 5 mm NMR tube filled with 1 atm Ffc and capped with a rubber septum. The NMR tube was then placed into a Schlenk tube containing hydrogen, and the capped Schlenk tube placed into a dry ice/acetone bath; freezing of the solution in the NMR tube occurred slowly. Within 20 min of sampling, the solution in the NMR tube was carefully thawed and placed into the probe of the NMR spectrometer, and then the probe was cooled as quickly as possible to ~5 °C. The NMR spectra were then recorded, the acquisition being complete within ~2 h of starting the catalytic hydrogenation. The following systems, all under 1 atm H2, were studied in this way: a) The la system with experiments at 0.04 M and 1.0 M 1-hexene in 4 mL 50:50 DMA/C6D6 with la (0.02 g, 0.004 M), and another experiment at 1.0 M 1-hexene in 4 mL 1:7 DMA/toluene-dg at the same ruthenium concentration. b) The 3a system: 0.1 M 1-hexene in 0.5 mL n-butanol/3.5 mL Cg>6 with 3a (0.050 g, 0.014 M). c) The 3b system: 0.05 M 1-hexene in 4 mL of 50/50 DMAJC^D6 with 3b (0.015 g, 0.005 M). Hydrogen-uptake measurements for 1-hexene hydrogenation All rates of hydrogenation of 1-hexene, using 3a, the species A described below, 3b and la as catalysts were measured by means of hydrogen gas uptake experiments (sect. 2.2). 4.3 Results Reactions of RuCl2(PPh3)(PPFA), 3a, with hydrogen in toluene-dg, toluene-ds/butanol and butanol The JH NMR spectrum of the toluene-dg reaction mixture formed after 12 h from rnixing showed no upfield, hydride signals. The 31P{1H) NMR spectrum (Fig. 4.14) shows that no starting complex reniained, and the main features were a pair of doublets at 8 42.7 and 45.7 with 2JPp = 40 Hz, and a less intense singlet at 8 28.9. No reaction occurred at ambient temperature in toluene-dg/butanol in the absence of H2, but after 6 h in this solvent in the presence of H2, several species were observed in the 31P{1H} NMR spectrum (Fig. 4.15). The starting complex 3a was present (8 77.4 d, 44.2 d, 2JPP = 38 Hz), and there was also a species, A, (8 71.1 d, 65.3 d, 2Jpp = 40 Hz), associated, via data from other experiments (see below), with a ]H NMR triplet at 8 -20.3 Ohn ~ 32 Hz); there is also evidence for the presence of la in the 31P{XW\ NMR spectrum (8 71 br s, under the "doublet" at 71.1, and 8 46 br s). Supporting evidence for this assignment coines from the observation of a broad *H NMR signal at 8-13.2 (Fig. 4.16; cf. 8-12.9 for la in toluene-ds, sect 3.3). The Ti measurements, given below, at 6 °C and -20 °C, indicating a dihydrogen complex, are also consistent with this assignment In addition, low temperature 31P{ JH} NMR data show that la is formed in the reaction of 3a with H2 in butanol. Another major feature in the 31P{1H} NMR spectrum (Fig. 4.15) is an AB quartet with 8 69.8 d, 69.0 d, 2JPp = 40 Hz (due to an unknown species, B) and a singlet at 8 63.8 of comparable intensity; there is also a strong, broad singlet due to PI13PO at 28.5, very similar to that appearing in the 31P{ !H} spectrum for the reaction in pure toluene-dg, where no hydrides were detected. Also, after 6 h, four upfield resonances appeared in the !H NMR spectrum (Fig. 4.16): two resonances, at 8-13.2 (br s, 'ii'), due to la and at 8 -20.3 (t, 'iv'), due to A, are mentioned above, and the other two have the following characteristics: 8 11.5 br s, with Ti = 50 ms at 300 MHz, and 8-18.51, with 2JPH = 30 Hz. The 31P{ ^H} NMR AB quartet, 'c', at 8 69.8, 69.0 for the unknown species, B, is thought to be associated with this *H NMR triplet, 'iii', at 8 -18.5 because: 1 1 1 I ' 1 1 1 I 1 1 F"' n-p-rn-p-rTTT i rr r |-rmi i i i i | t i ft rrt-rtyr-t . : | I . I I | iTn-iTrrq-rn-r| i i i i |-1 i i I | i I I i [ I I I I | '00 80 60 40 20 0 -20 -40 PPM 140 120 Figure 4.14 The 31P{1H} spectrum (121 MHz, toluene-d8, ambient temperature) of the in situ product(s) of the reaction of 3a with H2, after ~12 h from mixing; the peaks around 'a' are not identified, although one of these is probably due to Ph3PO, present as an impurity in the sample of 3a used. HO expanded spectrum A, -i 1 1 1 r-7b 70 -i 1 1 1 1—T—r—r—i 1 — i r- 55 A A' i i 1 1 1 1 1 1 1 1 r 50 45 PPM 1—i—i—i—n i—i—i—i—p lot i—i—i—i—i—i—i—r—r—i—i—'—i—I—i—i—rH—I M i—rH r~ i—|—i—i—i—i—|—i—I—I—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—r—y~i—i—i—i—|—i—i—r~i | TTT T+T p BO 60 40 20 0 I -20 PPM Figure 4.15 The 31P{1H} NMR spectrum (121 MHz, toluene-d8/«-butanol, ambient temperature) of the reaction mixture formed from the reaction of 3a with Hj, after ~6 h from mixing; peaks are identified as follows: 'a', due to 3a; 'b', due to complex A; 'c' (AB quartet), due to complex B; 'd' (partially obscured by resonances due to A), complex la; 'e', Ph3PO; unmarked peaks are unidentified. 100 (1) the ratio of the area under the 31P{1H) NMR resonances at 5 71.1 and 65.3, to the area under the AB quartet 'c', at 8 69.8,69.0 is similar to the ratio for the integrations of the two *H NMR triplets, 'iii' and 'iv', at 8 -20.3 and -18.5, respectively, and (2) the triplet 'iii', implying coupling to two equivalent phosphorus atoms, could result from coupling to the two, almost equivalent, phosphorus atoms giving rise to the AB quartet It should be noted that in other experiments, the four lH NMR upfield resonances appeared to be independent of each other numbering these resonances in order of increasing field iv Figure 4.16 The upfield 'H NMR spectrum (300 MHz, toluene-dg/n-butanol, ambient temperature) of in situ products from the reaction of 3a with H2, after 6 h from mixing; resonances T - 'iv' are discussed in the text. (Fig. 4.16), T, 'ii' (due to la) and 'iv' (due to A) appeared without 'iii'; 'ii' and 'iv' appeared without 'i' and 'iii'; and also, 'ii' appeared on its own. The Ti values (Table 4.1) for T - 'iv' indicate that T and 'ii' are truly due to dihydrogen complexes, and 'iii' and 'iv' are due to classical hydride complexes. The *H and 31P{ lH) NMR spectra, after two weeks, showed little change. Other minor species were also observed both after 6 h and after two weeks in the 31P{ 1H) NMR spectrum. On one occasion, in a similar experiment, other additional minor hydride resonances, observed at about 4 h, had disappeared after about a further 1 h, leaving only those upfield resonances shown in Fig. 4.16. The golden yellow precipitate, obtained from the reaction of H2 with 3a in neat n-butanol was not a pure compound. The NMR spectra (Figs. 4.17,4.18) of solutions of the yellow precipitate in CD2CI2 under hydrogen, even after careful washing, revealed, in Table 4.1 Ti (300 MHz) values, in ms, estimated by the inversion-recovery method, for upfield *H NMR resonances observed in the reaction of RuCl2(PPh3)(PPFA), 3a, with H2 in toluene-dg/butanol at ambient temperature Recording chemical shift (5) temperature (°Q -11.5 -13.2 -18.5 -20.3 20 50(2) 21(1) 222(8) 145(5) 6 60(2) 19(2) 264(7) 164(5) addition to resonances due to the yellow complex, A, resonances due to 3a and la: For the new yellow species, A: 3iP{ *H} NMR (121 MHz) 8 (20 °C) 71.11 (d, 2JPP = 40 Hz), 70.97 (d, 2JPP = 43 Hz), 64.72 (d, 2jPP = 40 Hz), 64.59 (d, 2JPP = 43 Hz). IH NMR (300 MHz) 5 -20.4 (t, 2jPH = 32 Hz). For 3a: 3ip{ iH) (121 MHz) 8 76.18 (d), 42.18 (d), 2JPP = 38 Hz. ^H NMR (300 MHz) 8 1.57 (d, 3JHH = 7 Hz, C-CH3), 2.47, 2.68 (2s, K-(CH3)2), 3.60 (s, Fe-Cstfj), 4.12, 4.34,4.58 (3 s, Fe-Csf/j), 622 (q, 3JHH = 6 HZ, FC-C/7); integrations were hampered by overlap with other resonances from other species in the mixture, but the data are close to those given in the Uteraturett (see also sect 2.4). For la: 3iP{ iH} NMR (121 MHz, 8 (20 °C) 70.8,46.8 (br s); 8 (-78 °C) 79.5 (br s), 64.5 (underneath the resonance at 64.5 due to the yellow species, A), 61.6 (br s), 34.8 (br s); the data correspond to those quoted in sect 3.4. !H NMR: 8 -13.1 ttFrom reference 25: Scrxaj 1.52 (d, 3H, 3JHH = 7 HZ, C-CH3), 2.44,2.66 (2d, J = 2.4 Hz, N^O/j)^, 3.56 (s, 5H, Fe-Csfy). 4.12,4.27,4.48 (s, t, s, 3H, J = 2 Hz, Fe-Cs//3), 623 (q, 1H, 3JHH = 6.4 Hz, Fc-C//), 6.9 - 7.8 (m, 25H, Ph). Figure 4.17 The 31P{ JH} spectrum (121 MHz, CD2C12, ambient temperature, under hydrogen) of the yellow precipitate from the reaction of 3a with H2 in n-butanol; peaks are assigned as follows: 'a', complex 3a; 'b', complex A; 'c' (partially obscured by the 'b' resonances), complex la. Resonances 'a' increased with time, and resonances 'b' decreased with time, as the solution turned greenish. Figure 4.18 *H NMR spectra (300 MHz, CD2CI2. under hydrogen) of the yellow precipitate from the reaction of 3a with H2 in n-butanol: a) downfield region at ambient temperature, b) upfield region at ambient temperature, and c) upfield region at -78 °C. Peak assignments are: 'i', complex 3a; 'ii', complex la; Iii', CH2C12; 'iv', complex A; V, silicone grease; unmarked peaks are unidentified. (br s); cf. cWne-dg "12-9 0* s) <sect 3-3)-Another broad singlet (8 29.3), probably due to Ph3PO, appeared in the 31P{ lH} NMR spectrum of the solution of this mixture. This singlet also appeared in the spectrum of the in situ toluene-dg reaction solution of H2 with 3a, in the absence of butanol (see above), as well as in the spectrum of a 3a catalyzed 1-hexene hydrogenation solution with 1:8 butanol/C6D6 as solvent It should also be noted that a peak at 8 24.9, due to PhsPO,21-26 appeared in the spectrum of 3a in and this peak was shifted to 8 27.4 when the spectrum was recorded in a solvent made up of 5 drops /i-butanol/1 mL C&>&. The 31P{ 1H) NMR signals due to the yellow species, A, decreased with time, while the signals due to 3a increased with time, even under H2. At the same time, the !H NMR spectra of this mixture show a decrease in the triplet associated with species A ('iv', Fig. 4.19, 8 -20.3), relative to the broad singlet at 8 -13.1 ('ii', Fig. 4.19). Figure 4.19 shows the !H NMR spectrum of a sample of the mixture within about one hour, and then after about 6 h. In addition to the above-mentioned upfield changes, corresponding changes can be seen downfield; changes in intensity can be related to the solvent resonance (CH2C12, 'iii', at ~8 5.3) and another unassigned resonance at ~8 9. The resonance T at ~8 6.2, due to the CHiCHj) in 3a, is seen to increase markedly during this time, while the 8-13.1 peak 'ii' due to la remains about the same. Further examination of the !H NMR spectrum (80-5) confirms that the concentration of 3a is increasing.with time. The NMR data show that species A decomposes to regenerate 3a. The yellow solid remains as such if kept dry and under hydrogen. But instability, when moist and iii solution, and the precipitation of other, additional species during the preparation, made work-up of the pure yellow product, A, difficult. Nevertheless, one sample of crystals was submitted for elemental analysis (Found: C 59.37, H 5.93, N 1.30, Q 9.50). The yellow crystals were only slightly soluble in dry toluene-dg, but dissolved slowly with shaking under H2 to give a pale yellow solution, which turned green a) i 1 1 1 1 1 ] 1 r—i 1 1 1 1 1 1 1 1 1 1 1 1 1 r 1 1 1 1 1 1 1 1 1 1 1 1 ] r— 10 3 6 -5 -iO -IS -20 -25 PPM Figure 4.19 !H NMR spectra (300 MHz, CD2C12) of the yellow precipitate from the reaction of 3a with H^ in H-butanol, within a) 1 h, and b) 6 h of making up the solution; peaks to note are: T, due to 3a; 'ii', due to la; 'iii', due to O^Clj; 'iv', due to complex A. Peaks due to 3a increase with time relative to those due to A. b) in ii Figure 4.19b (see legend, previous page) 1 1 1 1 1 1 1 1 r- -1 1 r within 30 min. The !H and 31P{ *H} NMR spectra of this green solution proved it to contain 3a. Signals in the *H NMR spectrum at 8 0.82(t), 1.2 -1.3 (m) and 3.3 (m) bore some similarity to butanol signals found in more concentrated solutions of butanol in toluene-dg (~2 drops in 0.7 mL) at 8 0.89 (t), 1.3 -1.5 (m), 3.1 (br s), 3.6 (t), suggesting the presence of perhaps 1 equivalent of n-butanol/Ru in the yellow crystals of A (discussed in sect 4.4). This latter spectrum also showed broad signals at 8 0.5 (s, ~5H) and 4.5 (s, ~3H). The 31P{ JH} NMR spectrum showed that no la was present, indicating that the washing of the yellow crystals of A had been successful in removing at least this impurity. This 31P{iH} NMR spectrum was the same as that of a solution of 3a, except that the integral of the 8 60 - 70 region suggested that there were possibly some signals hidden in the baseline. The volume of gas evolved from 0.0146 g (0.0015 mmol, based on the suggested formulation given later, sect 4.4) of the yellow precipitate corresponded to 0.0013 mmol of gas. The gas evolved from a toluene solution of the yellow crystals of A was identified by GC as H2. The equivalent conductance, A of a sample of the yellow crystals of A, calculated on the basis of the formulation, was 30 SHmoHcm2, which compares with 63 Q-^moHcm2 found for NEtiCl using the same solvent and molar concentration. The reaction of 3a with H2 to give the yellow species was also carried out in benzene/hexane/methanol (6:7:2) and in benzene/methanol (~6:1) solvent systems. The variable-temperature *H NMR spectrum of an isolated solid mixture formed in benzene/hexane/methanol solution is shown in Figure 4.20. The main upfield resonances were: OCD2CI2 (18 °C) -11.4 (t, JPH = 17 Hz), -13.1 (br s), -18.55 (t, JPH = 30 Hz), -20.3 (t, JPH =32 Hz); this last noted triplet corresponds to species A, formed in the reaction in the presence of butanol. "i""'"M|,i -1» -20 ' I'1'1'111 1 I -21 PPU -22 Figure 4.20 Variable temperature *H NMR spectra (300 MHz, CD2CI2) of me yellow deposit from the reaction of 3a with hydrogen in 6:7:2 benzene/hexaneAnethanol; spectra were observed at a) ambient temperature, b) -15 °C and c) -55 °C. Reactions ofRuCh(PPh3)(isoPFA), 3b, with hydrogen in DMA, DMAIC£>6, and benzenelmethanol The *H NMR spectrum of Rua2(PPh3)(isoPFA), 3b, is shown in Figure 4.21. The violet precipitate, obtained from the reaction of 3b H2 in DMA in the presence of PPh3, was identified as RuHa(PPh3)3 by lH NMR (80 MHz, CDC13, ambient temperature, under nitrogen): 8 -17.4 (q, 2JPH = 25 Hz, Ru-H) (cf. reference 81); the yield was 90%. After 3b had reacted with H2 for 2 h in DMA/C^D^ in the absence of added PPh3, peaks due to the following four major species were observed in the 31P{ 1H} NMR spectrum (Fig. 4.22): (i) unreacted 3b: identified by the two doublets at 8 35.7 and 94.7 with 2JPP =32 Hz. The spectrum compares closely with that given in section 2.2 and with that for a CDCI3 solution containing 2 drops DMA per mL (8 36.05 (d), 95.25 (d), 2JPP = 32 Hz). (ii) (rj2-H2)(isoPFA)Ru(u-a)2(ii-H)RuH(PPh3)2,4: identified by the resonances at 8 78.6 (dd), 76.8 (s), 64.5 (dd). These and the coupling constants compare closely with the values for the CD2CI2 spectrum (sects. 2.4,3.3). (iii) la: identified by the two broad singlets at 8 70.3 and 46.0 (cf. sects. 2.4 and 3.3). (iv) an unidentified species with a broad singlet at 8 -13.9. Also observed were: (vi) RuHCl(PPh3)(isoPFA), 7: identified by two doublets at 8 98.0 and 61.0 (cf. values given in sect 2.4). (vi) RuHCltPPh^: the singlet at 8 59.3 is close to that reported for the CH2C12 solution of this complex (8 59.0).74 and several unidentified minor resonances. The unwashed orange solid deposited during the reaction of 3b with H2 in methanol/benzene contained mainly 4, identified by *H NMR (sect 2.4), and also another unidentified hydride Z, (*H NMR: 8 -12.8 (t)). The *H NMR (CDC13,400 MHz) Figure 4.21 The *H NMR spectrum of complex 3b. Detailed assignments are given in sect. 2.2; peak 'a' is an unidentified impurity, 'b' is due to silicone grease and 'c' is due to CHC13. -i—i—i—i—|—I—i—i—i—|—i—r—i—i—I—i—i—i—i—r—I—i—I—i—r~r—r—i—i—i—i—i—i—i—|—t—i—r 65 60 55 50 1 IWlHiMOli^OI •in TTiT7Tn-|TTi II —; i i i r r i i i i i r i i i i i i i i i i I i i I i i i i [i 'nTrn-rn-r i i r-j-r i I i 1 i I i i r 120 100 80 60 40 20 0 -20 PPM 9 1 i | i I i r Figure 4.22 The 31P{ IH) spectrum (121 MHz, ambient temperature) of a solution obtained after ~6 h from mixing 3b in 1:1 v/v CgDg/DMA under H^, peaks are assigned as follows: 'a', complex 3b; 'b', complex 4; 'c', complex la; 'd\ complex 7;'e', RuHCl(PPh3)3. Unmarked peaks are unidentified. An expansion of part of the spectrum is shown in the inset. spectrum of the residue of the supernatant of the reaction of 3b with H2 in methanol/benzene is shown in Figure 4.23. Although some decomposition undoubtedly 1 d -10 -20 •30 Figure 4.23 The 'H NMR spectrum of the residue obtained by removal of solvent from the supernatant of the reaction of 3b with H2 in 1:15 v/v methanol/benzene. Peak assignments are as follows: 'a', complex Z; V and 'c', complex 4; 'd', complex Y. Unmarked peaks are unidentified. Some decomposition has probably occurred as a result of removing the H2 atmosphere. occurred as a result of loss of hydrogen pressure, some peaks are recognizable: the 8 -14.5 (v br s) and -18.6 (t) peaks of 4 (see above), the 8 -12.8 (t) peak of the unidentified hydride Z, and a 8 -21.8 (q) peak of species Y, not yet identified but discussed more fully below. The white crystals of dirrxthylarnmonium chloride, which grew in the benzene/methanol supernatant, to which some dUchloromethane/hexane solvent had been transferred, exhibited a single lH NMR (CD3OD, 80 MHz, ambient temperature) singlet at 8 2.75, the same as mat of an authentic sample. The mass spectrum (Fig. 4.24) shows peaks at m/z 46,45,44,38 and 36,37 and 35; the mass spectrum of an authentic sample showed the same peaks, with the same relative intensities. The unidentified black-red, diamond-shaped crystals of Y exhibited the 31P{1H} NMR spectrum shown in Figure 113 u u I 1 I I I I I I I I I III I I 1 I I IW I I I I I IU 1111] " 1 ".i " J " -i i i "Ml II i i .. I i i i i i i i i i I i i i i i i i i i I Iff III Ml m/z Figure 4.24 The EI mass spectrum of the white crystals of dimethylammonium chloride isolated from the reaction of 3b with H2 in 1:15 v/v methanol^enzene. 4.25; the *H NMR spectrum shown in Figure 4.26 is partly obscured by hexane signals around 8 1. No signals were observable in the IR (KBr pellet) in the region 1700 to 2300 cm*1. 114 Figure 4.25 The 31P{1H) NMR spectrum (121 MHz, CI^Cl* ambient temperature, under hydrogen) of the blackish red crystals of complex Y isolated from the reaction of 3b with H2 in 1:15 v/v methanol/benzene. Expansions of the three peaks due to Y are shown in the insets; minor peaks marked with an 'a' are unidentified impurities. Figure 4.26 The 'H NMR spectrum (300 MHz, CD2O2. ambient temperature, under hydrogen) of Y. Peaks are assigned as follows: 'a', CH2CI2; V, hexane; 'c', methanol; 'd', peaks tentatively assigned to complex Y. Catalytic hydrogenations ofprochiral substrates catalyzed by (SJi)-RuCl2(PPh3)(PPFA), 3a, and (SJi)-RuCl2(PPh3)(isoPFA), 3b Results of these studies are presented in Table 4.2. NMR studies of the catalytic hydrogenation of 1-hexene by (rt2-H2)(PPh3)2Ru(ri-Cl)2(H-H)RuH(PPh3h, la, RuCl2(PPh3)(PPFA), 3a and RuCl2(PPh3)(isoPFA), 3b The la system in DMA/CyD^ a) at 0.04 M hexene, the 31P{1H} NMR spectrum (Fig. 4.27) of the active hydrogenation solution, measured as described in section 4.3, showed the presence of mainly la. The minor peak at 8 24.2 is probably due to Pl^PO,21-26 while other minor unassigned peaks are present at 8 -72.5,65.5, 60.2 and 51.2. rm t 'T~t t—r Figure 4.27 The 31P{1H) NMR spectrum (121 MHz, 1:1 v/v DMA/C$D6,5 °C, under hydrogen), acquired within 2 h of starting the reaction, of the la-catalytic hydrogenating solution containing initially 0.04 M 1-hexene. Peaks marked 'a' are due to la; peak V is due to PI13PO; other peaks are unidentified. b) at 1.0 M hexene, the total H2 uptake after 2 min was 0.024 M and at this rapid reaction rate the capacity of the buret would have been exceeded in another 4 min. Therefore, the uptake apparatus was adjusted to allow free, unmeasured flow of H2 to the reaction solution. The 31P{1H} NMR spectrum (Fig. 4.28) of this active hydrogenating solution showed no trace of la. There is a minor peak at 8 24.2 due to Ph3PO, and a peak at 8 58.5 possibly due to RuHCKPPlvjb; the other peaks in this spectrum at 8 72.5,65.5,60.2 and 51.2, which also appear in the 0.04 M Table 4.2 Hydrogenation results for various substrates using (S,/c>RuCl2(PPh3)(PPFA), 3a, and C5^;-RuCl2(PPh3)(isoPFA), 3b, as catalysts8 With (S#)-3a as catalyst:b catalyst (MxKP) substrate (M) 1.4 CH2=C(COOH)NHCOMe (0.11 M) 1.7 (£;-PhCH=C(COOH)NHCOMe (0.076 M) 1.6 CH2=C(CCX)H)CH2COOH (0.12 M) 1.6 EtC(0)Me (0.13 M) 1.6 PhC(0)Me (0.17 M) With (SJl)-3b as catalyst:b 1.1 CH2=C(COOH)NHCOMe (0.10 M) 1.9 ('£;-PhCH=C(COOH)NHCOMe (0.076 M) 1.7 CH2=C(COOH)CH2COOH (0.12 M) 1.3 EtC(0)Me (0.13 M) 0.9 PhC(0)Me (0.17 M) time (h) 23 23 27 23 23 25 23 27 18 22 solvent butanol methanol/ benzene (1:9) methanol/ benzene (1:9) methanol/ benzene (1:9) butanol methanol/ benzene (1:9) yield (%) 100* 90 eec 0 0 100 0 0 0 100 0 100 100 0 0 0 •At 1000 psi H2. bThe optical purity of the chiral ligands (84%) for catalysts was based on rotation measurements on (S^>l-[a-/v'^-dimethylammc«thyl]-2-diphenylphosphinoferrocene and (S>-l-[a-N^-dimethylammc«thyl]ferrocene, giving optical purity = 84 %. •The optical purity of the hydrogenated products was measured by rotation. dAt 10 atm H2 for 24 h, yield = 80%. 117 a 90 70 60 b0 40 30 20 PPM Figure 428 The 3,P{1H) NMR spectrum (121 MHz, 1:1 v/v DMA/CgDg, 5 °C, under hydrogen) of the la-catalytic hydrogenating solution containing 1.0 M 1-hexene, the NMR data being acquired within 2 h of starting the reaction. The peak marked 'a' is due to RuHCKPPh^; peak V is due to PhjPO; other peaks are unassigned. hexene spectrum, are unassigned; also unassigned is a broad resonance in the 8 78 -81 range. The la system in DMA/toluene-dg: After ten minutes of shaking under hydrogen and 0.013 M H2 had been absorbed, some of the la was still undissolved, showing that the amount of la used in this experiment exceeded the solubility limit The rate of uptake was approaching a maximum at this time, and the shaking of the flask was stopped to allow transfer of solution to the NMR tube. After the shaking had been stopped for about 1 min, the solution colour had changed from red to a brownish-maroon and the remaining solid then dissolved. Uptake of >0.08 M H2 was also observed within 3 min of stopping the shaking; the exact amount of H2 could not be measured as it exceeded the capacity of the buret. (Behaviour similar to that described above was also observed when shaking was stopped in other experiments at 1 M 1-hexene in 1:1 DMA/C^.) The 31P{1H} NMR spectrum (Fig. 4.29), acquired at 5 °C within 2 h of starting the DMA/toluene-ds reaction, showed the presence of Ph3PO at 8 24.1, unidentified peaks at 8 52.1, -15.3 and -30.2, and possibly resonances, scarcely above noise level, in the 8 55 - 80 range. The *H NMR spectrum, acquired at 5°C after 2 h from the start of the reaction, showed only one upfield hydride resonance at 8 -17.85 (q, 118 2JPH = 26 Hz) due to RuHCl(PPh3)3. Subsequent acquisition of the 31P{ ^H] NMR spectrum at -58 °C led to confirmation of the presence of RuHCl(PPh3)3 (8 94.2 (t, IP), 38.8 (d, 2P), 2JPP = 27 Hz; cf. reference 81); the -58 °C spectrum also included peaks at 8 52.0 and 24.5, due to an unidentified species and Ph3PO, respectively. 5 t 1 * 1 i, T-r^-JT-»-.-»7 "TTpr^T-p-J, qiT" rTr^|WTnrrrrrTT-- I t • 1 I-) t T t 1 t • • " 1 T-T n 1 ' • I; T • • • I • T I • I T T'TTf J • | • T T • T ^ * * ] ' ' ' * 'n • | ' 1 ' 1 | " ' * \ '(••• 140 120 10~ PO I'O 40 20 0 -20 -40 -60 PPU -AO Figure 4.29 The ''P^H) NMR spectrum (121 MHz, 1:7 v/v DMA/toluene-d8,5 °C, under hydrogen) acquired during the second hour after the start of the la-catalyzed hydrogenation reaction with 1.0M1-hexene. The peak marked 'a' is due to Pti3PO; other peaks are unassigned. The 3a system: When, at 9.5 min from starting the reaction, the hydrogen-uptake part of the experiment was interrupted in order to obtain an NMR sample, the hexene-to-hexane turnover number had reached 3.3. The 31P{!H} NMR spectrum (Fig. 4.30) of this hydrogenation solution shows peaks indicating the presence of mainly 3a and PhjPO, as well as a minor peak at 8 61.7, all of which were also observed at ambient temperature for a solution of 3a in a solvent made up of 5 drops butanol and 1 mLQDg, without H2 or hexene present; the Ph3PO was also present in a QD6 solution of 3a. Unidentified peaks barely above noise level, appear at 8 42.1 and 57.9 in the spectrum in Figure 4.30. No upfield hydride signals were observed in the *H NMR spectrum, which was acquired at 5 °C two hours after starting the reaction. a a f -:.J. •" ....... 120 •••'('••' 100 ' 1 ' ' 1 ' to • •' i • •' • i to ,,,,,,, 40 b Figure 4.30 The 31P{!H} NMR spectrum (121 MHz, 1:7 v/v n-butanol/QjDe, 5 °C, under hydrogen), acquired within 2 h of starting the reaction, of a 3a-catalytic hydrogenating solution containing 0.1 M 1-hexene and 0.014 M 3a. The peaks marked 'a' are due to 3a; peak V is due to PI13PO; other peaks are unassigned. The 3b system: The 31P{1H} NMR spectrum (Fig. 4.31) of this hydrogenation solution shows the presence of mainly 3b and PI13PO, as well as other unassigned minor peaks. <G C V "a V mm\ mm iiwUli 11 ii \ lf^ifn i' -* 1 -* ~— 1 jtjtJUk I j I I I I | I M I | I I I I | I I M | I I I I | I I I I | I I } I | I M I | I I I I | I I M | I I I I | I I I I | I I I I | I I II | I I II | I I I I 120 100 8C 60 40 20 0 -20 PPM Figure 4.31 The 31P{1H) NMR spectrum (121 MHz, 1:1 v/v DMA/C<p6, 5 °C, under hydrogen), acquired within 2 h of starting the reaction, of a 3b-catalyzed hydrogenating solution containing 0.05 M 1-hexene and 0.005 M 3b. The peaks marked 'a' are due to 3b; peak V is due to Pti3PO; other peaks are unidentified. Kinetic data from hydrogen-uptake measurements for 1-hexene hydrogenation With (Tt2-H2)(PPh3)2Ru0i-a)2(u-H)RuH(PPh3)2, la, as catalyst at a total ruthenium concentration of 6.2 x 104 M and a 1-hexene concentration of 0.02 M in DMA at 30 °C under 1 atm H2, the maximum rate of uptake was 1 x lf>5 M s1. With compound A as catalyst, employing 3 x 104 M compound A and 0.1 M 1-hexene in dry n-butanol under 1 atm H2, the total H2 uptake after 2 h corresponded to <10% of a quantitative conversion of 1-hexene to hexane. The golden colour of the solution obtained on dissolution was retained overnight. More extensive data were obtained using RuCl2(PPh3)(isoPFA), 3b as catalyst. For a 1-hexene hydrogenation system in DMA at 30 °C, the total gas uptake corresponded to virtually complete reduction of the 1-hexene to hexane. Isomerization of the 1-hexene was not detected at stages prior to complete hydrogenation. The uptake plots (Fig. 4.32) show an initial autocatalytic period prior to reaching a maximum rate, usually after 500 to 4000 s; the rate then gradually falls off. The maximum rates were used to determine the 2.4 _ 2.0 _ £ 1.6-& 1.21-3 0.8 r-0.4 h J L l l I I 234 5 6 7 8 9 10 I! Time (s x 10"3) Figure 4.32 H2 uptake plot for 1-hexene hydrogenation catalyzed by 3b, with 1 atm H2,0.030 M 1-hexene, 4.3 x 10"4 M PPh3 and 52 x 10"* M 3b in DMA at 30 °C. kinetic dependences. The data are summarized in Figures 4.33 to 4.37. The dependence on [Ru]toui goes from first to half-order with increasing concentration; with added PPh3, the dependence is more nearly first-order (Figs. 4.33,4.34). There is a first-order dependence on 1-hexene and on H2 (Figs. 4.35,4.36), and a complex dependence on added phosphine, with the rate initially increasing somewhat with phosphine concentration, and then becoming inhibited at [PPI13] > 1 x lO3 M (Fig. 4.37). A colour change (from green to violet) of the hydrogenating solutions containing added phosphine was observed during gas-uptake measurements. 2.0 4.0 6.0 8.0 [Ruh- (M x 104) 10.0 12.0 Figure 4.33 Dependence of maximum rate of hydrogenation of 1-hexene in DMA on total Ru concentration (3b): °, with no added PPh3; with 6.5 x 10"3 MPPha added VO O 70 r 6.0 -S 50 % 4.0 £ 20 3.0 15 20 25 30 35 [Ruix (M1/2 x 103) Figure 4.34 Dependence of maximum rate of hydrogenation of 1-hexene in DMA on the square root of the total Ru concentration (3b). [1-hexene] (M) Figure 4.35 Hexene dependence of maximum rate of hydrogenation of 1-hexene in DMA at 30 °C, 1 atm total pressure and total Ru concentration (3b) = 5.1 x lO^M. 0 0.5 1.0 1.5 i.0 [HJ (M x 103) Figure 4.36 H2 dependence of maximum rate of hydrogenation of 1-hexene in DMA at 30 °C, with total Ru concentration (3b) = 5.1 x 10^ M and [1-hexene] = 0.06 M. I O »—i X C/3 a, 9 1.0 -.8 -.6 -.4 -.2 -0 o o _L 0.2 0.4 0.6 0.8 [PPhJOlxlO2) Figure 4.37 Phosphine dependence for the reciprocal of the maximum rate of hydrogenation of 1-hexene (0.03 M) in DMA at 30 °C and 1 atm total pressure, with total Ru concentration (3b) = 5.1 x 10"4 M. With (Tj2-H2)(isoPFA)Ru(u-Cl)2(u-H)RuH(PPh3)2,4 as catalyst, using a total ruthenium concentration of 3 x 10"4 M and a 1-hexene concentration of 0.02 M in DMA at 30 °C under 1 atm H2, the maximum rate of uptake was 8 x 10"6 Ms"1. 4.4 Discussion The (7}2-H2)(PPh3)2Ru(p:-Cl)2(^-H)RuH(PPh3)2, la, 1-hexene hydrogenation system At the relatively low hexene concentration of 0.04 M and at 1 atm H2, the hydrogenating solution exhibits a 31P{ lH) NMR spectrum showing that la is the main species present. This species was not observed at the high hexene concentration of 1 M; however, it must be remembered that stopping the shaking in order to sample this hydrogenating solution, even while under 1 atm H2, would result in depletion of dissolved H2. That this H2 depletion actually perturbed the distribution of ruthenium species in solution is evidenced by the colour change from red to brownish-maroon (p. 117) and -immediate dissolution of undissolved la, present above its solubility limit in experiments at high hexene concentration. Furthermore, it was shown in section 3.3 that la reacts, in the absence of hydrogen, with 1-hexene to give hexane and RuHCl(PPh3)3; thus, in the 1 M hexene hydrogenating solutions under 1 atm H2, the same reaction appears to occur, when shaking under hydrogen is stopped in order to sample, and the presence of RuHCl(PPh3)3 is, in fact, indicated by the 31P{ lH) and lH NMR spectra of these solutions. That a somewhat altered catalytic system is present after stopping and then resuming the shaking is evidenced by the rapid uptakes observed and the fact the colour remains brownish-maroon (p. 117). It could be inferred then that, in the 1-hexene hydrogenation system present with continuous shaking, complex la is catalytically reacting with 1-hexene both at high and low hexene concentration and is being regenerated by reaction of an intermediate with hydrogen. Other rninor species are indicated by the 31P{ lH) NMR spectra, but have not been identified; neither is it known whether these species are intermediates in the catalytic cycle involving la, or whether they are side-reaction products. The mechanisms for hexene cwrdination and hydride transfer are not known; however, breaking of a Ru-((J.-H) bond, which might occur during the rapid hydrogen ligand exchange process (sect. 3.3, 3.4), could provide a vacant site for coordination of hexene. Rapid transfer of rj2-H2 or hydride ligands to the hexene, elimination of the hexane and regeneration of la could then occur. A possible intermediate might be the hypothetical species (PPh3)2Ru(ji-Cl)2(u.-H)RuH(PPh3)2, discussed in section 3.4, which is thought to react rapidly with Kfe to regenerate la, and may not be easily observed under H2. Thus, in explanation of the near first-order dependence on [RU]T, the first- to zero-order dependence on 1-hexene, the mainly first-order dependence on hydrogen at high substrate concentration, and the first- to zero-order dependence at lower substrate concentration, the following mechanism is proposed: (Ti2-H2)(PPh3)2Ru(^-Cl)2(u-H)RuH(PPh3)2 + hexene -—- (PPh3)2Ru(u-Cl)2(u-H)RuH(PPh3)2 + hexane , (4.10) ,(PPh3)2Ru(H-Cl)2ai-H)RuH(PPh3)2 + H2 =£3= (Tl2-H2)(PPh3)2Ru(U-Cl)2(^-H)RuH(PPh3)2 (4.11) where ki, k2 and k.2 are rate constants. Evidence for the equilibrium 4.11 is obtained from a study21-23 of the reaction of la with PPh3 (discussed in sect. 3.4), and the k_2 step is therefore written into the mechanism proposed here. Applying a steady-state treatment to the intermediate, (PPh3)2Ru(jx-Cl)2(li-H)RuH(PPh3)2, gives the rate equation: P:itr _d[H2] _ kik2[hexene][H2][Ru2jT  Kale~ dt " k.2 + k2[H2] + ki[hexene] where [RU2]T = [(Tl2-H2)(PPh3)2Ru(u-a)2(^-H)RuH(PPh3)2] + [(PPh3)2Ru(^Cl)2(^-H)RuH(PPh3)2] = V2[Ru]T (4.12) (4.13) 125 The initial induction period (Fig. 4.1) preceding the maximum rate of H2 uptake is probably due to the dissolution of the complex. The rate equation (4.12) can be written as: Rate = k*[Ru]T l (4.14) 2kik2[hexene][H2] where k* " k.2 + k2[H2] + kithexene] (4-15Thus, the above mechanism would produce a first-order order dependence on [RU]T- The data in Figure 4.2, obtained by Thorbum,71 follow approximately first-order behaviour, at least at lower [RU]T (£ 2 x 10"3 M, the concentration used in most of Thorburn's uptake experiments). The somewhat less than first-order Ru dependence at higher [RU]T, however, is probably real and is unexplained by this mechanism. A possible reason, admittedly speculative,, for this fall-off in rate with increasing Ru concentration would be an interaction between one of the dinuclear complexes in equations 4.10 and 4.11, with a mononuclear species in solution to give a RU3 species. In any case, the fall-off in rate at higher [RU]T does not rule out the mechanism described above, at least when Ruj < 2 x •10-3 M. At lower hexene concentrations, the ki [hexene] term in the denominator of equation (4.12) could possibly become small compared to Qc.2 + k2[H2]), and the rate equation approximates to equation 4.16, which shows a first-order dependence on hexene: |kik2[Ru]T[hexene][H2] (4.16) Rate = k.2 + k2[H2] At higher hexene concentrations (>1 M), the ki[hexene] term becomes the most important in the denominator, and equation 4.12 approximates to: Rate = k2[Ru]TrH2] . (4.17) which shows independence of hexene concentration. Rearrangement of equation 4.12 gives: _j_ k.2 + k2[H23 1_ ?__ (4<18) Rate kkaEHdERnlT [hexene] 5k2[Ru]T[H2] In fact, in accordance with equation 4.18, a plot of 1/Rate versus l/[hexene] at constant [RU]T and [H2] yields a straight line (Fig. 4.4), the intercept of which is used to calculate a k2 value of 160 M-h'} At the lower olefin concentration conditions where the rate equation approximates to 4.16, the proposed mechanism requires that the dependence on hydrogen should go from first- toward zero-order as the [H2J increases, as seen experimentally (see data for 0.3 M hexene experiments, Fig. 4.5). When the olefin concentration is high (>1 M), the rate equation approximates to 4.17, and the mechanism requires that the rate should show a first-order dependence on hydrogen over the complete range of hydrogen concentrations. While experiments on H2 dependence at [hexene] > 0.7 M have not been carried out, the data for experiments at 0.7 M hexene do show the increased tendency toward first-order behaviour. Values for rate constants can be obtained using a rearranged form of equation 4.12: _1 k.2 + ki [hexene] 1 1 ^ ^ Rate " ^k1k2[hexene][Ru]T [H2] ^i[Ru]T[hexene] -In accordance with equation 4.19, plots of 1/Rate against I/TH2] at constant [Ru]x and [hexene] at both hexene concentrations yield straight lines (Fig. 4.6), the intercepts of which are used to determine ki values of 0.56 and 0.58 M^s-1. Since the value of k2 is known (vide supra), these values of ki and k2 can be used to derive the value of k.2 from the slope of one of the lines plotted according to equation 4.19 (Fig. 4.6), and hence the value of k2/k-2. Thus, for a hexene concentration of 0.2 M, a value for k.2 is calculated to be 0.020 s-1, which yields a value for k2/k-2 of 7.9 x 103 M*1. Calculation of k.2 could also be carried out, if necessary, for a hexene concentration of 0.7 M. Further, equation 4.18 and the slope of the line in Figure 4.4 could also be used to obtain a value for L2; however, corresponding calculations carried out by Thorburn show that the data are self-consistent, and thus the alternate calculations are unnecessary in the present work It is of interest in connection with this value of K that work on the reaction of la with PPh3 yields an estimate of K of 5 x 104.23 For experiments at constant [RU]T, [H2] and [1-hexene], the rate of hydrogenation decreases with increasing [PPh3]. Addition of PPh3 to la is known to give a solution containing the monomeric RuHCl(PPh3)3 complex, with evolution of hydrogen.23 The colour of the solution changes with increasing added PPI13 from the red of la to the violet colour of RuHCl(PPh3)3 and is now thought to proceed by an initial equilibrium involving loss of the T|2-H2 ligand, followed by reaction with PPh3: (Tl2.H2)(PPh3)2Ru(p-a)2(p-H)RuH(PPh3)2 H2 +. • (PPh3)2Ru(p-CD2(p-H)RuH(PPh3)2 (3.14) (PPh3)2Ru(p-Cl)2(p-H)RuH(PPh3)2 + 2PPh3 2RuHCl(PPh3)3 (3.15) Added PPh3 also suppresses the rate of hydrogenation catalyzed by RuHCl(PPh3)3;2 thus, although RuHCl(PPh3)3, one of the most active platinum metal catalysts known for the hydrogenation of terminal olefins (see section 4.1), is probably formed on addition of PPh3 to the hexene hydrogenation system, its formation is not inconsistent with loss of activity. The reason for the increase in rate with increasing [LiCl] is still unknown, but the increased rate is probably due to reaction of one of the dimeric species in equations 4.10 and 4.11 with chloride to give a more active catalyst Hydrogenations involving RuCl2(PPh3)(PPFA), 3a Reactions of 3a with H2: The complex 3a reacts with hydrogen in toluene in the absence of alcohol to give non-hydridic species; hydride and dihydrogen species are presumably produced by the reaction of 3a with H2 in butanol. The hydrides were not observed in situ with 1 atm H2 in the previous work,25 and the reported absence of 31P{ lH} NMR signals25 must have resulted from using solutions that were too dilute. The only product positively identified in the present work is the complex, (T|2-H2)(PPh3)2Ru(u,-Cl)2(H-H)RuH(PPh3)2, la, discussed in chapter 3, which must occur following a phosphine ligand redistribution reaction of 3a. The isolated golden yellow hydride, A, was not fully characterized, but must be closely related to 3a, since it decomposes in CD2CI2, and in toluene-dg, even under H2, to regenerate 3a, which was detected by its green colour and JH and 31P{ lH] NMR spectra. Of importance, a butanol solution of the yellow hydride under 1 atm H2 retained the yellow colour overnight (vide supra), suggesting that the alcohol solvent stabilizes the hydride. This behaviour is also consistent with the report of regeneration of the green colour of 3a, even under H2, when work-up of hydrogenated butanol solutions was attempted25 Spectral changes in Thorburn's UV-vis experiment of subjecting hydrogenated solutions of 3a to argon in toluene/methanol65 (sect. 4.1) were consistent with regeneration of 3a. Also, in the in situ experiments in toluene-dg/butanol, the reaction between and H2 did not go to completion even though hydrogen was still available. These findings, coupled with Thorburn's H2 uptake experiments which showed decreasing H2:Ru stoichiometries with increasing temperature,65 suggest that an equilibrium exists between 3a and the yellow hydride complex, A. Based on the above observations, a possible formulation for A is shown within a suggested equilibrium (eq. 4.20). The proposed structure involves a zwitterionic species stabilized by coordinated butanol of some type, either as shown, or involving the hydride ligand. In CD2CI2 or toluene-ds, loss of the coordinated butanol could result in destabilization, and subsequent reverse reaction between the hydride and nitrogen-stabilized proton to give H2 and the reactant complex, 3a, as observed in the gas evolution experiments. The presence of a hydride ligand cis to two almost equivalent, H Ph3P. P BuOH a R S. H RuCl2(PPh3)(PPFA) + H2 a / H Bu (4.20) mutually cis phosphine ligands is consistent with the observed highfield pseudo-triplet (2JPH = 32 Hz) in the lH NMR spectrum, and the magnitude (2jpp = 40 Hz) of the 31P{ ]H} NMR coupling constants.89-91 The pale yellow colour of this hydride is consistent with a six-coordinate complex; five-cc»rdinate Ru(IT) hydrides are typically intensely red-coloured,90 an observation born out in the present thesis work (sect. 5.3). The elemental analysis for A is generally reasonable, but the carbon analysis is low, and the chlorine analysis is high: Calcd. for C48H530NCl2FeP2Ru: C 60.44, H 5.56, N 1.47, Cl 7.44. Found: C 59.37, H 5.93, N 1.30, Cl 9.50. The 3ip{lH} NMR spectrum indicated that some PI13PO was present. Also, the conductivity measurement corresponds to half a mole of a 1:1 electrolyte; this might be due to 0.5 mole of an ionic chloride impurity, consistent with the chloride analysis which is high by roughly 1/2 Cl. The proposal implies a reaction showing heterolytic cleavage of H2 promoted by the amine functionality of the coordinated PPFA ligand; indeed, one goal of the original P-N ligand work25 was to (tetermine if such a reaction might occur. There is limited direct data to support heterolytic cleavage of H2 at a metal centre,110 and the subject is particularly important at Ru(U) centres where monohydride formation within a mononuclear species would otherwise have to invoke Ru(TV) dihydride intermediates.25 Involvement of the 130 amine group in the reaction of the analogous complex, RuCl2(PPh3)(isoPFA), 3b, with H2 in benzene/methanol (15:1) is demonstrated above (sect 4.3), where protonation of the -NMe2 group, results in the formation of H2NMe2+Cl~. The yellow complex formed in the present PPFA system may represent an arrested mtermediate in a reaction analogous to that in the isoPFA system The presence of two overlapping doublets in the 3 ^P^H} NMR spectra of A at 8 71.11 and 70.97, and also at 8 64.72 and 64.59 may be due to the presence of diastereomers in solution: the PPFA ligand may be R,S or SJI, and the ruthenium centre in the above structure is also chiral. Hydrogenation systems catalyzed by 3a: Complex A was found to be relatively inactive as a 1-hexene hydrogenation catalyst in H-butanol under 1 atm hydrogen. It is therefore not a possible intermediate in the 1-hexene hydrogenation system catalyzed by 3a in neat butanol, nor is it observed in the 31P{ !H) NMR spectrum, observed within two hours of starting the reaction, of a hexene-hydrogenating solution with 1:7 /j-butanol/C6D6 as solvent (Fig. 4.30). The complex (T|2-H2)(PPh3)2Ru(p-Cl)2(p-H)RuH(PPh3)2, la, formed by the reaction of 3a with several atmospheres H2, is also not observed in the spectrum shown in Figure 4.30, and the activity of la is not high enough for traces of it to account for the activity of the 3a system. The complex 3a is the main ruthenium species clearly indicated in the spectrum shown in Figure 4.30, the period of acquisition being from 0.5 to 2 h from the start of the 1-hexene hydrogenation reaction. Although the appearance of the 31P NMR spectrum is very similar to that of the butanol/C6D6 solution of the starting complex 3a, there is a very low intensity broad resonance in the 8 58 - 68 range, in addition to weak resonances at 8 58 and 42, and any of these may be associated with intermediates in the catalysis. There is no indication of the presence of PPh3 to support the proposal (sect. 4.1) of a pre-equilibrium loss of PPh3 to account for the inverse dependence on added PPh3, the approach toward 131 half-order in ruthenium dependence at higher ruthenium concentrations, and the first-order ruthenium dependence in the presence of added phosphine. However, the dissociated PPh3 may undergo a subsequent reaction; or it may be that free phosphine is not observed because concentrations of catalytically active species are low and also unobserved. No hydrides were observed in the *H NMR spectrum of a hexene-hydrogenating solution with 1:7 butanol/C6D6 as solvent But the fact that Ru(IT) hydrides and dihydrogen complexes are formed in the presence of butanol under 2-3 atm H2, makes it likely that they are also formed under 1 atm H2, though perhaps at concentrations too low to be observed. It is therefore proposed that an intermediate Ru(TJ) hydride is initially formed, which can undergo a reversible phosphine dissociation, and that the dissociated ruthenium species is involved in the catalytic hexene hydrogenation cycle. A modification of the mechanism outlined in equations 4.7 and 4.8 is therefore proposed: Ru(II)H(PPh3). KB. Ru(JJ)H PPh3 (4.21) Ru(II)H K2 hexene Ru(TI)(hexyl) H2 Kj Ru(n)H(Ti2-H2) hexene k2 H2 Ru(U)H + hexane (4.22) The question was raised earlier concerning the accessibility of RufTV) in catalysis at Ru(U) centres. The general tendency of Ru(TJ) to form ri2-H2 complexes (sect. 3.1), the tendency in particular of 3a to form such complexes, at least under 2-3 atm H2 (Fig. 4.3, Table 4.1), and the work discussed above on the (rj2-H2)(PPh3)2Ru(u-Cl)2(u-H)RuH(PPh3)2, la, hexene hydrogenation system, where a rj2-H2 complex seems to be involved in a hexene hydrogenation cycle, suggest that a Ru(II)H(r|2-H2) intermediate is more likely than a classical Ru(TV)H3 species. Hydrogenations involving RuCl2(PPh3)(isoPFA), 3b The reactions of 3b with H2 in DMA, DMA/toluene-dg or methanol/benzene are complex, each giving several products, at least partly owing to phosphine ligand redistribution. Not all the products have been identified. Two products, which are considered in chapter 3, are well-established as la and (r|2-H2)(isoPFA)Ru(}i-Cl)2(jj.-H)RuH(PPh3)2,4, and are among the products formed in the presence of DMA; 4 is also formed in methanol/benzene, along with H2N(CH3)2+C1" and other products, one being a black-red phosphine ruthenium hydride, Y. In this reaction, the arnine functionality of the isoPFA almost certainly accepts a proton, as judged by the formation of H2NMe2+Cl_, and promotes a 'heterolytic' H2 reaction. It is important to note that, in the absence of base, the complexes 3a and 3b react with hydrogen in butanol or methanol/benzene to give, among other products, the dinuclear Ti2-H2 complexes la and 4, respectively. In contrast, the reactions with hydrogen of the complexes, Ru2Cl4(PR3)4 (R = Ph, p-tol), 2 (a, b), to give such dinuclear T|2-H2 complexes, require Proton Sponge®, a stronger base than DMA. This difference in reactivity toward H2 between the complexes of type 3, and the complexes of type 2, must be in part related to involvement of the isoPFA amine functionality in promoting reaction with H2. This isoPFA arnine involvement probably occurs in DMA as well, although isoPFA fragments were not identified in this system. Hydrogenation of 1-hexene catalyzed by RuCl2(PPh3)(isoPFA), 3b: The 31P{ lH) NMR spectrum (Fig. 4.31) of a hydrogenating system with 1:1 DMA/C0D6 as solvent, shows 3b to be the main ruthenium species present, with rninor resonances at 8 100.1,48.2,26.2,15, and possibly a weak, broad resonance at 8 ~63. The minor signals may be related to intermediates in the catalytic cycle, but have not been identified. The similarities in reactivity toward H2, between 3a and 3b, and the similarity 133 of [RU]T, [H2] and [hexene] dependences of the rates of 1-hexene hydrogenation for these complexes, suggest a mechanism similar to that proposed for 3a in equations 4.21 and 4.22. Again, free Ph3PO, but no free PPI13, is seen to be present (Figure 4.31). Hydrogenation of prochiral substrates catalyzed by RuCl2(PPh3)(PPFA), 3a, and RuCl2(PPh3)(isoPFA), 3b The catalytic hydrogenation experiments on prochiral substrates, using the complexes of type 3, were undertaken at elevated pressures, because the earlier studies25 had found the 3a/butanol system to be ineffective at 40 °C under 1 arm H2 for the hydrogenation of disubstituted olefins, including a-N-acetamidoacrylic acid. At 1000 psig H2, the systems summarized in Table 4.2 catalyze the hydrogenation of the procMral terminal olefins, itaconic acid, CH2=C(COOH)CH2COOH, a-N-acetamidoacrylic acid, CH2=C(COOH)NHCOMe, and also the internal olefin, (^-a-N-acemmidocinnamic acid, CHPh=C(COOH)NHCOMe. Although the number of substrates tried was limited, the lack of any chiral induction in olefinic acid systems is not encouraging. One obvious possible reason for lack of asymmetric induction is that the active species is not chiral. Experiments described earlier on the reactions of the complexes of type 3 with 2-10 atm H2 in the absence of substrates showed that redistribution of phosphine ligands occurs in butanol or benzene/methariol solutions, leading to the formation of dinuclear T|2-H2 complexes. The formation of (Ti2-H2)(PPh3)2Ru(u-Cl)2(u;-H)RuH(PPh3)2, la, for example, in the olefinic acid hydrogenation systems, could account, in whole or in part, for the lack of chiral induction. In section 4.1 and elsewhere,3 questions were raised concerning the selectivity of catalytic hydrogenation systems rontaining P-N chelating ligands, compared to P-P and N-N systems. The systems show a decided preference for hydrogenation of disubstituted olefinic acids over sterically similar ketones, which are, by definition, disubstituted. Because of the degree of phosphine ligand redistribution that can occur in the reactions of the complexes of type 3 with hydrogen in the absence of substrate, however, the evident selectivity of these systems for olefins over ketones must be interpreted with caution; it is possible that P-N ligands are not bonded to the catalytically active metal centre. Chapter 5 Reactions of RuCi2(PPh3)(isoPFA) and derivatives with methanol 5.1 Introduction The homogeneous dehydrogenation of alcohols by ruthenium, and other transition metal complexes, has been known for over twenty years.nu 12.113.114 indeed reviews2-4 have listed many solvents, including water, glycols, aldehydes, amides, carboxylic acids, ethers and amines, that have been shown to undergo dehydrogenation by reaction with transition metal complexes. Dehydrogenation of these hydrogen donors is often facilitated by the presence of unsaturated organic substrates, acting in the role of hydrogen acceptors; the above-mentioned reviews2-4 list many examples of such catalytic hydrogen transfer systems. The acceptors, or substrates, have been most often carbonyl-containing compounds, but alkenes, imines4 and alkynes115 have also been used. There are also many examples of decarbonylation of solvents by metal complexes, and this subject has been reviewed.116 Generally, in their reactions with alcohols, metal halides or halide complexes, often of platinum metals and containing also phosphine or arsine ligands, generate metal hydrido and hydrido-carbonyl complexes.3 These reactions are often base-assisted, with the base neutralizing the hydrohalic acid formed in the reaction. For example:117 [Ru2Cl3(PEt2Ph)6]a + CH3CH2OH K0H- [RuHCKCOXPE^Ph)^ (5.1) Examples of catalytic dehydrogenation of alcohols, to generate H2 and the corresponding ketone without the need for added base or hydrogen acceptor, are those catalyzed by rM(OCOR)2(CO)(PPh3)2], where M = Os, Ru; R = CF3, C2F5, C&s-113 Because the present work is mainly concerned with methanol reactions, we now turn to a brief review of developments in this more specific area Reactions of methanol with metal complexes include dehydrogenation to give hydrogen and formaldehyde (eq. 5.2) and/or related products, and decarbonylation, in which methanol is converted just into CO and hydrogen. CH3OH H2 + CH20 (5.2) These processes may involve heterogeneous or homogeneous systems, may or may not be base-assisted and/or photoassisted, and may be catalytic or stoichiometric. Before looking at details of the dehydrogenation reactions, it is useful to consider some products which are sometimes associated with the formaldehyde formed according to equation 5.2. A comprehensive review of the chemistry of formaldehyde has been compiled by Walker.118 Processes which lead to associated products are briefly: (i) dimethyl acetal formation: dry HQ CH20 + 2CH3OH H2C(OCH3)2 + H20 (5.3) (ii) methyl hemiacetal formation: H+ CH20 + CH3OH H2C(OH)(OCH3) (5.4) Under alkaline or neutral conditions, only the hemiacetal forms. (iii) polymerization: At room temperature, dry gaseous CH^ polymerizes slowly; however, trace water promotes rapid polymerization: CH20 + H20 :=r H2C(OH)2 HO(CH20)„H (5.5) methylene glycol paraformaldehyde (iv) thermal decomposition of paraformaldehyde: 2(CH20)n » HCOOCH3 + other products (5.6) methyl formate (v) the Cannizzaro reaction: 2CH20(aq) + H20 OH" • CH3OH + HCOOH (5.7) methanol formic acid which occurs via: 2CH20 HCOOCH3 (5.8) (vi) Condensation to give glycolic adldhyde: OH: 2CH20 ;=r CH2OHCHO (5.9) glycolic aldehyde In turn, process (vi) can lead to other products and even sugar-like derivatives; this process, however, is favoured more by the weaker alkaline earth bases, whereas the Cannizzaro reaction is favoured by a stronger base, such as KOH. Many of these processes listed under (i) - (vi) were considered in the present work during the search for organic products of the methanol dehydrogenation reactions. Heterogeneous dehydrogenation of methanol to give formaldehyde, by means of high temperature air-oxidation over a metal catalyst has been known for 120 years;119 production of methyl formate by passing methanol vapour over copper metal at 250 °C was reported in 1916.120 The air-oxidation of methanol, however, results in the formation of water: CH3OH + V2 02 J^j CH20 + H20 (5.10) and, as seen from equation 5.5, formaldehyde and water can react to form polymeric formaldehyde. These processes are similar to heterogeneously catalyzed dehydrogenation reactions of higher alcohols. The interests in the homogeneous, solution, dehydrogenation of methanol according to equation 5.2 are that relatively mild conditions can be used, and also the formaldehyde product can be water-free and therefore stable against polymerization. It was only recently, in 1985, that the first reports of homogeneous catalytic dehydrogenation of methanol appeared. At that time, there appeared reports of 138 photocatalytic dehydrogenation of methanol, initially to formaldehyde, using rhodium and palladium catalysts,115*121 and in the same year from the same group114 came the first reports of thermal dehydrogenation of methanol to formaldehyde using homogeneous catalysts, formed from [Ru2(OAc)4Cl] and the tertiary phosphines, PR3 (PR3 = PPI13, PEtPh2, PEt2Ph). The thermal dehydrogenations were carried out in anhydrous methanol at 66 °C. Addition of acetic acid enhanced the rate, giving at best initial turnover frequencies of about 1 h1 for the first 17 h. The initial organic product was formaldehyde and its dimethyl ace tai, but this was gradually converted into secondary products including methyl formate. Deactivation of the PPh3<ontaining catalyst was associated with formation of the carbonyl complex RuHCl(CO)2(PPh3)2. Maitlis' group122 reported the dehydrogenation of methanol, again in the absence of base or hydrogen acceptor, catalyzed by RuCl2(PPh3)3, but at a higher temperature than used by Saito's group.114 The main ruthenium species identified in this case122 at the end of the catalysis was the dinuclear cation, [Ru2(u,-H)(u,-Cl)2(CO)2(PPh3)4]+, with small -amounts of other species being tentatively identified as RuHCl(CO)(PPh3)3 and RuCl2(CO)2(PPh3)2. In contrast to the results from Saito's group,114 the predominant organic product was methyl formate. However, the apparent selectivity of Maitlis' group's system for methyl formate may be due more to the higher temperature (150 °C versus 66 °C) than to the difference in catalyst The most recent report123 of catalytic hydrogen production from methanol employed RuH2(N2)(PPh3)3, RuH2(PPh3)4, RuH2(H2)(PPh3)3 or K[RuH3(PPh3)3], at 150 °C, or [Rh(bipy)2]Cl at 120 °C. The reactions required the presence of NaOH base, and iUumination enhanced the rate. Without iUumination, turnover rates were 6 - 8 Ir1 for the first two hours. Surprisingly, the organic products were not reported, nor were the ruthenium products specifically, although Morton et a/.123 reported that the decarbonylation product, RuH2(CO)(PPh3)3, was observed after the catalysis, in the case of higher alcohols. Both this group and that of Halpern124 have reported the formation of the dihydrogen complex, RuH2(H2)(PPh3)3 via abstraction of hydrogen from higher alcohols by the [RuH3(PPh3)3]_ ion, and this may well occur in the case of methanol as well. Reports of transfer hydrogenation, using methanol as the hydrogen source, preceded the discovery of methanol dehydrogenation by many years. The earliest report125 of homogeneous methanol hydrogen transfer was in 1967, using platinum metal complexes as catalysts. There are a few other reports,126 most recently those of Smith and Maitlis.127'128 The dehydrogenation of methanol, in the absence of a hydrogen acceptor, is highly endothermic:127 CH3OH CH20 + H2 ^29SK = 92 ^Z™01 <5-10) Only the low solubility of hydrogen, leading its easy removal from solution, and the increase in entropy of the process, make the homogeneous dehydrogenation of methanol feasible. The participation of a hydrogen acceptor, however, makes the dehydrogenation considerably more favourable. The enthalpies for the process: RR'CHOH + cyclohexanone :===r RR'OO + cyclohexanol (5.11) for several alcohols have been calculated:128 alcohol AH°2 9 g K (kJ/mol) 2-propanol -25.9 ethanol -3.8 1-propanol +1.3 methanol +21 Thus, the participation of cyclohexanone as a hydrogen acceptor makes the dehydrogenation of methanol less endothermic by ~70 U/mol; yet, compared with these other alcohols, methanol is not a promising hydrogen source. The rates of transfer hydrogenation from methanol are also much lower than corresponding rates for higher alcohols.125-129 Far example,129 for the reduction of cycloheptene by various alcohols in toluene, catalyzed by RhH(PPh3)4, the relative rates were 2-propanol (200) > ethanol (8) > methanol (1). Similar trends are observed for dehydrogenation of alcohols in the absence of a hydrogen acceptor.123 While higher alcohols have thus received more attention as hydrogen sources, Smith and Maitlis127 have maintained that methanol still holds certain advantages over gaseous hydrogen and other alcohols, namely: (i) methanol is more easily handled than H2. (ii) the local concentration of reductant is several orders of magnitude higher for methanol than for H2. Thus, while the solubility of hydrogen gas in organic solvents is typically ~3 x lf>3 M atm*1 at 20 °C, methanol contains, potentially, 49.4 mol H2/L based on the reaction: CH3OH - CO + 2H2 (5.12) (iii) there are potentially 2 mol Hk/mol CH3OH, compared with 1 mol H2 for other alcohols, and (iv) methanol is relatively cheap (per mol H2), and is likely to remain so. Examples of decarbonylation of methanol are scant, but have been known for many years. In 1968, Wilkinson's group68 reported the reaction: RuC«PPh* \JS?L* R<*2<CO)(PPh,)3 (5.13) H2 or N2 This same dihydrido-carbonyl has been produced from several different reactions between methanol and ruthenium complexes (as discussed below), and was mentioned above as appearing in alcohol dehydrogenation systems catalyzed by several ruthenium hydrides (e.g., RuH2(N2)(PPh3)3 in NaOH/alcohol solution).123 Smith and Maitlis have related a series of complexes by the reaction scheme shown below for his Rua2(PPh3)3/methanol system in the absence of base: 141 RuQ^COXPPh^ - RuH^COXPPh^ RuH^jXPPhjk RuH2(PPh3)3 RuQ2(PPh3)3 - RuHCl(PPh3)3 RuHzCPPbii, RuH^jXP^ (5.14) The RuH2(CO)(PPh3)3 species is also formed by stirring RuH(C6H4PPh2)(PPh3)2(Et20) in methanol at room terrrperature for several hours,91 and also in the following reactions:90 RuH2(PPh3)4 + CH20(g) RuH2(CO)(PPh3)3 (5.15) RuHCl(PPh3)3 + NaOMe ^^"p' RuH2(CO)(PPh3)3 (5.16) Other examples of reactions generating RuH2(CO)(PPh3)3 are given elsewhere.130,131 The ubiquity of this dihydrido-carbonyl complex suggests it is a ruthenium sink in these alcohol reactions, as indicated by its irreversible formation in equation 5.14. Another common hydrido-carbonyl formed in methanol systems is RuHCl(CO)(PR3)3; for example, Saito's group114 reported that RuHCl(CO)(PEtPh2)3 was formed quantitatively from Ru(OAc)a(PEtPh2)3 solutions that had lost their catalytic activity for methanol dehydrogenation. As mentioned previously, the analogous RuHCl(CO)(PPh3)3 species, along with Rua2(CO)2(PPh3)2, was tentatively identified by Maitlis' group122 as being formed from RuChOPPhak m methanol at 150 °C, along with the dinuclear, canonic complex, Ru20i-H)(p-a)2(CO)20?Ph3)4+. The reaction of RuCl2(PPh3)3 with alcohols to give RuHCl(CO)(PPh3)3 via the intermediate RuHClOPPh^ has also been reported.132 Thorburn133 obtained a neutral, dinuclear hydrido-carbonyl by the action of Proton Sponge® and methanol on Ru2Cl4(dppb)2. He proposed the following stoichiometry: PS Ru2CU(dppb)2 + CH3OH » Ru2H(p-H)(p-a)2(CO)(dppb)2 + 2HC1 (5.17) 142 Mechanistic Studies Many papers have appeared on the subject of the mechanism of transfer hydrogenation, and this subject will now be reviewed briefly. The mechanism of transfer hydrogenation is still poorly understood; for example, publications by Beaupere et a/i34,i35,i36 nave featured mechanistic studies on catalytic hydrogen transfer to a,(3-unsaturated ketones using alcohols as die hydrogen source. Listed are 21 references, appearing between 1975 and 1983, which describe attempts to clarify the mechanism of transfer hydrogenation reactions using various acceptors, donors and catalysts. It is stated that134 "usually, the equations describing the postulated reaction schemes are oversimplified and result only in the determination of product-formation rate constants." Details of the various steps depend on the catalyst, the donor (D), and acceptor (A) involved, but generally two main routes for hydrogen transfer from D to A, to give D-H2 (i.e., D less 2 H atoms) and AH2, have been proposed: the donor (D) route and the acceptor (A) route,2 and are exemplified by a RhH(PPh3)4 system, studied by Beaupere et •al. The active species in the cyclic part of the mechanism, after initial phosphine dissociation and transfer of hydride atoms from two RhH(PPh3)4 molecules to a single A,137 is thought to be Rh(PPh3)3, and the D and A routes are outlined below: Rh(PPh3)3 Rh(PPh3)3D Rh(PPh3)3A Rh(PPh3)3DA Rh(PPh3)3 + D-H2 + AH2 143 The RhH(PPh3)4 catalyzed reduction of cyclohexenone, in the absence of base by 1-phenylethanol in toluene at 40 °C under argon, was sensitive to the mixing order.138 The rate was fastest when alcohol and catalyst were mixed first, followed by addition of the acceptor, cyclohexenone; slowest when acceptor and catalyst were mixed first, followed by addition of the donor alcohol; and in between when the catalyst was mixed with a donor-acceptor mixture. With cyclohexenone, the D route was considered to be the faster one. Also, whereas transfer of the pVhydrogen of the alcohol has usually been considered rate deterniining in alcohol hydrogen transfer systems,139,140 O-H bond cleavage was the rate-deterrnining step in the above system,137,141 and the hydroxyl H atom is transferred to the a-C atom of the ketone. This concerted H transfer, of course, requires the simultaneous coordination of both donor and acceptor. Spogliarich et al.142 later reported on the base-assisted transfer hydrogenation of cyclohexanone catalyzed by [Rh(COD)dppe]+, using 2-propanol as hydrogen source. The catalyst was refluxed in KOH/2-propanol solution, and then the substrate added—the same optimum order of addition found by Beaupere et al. In -contrast, Kvintovics et a/.143 found that if the ketone was not added first, Rh metal precipitated in a similar transfer hydrogenation from 2-propanol to acetophenone catalyzed by an in situ rhodium catalyst containing iV-acetyl-fSj-metWonine-f/^^^-sulfoxide, in the presence of KOH. This comparison illustrates the contrasting dependence of optimal order of addition of reactants on catalyst, donor and acceptor, in the Rh-phosphirie systems, the alcohol is presumed to form an active complex with the catalyst. In the rhodium-sulfoxide case, however, the interaction between basic 2-propanol and Rh complex gave reduction to metal. In contrast to the concerted D—"~A hydrogen transfer proposed for the RhH0?Ph3)4 system, hydrogen transfer via hydride formation from catalyst and alcohol has been inferred for other systems.131'132 In one system,144 preference for hydride formation versus concerted hydrogen transfer depended on the choice of hydrogen 144 acceptor. With hydrogen transfer from 1-phenylethanol to the a,fJ-unsaturated ketone, benzylideneacetophenone, the following steps were inferred: (i) phosphine dissociation, (ii) coordination of acceptor, (iii) coordination of the alcohol and formation of a metal alkoxide, (iv) a rate-determining hydrogen transfer from the alkoxide ligand to the coordinated ketone and (v) release of product via protonation. With dibenzylketone as acceptor, however, kinetic studies indicated that while steps i to iii were similar to those of the previous case, step iv appeared to involve hydrogen transfer from alkoxide ligand to the metal. The coordinated hydride was then thought to attack the coordinated ketone with the lower reduction potential. The above discussion has been limited to a description of general proposals and scattered results from studies on hydrogen transfer from higher alcohols. There are no examples of mechanistic studies on methanol hydrogen transfer. Smith and Maitlis have speculated on possible intermediates in the Rua2(PPh3>j/methanol hydrogen transfer system, but lacked data to distinguish between the possibilities.128 For the dehydrogenation of methanol catalyzed by Ru(OAc)Cl(PPh3)3, the proposed mechanism145 is analogous to that described for higher alcohols using M(OCORF)2(CO)(PPh3)2 (M = Ru, Os; OCORF = a perfluorocarboxylate):113 (i) solvolysis of the catalyst (alkoxide formation): MLL' + RR'CHOH ML(OCHRR') + HL' (ii) (Elimination from the alkoxide ligand: MUOCHRR') MHL + RR'CO (iii) acid attack on the hydride intermediate: MHL + HL' MLL' + H2 (5.20) where R, R' = alkyl; L = ligands; L' = carboxylate. The methanol case is depicted in Figure 5.1.145 As mentioned earlier, HOAc addition enhanced the dehydrogenation rate, while carbonyl formation was associated with catalyst deactivation. Addition of paraformaldehyde also deactivated die catalyst to the hydrido-carbonyl, by accelerating the process: Ru(OAc)Cl(PPh3)3 + CH2O • HOAc + RuHCl(CO)(PPh3)3 (5.21) It was therefore concluded that facile liberation and difficult coordination of CH2O was required to increase turnover frequency to H2 and CH2O, and, at the same time, avoid deactivation to the hydrido-carbonyl via decarbonylation of the formaldehyde (see Fig. 5.1). If incipient metal-carbon bond formation is allowed to occur, then CH2O would Ru = Rua(PPh3)3 Figure S. 1 Scheme proposed in reference 14S for the dehydrogenation of methanol catalyzed by Rua(OAcXPPh3)3. probably react via coordination of the CH2O through the carbon atom, followed by hydrogen transfers to the metal to give, finally, the hydrido-carbonyl and H2. htermediates for the process: RuHCl(CH20)(PPh3)3 H2 + RuHCl(CO)(PPh3)3 . (5.22) were not specifically suggested. Some MO calculations were carried out to determine the relative tendencies toward metal-carbon bond formation for the CH2O moieties coordinated to ruthenium in [RuHCl(CH20)(PH3)3] and for rhodium in rranj-[Rh(CH20)(PH3)2].145 As the Ru-O-C angle was decreased from 180°, initially repulsive M-C interactions gradually gave way to attractive M-C interactions. The Ru-C repulsive interactions were greater than the corresponding Rh-C repulsive interactions. Thus, theoretical calculations indicate that the rhodium complex would have the greater tendency toward carbonyl formation and concomitant loss of catalytic dehydrogenation activity. Furthermore, the calculations showed that prc(C>~—d:t(M) interactions were important in promoting the formation of the M-CO bond, and this was considered to be favoured for Rh(I) compared to Ru(H). These calculations were consistent with the generally observed higher methanol dehydrogenation activity for ruthenium(II) versus rhodium(I) complexes.1 K122,123.145 These studies on methanol dehydrogenation and decarbonylation are also of interest in connection with mechanistic work on the Fischer-Tropsch reaction: 2H2 + CO —A— MeOH (5.23) catalyst Germane to all these reactions, organometallic chemistry has provided structural precedents, which act as models for the proposed catalytic intermediates. Thus, examples of methanol-containing complexes of ruthenium have been characterized,901146 as have r|2-CH2O and formyl complexes of osmium,147 iron148 and more tentatively, ruthenium,90 and methoxo complexes of ruthenium149'150'151 as well as other alkoxo complexes of the iron triad.152'153 5.2 Experimental Experiments on the effects of various solvents, added PPh3t and air, respectively, on visible spectra of RuChfPPhsXisoPFA), 3b, in solution The visible spectrum was recorded for a solution made up of 3b (2.7 mg, 0.0033 mmol) in ~3 mL of CHCI3 under Ar. Added PPh3 (20 mg, 0.076 mmol) was then dissolved in this solution and the visible spectrum recorded again. The visible spectrum of an identical solution, but without added phosphine, was also recorded before and after exposure to air for a few minutes at ambient temperature. The visible spectrum of 3b in THF was recorded as follows: the complex 3b (17 mg, 0.020 mmol) was dissolved in freeze/thaw-degassed THF (1 mL) by heating to reflux temperature under nitrogen in the anaerobic cell shown in Figure 2.4, and the visible spectrum recorded within 0.5 h at ambient temperature. A similar experiment was carried out with the differences being that the solution was made up of 3b (6.9 mg, 0.0085 mmol) in THF (1.5 mL)/MeOH (0.5 mL). Experiment on the effect of styrene on 3b in solution Complex 3b (~30 mg, 0.035 mmol) and nitrogen-purged styrene (0.035 mmol) were dissolved under nitrogen in ~1 mL nitrogen-purged CDCI3 at ambient temperature in a 5 mm NMR tube, and the 31P{ lR] NMR spectra recorded. A similar experiment was carried out with a higher styrene:Ru ratio: 3b (~60 mg, 0.070 mmol), styrene (0.90 mmol). Reactions ofRuCl2(PPh3)(isoPFA), 3b, and the derivative, RuHCl(CO)(PPh3)(isoPOF), 5, with methanol in the absence of added base The preparation of the product, 5, from the reaction of 3b with methanol, was as described in section 2.4. The same product was obtained when the reaction was carried out with 0.005 mmol 3b in 2 mL methanol in a sealed tube under vacuum at 100 °C for 1.5 h in the presence of 0.4 mmol 1-hexene. The preparative reaction was also carried out in methanol-aU under otherwise the same conditions as described in section 2.4. The distillate from the methanol-cU preparation was tested for formaldehyde using the Eegriwe test118 and a sample submitted for GCMS analysis. A single crystal X-ray diffraction study of the isolated complex 5 from the methanol-do reaction was carried out by the crystallographic group at SFU, Burnaby, Canada.154 Catalytic activity of 5 and 3b: (i) isomerization of 1-hexene The complex 5 (0.004 g, 0.005 mmol) was heated overnight at 105 °C with 0.4 mmol 1-hexene in 2 mL dry methanol under vacuum; at the beginning of this experiment, liquids were freeze/thaw degassed, then vacuum transferred onto the catalyst contained in a glass tube, which was then sealed. The distillate obtained after reaction was analyzed by gas chromatography using a packed column, packed with tricresylphosphate on chromosorb. An almost identical procedure was used for the corresponding experiment with 3b and various substrates, with the specific substrates and reaction conditions given later in Table 5.3 within the Results section (sect 5.3). Distillates obtained after reaction were analyzed as indicated in the table. (ii) hydroformylarion of 1-hexene The complex 5 (0.012 g, 0.015 mmol) was heated at 100 °C for 2 d with 1-hexene (0.16 mmol) in 2 mL CH2CI2 under 1500 psig 1:1 H2/CO. At the beginning of this experiment, the liquid reagents were placed in a glass liner, which was then inserted into a steel bomb, and these liquids then purged with a flow of nitrogen gas prior to addition of the solid catalyst; the bomb was then sealed, flushed at least twice with H2/CO and filled with this gaseous mixture to the pressure indicated above. The distillate obtained after reaction was analyzed by gas chromatography using an OV-17 column for identification of alcohol and aldehyde products, and a tricresylphosphate column for identification of hexane and hexenes. Reactions ofRuCkiPPhsXisoPFA), 3b, and derivatives with methanol in the presence of added base With Proton Sponge® as base: A typical reaction at room temperature is described in section 2.4 under synthesis of 7. Single crystal X- ray diffraction studies on two of the products of this reaction were carried out; the structure of RuHCl(CO)(PPh3)(isoPFA), 6, was determined by S. Rettig of this department, and the structure of RuHCl(PPh3)(isoPFA), 7, by J. Simpson of the University of Otago, New Zealand55 Distillates of reaction mixtures obtained after reaction were analyzed by the Eegriwe test118 and by GCMS. The gas evolved during a similar reaction was monitored in a gas evolution experiment using the standard procedure and apparatus indicated in chapter 2. Some of complex 3b (0.037 g, 0.046 mmol) was shaken with Proton Sponge® (0.040 g, 0.19 mmol) in methanol (10 mL) under -800 mm Hg nitrogen at 30 °C for mare than 4 h, and the gas evolution measured at intervals. With KOH as base: A typical reaction of 3b with methanol in the presence of KOH, with KOH:Ru = 1.0, to give 6 as the main ruthenium species, is described in section 2.4. Some reactions employing a KOH:Ru mole ratio of >1:1 were carried out in a similar manner. Wine-coloured supematants obtained from such reactions were analyzed by column chromatography, in experiments such as follows: the complex 3b was heated at 80 °C under nitrogen in previously freeze/thaw-degassed KOH/methanol solution (3 mL of 1.8 M s 5 mmol KOH) till the green, insoluble 3b had completely reacted, and a wine-coloured solution was produced, along with a white precipitate, which was filtered off and shown to be RuH2(CO)(PPh3)3. The filtrate was vacuum-pumped and, as the solvent decreased, more white precipitate was deposited and this was filtered off to increase its yield. When the solvent volume reached about 0.5 mL, a red deposit was visible and so the remaining red supernatant, containing a high concentration of KOH, was decanted and discarded. The red solid was dissolved in methanol (~3 mL), and nitrogen-purged, deionized water (~50 mL) added to deposit the dissolved red material. This red deposit was washed with nitrogen-purged deionized water and dried under vacuum, before dissolving in a minimal amount of acetone prior to chromatography, initially in air, conditions were optimized using TLC (silica). Yellow band(s) could be eluted with acetone, but methanol was required to move red materials—a total of four red bands. The first eluted red methanol solution increased in intensity with increasing exposure of the initial acetone solution to air. Therefore, subsequently the column was prepared by removing air from the silica by vacuum-pumping, and thereafter keeping the silica under nitrogen. All solvents were purged with nitrogen and all column operations were conducted under nitrogen. The column was packed with silica and methanol (spectro-grade). An acetone solution of the red solid was transferred to this column; two yellow bands were first eluted with acetone, then three red bands with methanol, then two 'weak' yellow bands (which were discarded), then finally a fourth red band was collected. The solvent was removed from these fractions by evaporation under vacuum and spectroscopic data recorded for the CDCI3 extracts of the residues. On recrystallization of the white precipitate of RuH2(CO)(PPh3)3 from dichloTomethane/methanol, a few single crystals of the dichloromethane solvate of RuH2(CO)(PPh3)3 suitable for X-ray analysis were obtained on one occasion. Also, a series of similar reactions, but with styrene also present, were carried out to show variations in appearance and styrene-hydrogenating activity with increasing KOH concentration, and the procedure in detail was as follows: Some of complex 3b (0.006 g, 0.007 mmol), and a freeze/thaw degassed mixture of dry methanol, KOH/methanol solution (to give a total of 1 mL methanol) and styrene (0.087 mmol) were mixed under nitrogen in a Carius tube equipped with a septum for GC sampling. The tube, with the Teflon valve closed, was partially immersed in a thermostatted oil bath at about 80 °C; the clock was started, and the reaction mixture samples were injected at intervals directly into a Varian A-90 GC equipped with trap for dissolved and suspended solids, and a carbowax or SE-30 packed column. Experiments on the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/methanol, in THF in the presence or absence of styrene The KOH/methanol solution was made by dissolving KOH (7 g) in dry methanol (35 mL) under nitrogen and the suspended carbonate was filtered off. The KOH concentration of the filtrate was determined by titration of an aliquot (typically, 180 uL of KOH/methanol solution dissolved in 40 mL water) against potassium hydrogen phthalate (12-20 mg.) dissolved in water. The KOH/methanol solution thus prepared (~3 M) was stored under nitrogen and the titration repeated occasionally to monitor any change in concentration. Generally the concentration remained unchanged, at least over a couple of weeks, and even slight yellowing of the solution did not noticeably affect either the KOH concentration or the consistency of subsequent kinetic experiments. A JH NMR experiment was carried out to determine if there is any reaction between methanolic KOH and THF at ambient temperature, as follows: THF (3 mL), 0.1 mL of 3 M KOH/MeOH and 0.9 mL of MeOH were mixed at ambient temperature under air. The *H NMR spectrum of this mixture in C6D6 was recorded after 2 h, and again after 4 d. The title reactions were set up using one of the tubes shown in Figures 2.3 and S.2 connected to a double-manifold vacuum system. The reactions were conducted in 3:1 THF/methanoL the two components being separately prepared according to the following general procedure: a weighed amount of 3b was dissolved in section "B" of the specially designed tube (Fig. 2.3 or 5.2, Teflon valve closed) by refluxing, with a hot-air gun as heat source, under nitrogen in freeze/thaw-degassed THF (or in a THF/styrene mixture, when styrene was involved; the reactions involve dehydrogenation of methanol and styrene Figure 52 Glass tube used to monitor changes in the styrene/ethylbenzene ratio during transfer hydrogenation of styrene. Solutions are shown as they would appear before mixing. was used as an appropriate substrate, being converted to ethylbenzene). To the THF solution thus formed, phosphine (PPh3 or isoPFA) could be added, if necessary, through the Teflon valve opening. The tube was clamped at approximately the angle indicated (Figs. 2.3,5.2), the THF solution was then frozen in section "B" of the tube and the required volume of KOH/methanol solution was introduced through the Teflon valve opening into section "A" of the tube. Solid potassium chloride or methanolic lithium 2.5 mL KOH/ mothanoi solution RuCljfPPr^XisoPFA) and styrene in THF 7.5 mL of a solution of nitrate could also be introduced through the same opening, and also washed into section "A" with the volume of methanol required in order to bring the methanol/THF volume ratio up to 1:3. The methanol solution in section "A" was then degassed by means of three freeze/pumpAhaw-cycles. The freeze/thaw cycles were applied to both sections "A" and "B" together to minimize premature mixing of the two solutions. Finally, the tube, still at the angle indicated, was brought to the desired temperature under nitrogen by immersing in a water bath, and the Teflon valve closed. The tube was removed from the vacuum line, and the reaction(s) initiated by tilting the tube to mix the two solutions. Where styrene was used, it was important to mix the styrene with the solution of 3b before adding the KOH/methanol solution if maximum hydrogenation was desired. Mixing the solution of 3b with the KOH/methanol solution first, followed by addition of styrene, resulted in decreased hydrogenation rates and conversions. Monitoring changes in the 31P{1H} NMR spectrum: The reaction of 3b with methanolic KOH and styrene, in methanol/THF (1:3 v/v), was carried out as described above, initially utilizing the tube shown in Figure 5.2; however, as soon as all reagents were mixed, about 1 mL of the yellow, reaction solution was transferred by a nitrogen-flushed syringe to a 5 mm NMR tube fitted with a septum and previously filled with nitrogen. The NMR tube was placed immediately into the Varian XL300 probe and the 31P{!H} NMR spectra recorded at intervals. Monitoring changes in the visible spectrum: Changes in the visible spectrum during the reaction were monitored using the tube shown in Fig. 2.3. Typically complex 3b (3 mg, 0.004 mmol) was dissolved in THF, or THF/styrene (12 mL), thereby keeping the reactant concentration low to avoid subsequent clouding of the solution by the KCI that is produced in this reaction. For the same reason, the KOH concentration had to be £ 025 M (typically £ 1.5 mL of the 3 M KOH/methanol solution plus 2.5 mL methanol) or clouding of the solution made visible absorption measurements impossible. Even at these concentrations, a few tiny crystals of KCI did grow in the quartz cell during the reaction, but their presence did not affect the spectral changes; their refractive index appeared to be identical to that of the reaction solution, since the crystals were completely invisible when immersed in the solution. The reaction was monitored by setting the spectrophotometer at 605 nm, the absorption maximum of 3b, and starting the reaction (vide supra) and the chart drive simultaneously. When the two solutions were thoroughly mixed, the cell was placed in the thermos tatted cell holder in the spectrophotometer, usually within 10 to 15 seconds of mixing at time = 0. If spectra were to be recorded at various times during a run, the spectrophotometer could be zeroed before the run, with the above cell in place, but containing merely the 3:1 THF/methanol solvent mixture. This zero-correction was done for the data of Figure 5.25, but not for the data of Figure 5.26 (sect. 5.3). Monitoring changes in the styrene/ethylbenzene ratio: In order to use GC to monitor the hydrogenation of styrene in some of the reactions, it was necessary to quench the reaction at various points, to allow for distillation; "minimized evaporation from each sample withdrawn from the reaction mixture was also necessary, otherwise the samples gave erratic GC analyses. The reaction was carried out in the tube shown in Figure 5.2 as described above. Typically, 3b (25 mg, 0.03 mmol) in 7.5 mL THF/styrene, containing up to 1 mL styrene (with dissolved phosphine, where applicable), in section "B" of the tube, and the KOH/methanol solution (£ 0.8 mL of 3 M KOH/ methanol solution with dry methanol added to raise the volume to 2.5 mL) in section "A", were brought to the desired temperature in a mermostatted bath, the reaction started by tipping the tube to mix the two solutions, and the timer started. Samples (~0.6 mL) of the reaction solution were withdrawn, by means of a syringe, through the septum with the Teflon valve closed, or through the opened Teflon valve which was subsequently closed, and immediately expelled into a small heap of solid boric acid (-1.5 g) in a small flask; the mixture was then swirled to quench the reaction. The flask was immediately stoppered, or sealed with Parafilm, and 155 chilled to minimi 7/» evaporation. The volatile fraction of the reaction mixture was vacuum transferred to another small flask and then stored in a stoppered vial until the sample could be analyzed by GC. Monitoring gas evolution: Depending on conditions, reactions of 3b in basic methanol can lead to gas evolution (H2), and this was monitored as described elsewhere19*20 with a flask and procedure that was modified as follows: a) For experiments requiring a small volume (< 100 pL) of KOH/methanol solution: First, the desired volume of THF (~6 mL) was introduced into the round part of the flask ("A", Fig. 5.3a.) Then a glass bucket containing the desired amount of 3b was placed on a hook attached to a ground glass stopper, and inserted into the lower arm ("B") of the two-arm flask (Fig. 5.3a). An empty bucket was then placed on another hook attached to a ground glass stopper, this being inserted into the upper arm ("C") of the flask. The flask was then attached to the gas uptake/evolution apparatus by a narrow, spiral glass tube ("E", Fig. 5.3a), as described elsewhere.20 The THF in the flask (part "A") was degassed by three freeze/pump/thaw cycles, and the flask then filled with nitrogen (~1 atm), and isolated from the main part of the apparatus by closing the stopcock on the narrow glass tube ("E") which connects the flask to the rest of the apparatus. The bucket containing the complex 3b was then dropped into the THF, and portion "A" of the flask was heated, almost to reflux temperature, to dissolve the complex. When the flask was cooled to room temperature, the stopcock on the narrow tube ("E") was opened and the rest of the apparatus adjusted to allow free flow of nitrogen into the flask. The upper hook-and-stopper was then partly withdrawn from arm "C", and the required amount of freeze/thaw-degassed KOH/MeOH solution was injected by syringe into the upper bucket; the hook-and-stopper, still holding the bucket containing KOH/MeOH solution, was then a) b) Figure S.3 a) A 25 mL, modified gas evolution flask for following reactions between RuCl2(PPh3)(isoPFA), 3b, in THF, and a small volume of KOHAnethanol solution, b) Similar apparatus, for use with larger volumes of KOH/methanol solution. reinserted into arm "C". The reaction was initiated by dropping the upper bucket, with its contents, into the 3b/THF solution. The gas evolution was then followed in the usual manner. b) For experiments requiring a larger volume (1-2 mL) of KOH/methanol solution: As in (a), the desired volume of THF, and the complex 3b, were introduced into the flask (Fig. 5.3b) at "A" and "B", respectively. A small, closed, bent tube ("F"), fitted with a ground-glass joint, and containing the desired volume of KOH/methanol solution, was then inserted into the upper arm ("C") of the flask. Next, the THF in "A", and the KOH/methanol solution in "F", were simultaneously freeze/thaw-degassed. After being degassed, the flask was filled with nitrogen (~1 atm) as in (a). The bucket containing 3b was then dropped into the THF, and the complex dissolved by heating as in (a). The procedure then followed closely that described in (a), the reaction being initiated by rotating the bent tube ("F") containing the KOH/methanol solution. Procedure (a) was used for monitoring the gas evolution (H2) for experiments ih which the KOH:Ru ratio was ~1. Gas evolution was monitored in one such experiment at 27 °C, in the absence of styrene, in THF (6 mL)/MeOH (2 mL) with 3b (68 mg, 0.084 mmol) and KOH (0.096 mmol). After gas evolution had ceased, ruthenium complexes were isolated from this solution as described below, and the NMR spectra of the associated residue in CDCI3 recorded. Procedure (a) was also used in a related experiment at 27 °C, where the KOH:Ru mol/mol ratio was 0.84, and styrene was present (styrene:Ru =11). Procedure (b) was used to monitor gas evolution in an experiment at 27 °C with high base and styrene concentrations (3b (28 mg, 0.035 mmol), KOH (3.8 mmol), styrene (8.7 mmol)) in THF (6.5 mL)/MeOH (2.5 mL); total volume: 10 mL. After gas evolution had ceased (time = -40 min), analysis for ethylbenzene production was carried out by injecting the reaction solution directly into a GC with Carbowax column and a solute trap. A formaldehyde (ktermination was carried out using the Hantzsch reaction and is described in a later section. These methanol/THF reaction solutions, which had contained initially -50 mg 3b in 8-10 mL solution, were worked up by mixing them with -80 mL nitrogen-purged, deionized water. In the absence of styrene, precipitation of mtheruum-containing products occurred, and these were then filtered, dried under vacuum and analyzed in CDCI3 by 31P{ lH) and lH NMR spectroscopy. Precipitation did not occur when this procedure was used for reaction solutions containing styrene; instead, separation of -0.2 mL of a styrene phase less dense than the aqueous phase occurred, which was adsorbed on Celite® by passing the two phases through -2 cm (-10 mL) Celite® in a Schlenk filter apparatus. Solvent was then removed from the adsorbed phase under vacuum, and the adsorbed residue extracted with CDCI3 for analysis by 31P{!H} and *H NMR spectroscopy. Detection of KCI produced in the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/MeOH: i) in THF: The reaction was carried out in the general manner described above for monitoring changes in the visible spectrum, but in the tube shown in Figure 2.3. Conditions used were: [3b] = 0.019 M, [KOH] = 0.45 M, THF (1.0 mL), MeOH (0.33 mL), under nitrogen at ambient temperature. The reagents were mixed and the reaction monitored by recording the decay of the 605 nm band of 3b by visible spectrophotometry. When a white precipitate, which had formed during the reaction, had settled (time = 72 min), the visible spectrum of the solution was recorded. The supernatant was decanted and the precipitate washed with acetone and dissolved in deionized water, addition of AgN03 solution resulted in a white precipitate, whose solubility in dilute HNO3 was tested. ii) inCH2Cl2: The reaction was carried out as follows: 1 mL of a 5 x 10*3 M solution of 3b in degassed CH2CI2 was mixed at ambient temperature under nitrogen with 2 mL of a 1.8 M KOH/MeOH solution. A white precipitate, which had precipitated within a minute after mixing, was isolated by decanting the yellow supernatant which had also formed in the first minute, recrystallized from water/acetone, dried under vacuum and submitted for elemental analysis. Anal. Calcd. for KCI: Cl, 47.6. Found: 45.8. An aqueous solution of the white solid (16 mg/100 mL) was analyzed for potassium by atomic emission at 766 nm, and the reading was compared with readings obtained for 10 and 20 mg/100 mL KCI solution standards. Solvent was removed from the decanted yellow supernatant under vacuum, and the NMR spectra recorded for the associated residue in CDCI3. Formaldehyde detection: i) in the distillate from a reaction of 3b with KOH/MeOH in THF: The reaction of 3b with KOH/MeOH was carried out as follows: some of complex 3b (30 mg, 0.037 mmol) was mixed with styrene (1.8 mmol), 0.5 mL of a 1.8 M KOH/MeOH solution and 1.3 mL THF under nitrogen in a Carius tube, and the Teflon valve closed. The tube was heated for a couple of minutes at ~70 °C to dissolve the 3b complex, and then the temperature maintained at 50 °C for 5 d. The volatile components were then isolated from the reaction mixture by vacuum transfer, the Hantzsch reaction117 was carried out on this distillate, and the visible absorbance of the product solution at 412 nm compared with that of appropriate standards. ii) of non-volatile CH2O in a reaction mixture formed from a reaction of 3b with KOH/MeOH in THF: The reaction mixture formed in the gas evolution experiment described above, with high base and styrene concentrations, was analyzed for formaldehyde content as follows: the reaction mixture was mixed with deionized water to give ~40 mL solution which was passed through Celite®; a portion (17 mL) of this aqueous solution, after several days, was filtered and neutralized with aqueous HQ. The Hantzsch reaction was carried out on this solution, and the visible absorbance (412 nm) of the yellow product solution compared with that of appropriate standards. Experiment to detect gas phase CO by GC: Some of complex 3b (26 mg, 0.032 mmol) was mixed under nitrogen with styrene (0.88 mmol) and a solution of KOH in methanol (2 mL, 1.8 M), in THF (5 mL) in a Carius tube equipped with a septum, and the Teflon valve closed. The complex was dissolved by hearing briefly at 100 °C and then the tube was maintained at 45 °C for 25 min. A sample of the gas phase above the cooled solution was syringed onto a molecular sieves GC column (TCD detector), and the gas chromatogram compared to that for a CO gas standard. 2D NMR experiments on distillates from reactions of 3b with KOH/MeOH-dt and MeOH-di: These experiments were carried out to confirm that methanol was the true hydrogen source in the title reactions. The procedure was as follows: some of complex 3b (14 mg, 0.017 mmol) was mixed under nitrogen with THF (0.71 mL), styrene (0.88 mmol) and KOH/MeOH-d4 or MeOH-di (0.25 mL, 1.5 M) in the same apparatus as for the 'CO detection experiment'. Reaction was allowed to continue for 40 h at ambient temperature, and then the entire 1 mL reaction mixture was injected, in portions of 0.1 - 0.2 mL, into a Varian A-90 GC instrument equipped with a solute trap and a TCD detector. The ethylbenzene fraction was collected in a trap cooled with liquid nitrogen, and mixed with 1 mL acetone, before recording the 2D NMR (46 MHz) spectrum The extents of hydrogenation of styrene in the MeOH-cU and MeOH-di reactions were determined at the same time. For reference purposes, the *D NMR spectra of a 1:1 styrene/acetone mixture and a 1:1 ethylbenzene/acetone mixture were also recorded. An experiment to determine what would happen to CH20 generated in a solution similar to those used for the reactions of 3b with KOH/MeOH Formalin solution (-50 mL of 37% CH2O in methanol water) was mixed with KOH/MeOH solution (~50 mL, 1.5 M) and THF (-30 mL) at ambient temperature, and then the mixture neutralized with HCUq; acetone was then added until a considerable amount of white solid precipitated. About 0.5 g of this was collected, washed with water, acetone and diethylether and dried. The white solid was submitted for elemental analysis and mass spectrometrie analysis. Reactions ofRuHCl(CO)(PPh3)(isoPFA), 6, with methanol The tide complex, 6, (0.030 g, 0.037 mmol) was heated with methanol (1 mL) under nitrogen in a Carius tube at 80 - 85 °C overnight in the absence of base. Most of the yellow 6 remained undissolved. The yellow solid present after reaction was filtered off, dried under vacuum and the !H NMR (300 MHz, CDCI3, ambient temperature, under nitrogen) of the dried solid was recorded. Some of complex 6 (0.009 g, 0.011 mmol) was heated with KOH (0.01 mmol) in methanol (~1 mL) under nitrogen in a Carius tube partially immersed in a 80 °C oil bath overnight The supernatant decanted from the reaction mixture was then analyzed by UV-vis spectrophotometry, and the remaining solids dried under vacuum and analyzed in CDCI3 solution by *H NMR spectroscopy. Finally, complex 6 (18 mg, 0.022 mmol) and styrene (0.88 mmol) were heated under nitrogen in freeze/thaw-degassed methanol (1.0 mL) at -80 °C for 3 d and the solution analyzed by GC to determine the amount of hydrogenation of styrene to give ethylbenzene. 5.3 Results Figure 5.4 shows the visible spectrum of 3b (Ama* = 455, 605 nm) under nitrogen in CHCI3 in the absence (spectrum 'a') and presence (spectrum 'b') of added PPI13; the solution of 3b is air-sensitive fFig- 5.5). The visible spectrum of 3b in THF or a 3:1 v/v THF/MeOH solution shows: Anuu 605 nm, e = 500 M^cnr1; = 455 nm, e=1200 M-icm-i. The 31P{!H} NMR spectrum of 3b was unaffected by the presence of up to a 10-fold excess of styrene. 163 Reactions ofRuCl2(PPh3)(isoPFA), 3b, and RuHCl(CO)(PPh3)(isoPOF), 5, with methanol in the absence of base The distillate from the methanol-dt preparation of 5 in the absence of hexene was analyzed using the Eegriwe test and 2 - 5% of a stoichiometric amount of formaldehyde was detected. Most gas chromatogram peaks obtained from the GCMS analysis of the same distillate could not be identified; however, one fraction, that gave a highest mass fragment with m/z = 64, indicates of the order of 30% of a stoichiometric amount of methyl formate-di. The formulation of 5 based on spectroscopic and analytical data was confirmed by the crystal structure; the amine functionality of the isoPFA ligand had been replaced by a methoxy group during the reaction of 3b with methanol at reflux temperature under nitrogen. The crystal structure of 5 is shown in Figure 5.6, and selected bond distances and angles are given in Tables 5.1 and 5.2. Disorder of the unsubstituted cyclopentadienyl ring was satisfactorily accounted for by including two orientations of this group in the refinement All hydrogens were found in Fourier difference maps except for those in the disordered rings. The final residuals were R = 0.033 and Rw = 0.037. Other crystal data for the structure will be published elsewhere.154 The spectroscopic and analytical data for 5 as prepared from methanol-do are given in section 2.4, and the lU NMR spectrum is shown in Figure 5.7. The !H NMR spectrum CG4 C18 CIS Figure 5.6 The X-ray structure of RuHa(CO)<PPh3XisoPOF), 5. Table 5,1 Selected bond distances (A) obtained from X-ray diffraction data for three ruthenium products that have been isolated from the reactions of RuCl2(PPh3)(isoPFA), 3b, with methanol Bond Ph2 Ph2 Ph, typea 6b 5C 7d Ru-Cl 2.441(2) 2.521(2) 2.441(1) Ru-H 1.53(6) 1.57 1.52 Ru-P(Pri)2 2.329(2) 2.358(2) 2.178(1) Ru-PPh3 2.383(2) 2.354(2) 2.250(1) Ru-NMe2 2.420(6) - 2.265(5) Ru-OMe - 2.262(4) -Ru-(CO) 1.820(8) 1.793(8) -C-O(carbonyl) 1.142(9) 1.161(8) -O-Me - 1.439(7) -N-Me 1.469(9) - 1.488(7) N-Me 1.49(1) - 1.483(7) aCorresponding distances are listed side by side for comparison. bRuHCl(CO)(PPh3)(isoPOF) (Fig. 5.6): six-cc>ordinate, approximately octahedral. cRuHCl(CO)(PPh3)(isoPFA) (Fig. 5.9,5.10): six-coordinate, approximately octahedral. dRuHCl(PPh3)(isoPFA) (Fig. 5.12): five-coordinate, approximately trigonal bipyramidal. of the sample prepared using methanol-cU is shown in Figure 5.8. The OMe signal at 5 2.38 of the 'proton^ontaining* sample (Fig. 5.7) is not present in the spectrum shown in Figure 5.8, and the underlying PC//CH3 broadened multiplet at 5 2.36 (integrating at 1 proton) is clearly revealed. The hydride signal at 8 -13.9, however, is still present in the spectrum shown in Figure 5.8, but integrates to about half a proton. The orfno-phenyl protons meanwhile are reduced from 6 protons to about 1, giving an ortho/metajjara ratio of 1:9 instead of 2:3 as in the non-deuteriated sample. Table 5,2 Crystallographically determined bond angles associated with the bond distances shown in Table 5.1 /N Angles (°)B 6b 5C 7d P!-Ru-P2 158.67(6) Pj-Rui-P2 177.55(6) NrRurP2 157.1(1) N-Ru-Pi 92.8(1) OrRui-Pj 88.8(1) NpRui-P, 97.9(1) N-Ru-P2 108.3(1) OrRurP2 89.5(1) PrRurP2 101.5(1) C»-Ru-Pi 94.1(2) Ci-Rui-P, 92.9(2) C39-RU-N 96.6(3) Q-Rih-Oz 170.1(2) C39-RU-P2 86.3(2) CrRurP2 88.6(2) RU-C39-O 176.1(6) RUj^!-^ 176.4(6) C39-Ru-Hj 89(2) d-Rui-H, 82.5 C39-R11-CI 177.7(2) Ci-Rui-Qj 101.4(2) PrRu-H, 68(2) P^Ruj-H, 88.9 N-Ru-H, 161(2) 87.7 Cl-Ru-Hj 89(2) arRui-Hi 175.9 P2-Ru-H! 90(2) PrRurHi 89.4 N-Ru-Cl 85.3(1) QTRUI-QI 88.4(1) Nj-Ru^! 86.8(1) PrRu-Cl 87.11(6) Pi-Rui-Cli 89.61(6) PrRu!-Cli 124.9(1) p2-Ru-a 91.76(6) PrR^-Cli 92.03(6) PrRui-Cli 91.9(0) K^rresponding angles are listed side by side for comparison. Atom numberings are those given by the crystallographers. bRuHCl(CO)(PPh3)(isoPFA) (Fig. 5.6). cRuHCl(CO)(PPh3)(isoPOF) (Fig. 5.9, 5.10). dRuHCl(PPh3)(isoPFA) (Fig. 5.12). 8 7 6543 210 Figure 5.7 The lH NMR spectrum (300 MHz, CDC13, ambient temperature, under nitrogen) of complex 5. Peaks for complex 5: 'o', ortho phenyl H atoms; 'm, p', meta and para phenyl H atoms;.other peak assignments for 5 are given in sect 2.4. Other peak assignments are: 'a', CHC13; 'b', silicone grease; peaks marked 'c' are due to unidentified impurities. , i , ; ,— r- 1 ! i i '—. • ' • i 1 i i -4-8 7 6543 21 0 Figure 5.8 The !H NMR spectrum (300 MHz, CDC13, ambient temperature, under nitrogen) of deuteriated complex 5, prepared using methanol-d4. Peaks for complex 5: 'o', ortho phenyl H atoms; 'm, p', meta and para phenyl H atoms;.other peak assignments for 5 are given in sect. 2.4. Other peak assignments are: 'a', CHC13; 'b', silicone grease; peaks marked 'c' are due to unidentified impurities. In the experiment on isomerization of 1-hexene catalyzed by 5, the gas chromatograph of the distillate of the reaction mixture after reaction showed -95% isomerization, mainly to cis and *ra/ts-2-hexene in a 1:4 ratio, with -2% hydrogenation, presumably by methanol hydrogen transfer, to give hexane. In the exr>eriment on the hydroformylation of 1-hexene catalyzed by 5, the reaction mixture after reaction had a strong smell mat was certainly typical of either n-heptanol or n-heptaldehyde or a mixture of these. Comparison of the gas chromatogram, obtained using a tricresylphosphate column, of the distillate of this reaction mixture with standards showed that most of the 1-hexene was either isomerized or converted to other products. Comparison of the OV-17 gas chromatogram of this distillate with standards indicated the presence of heptaldehyde, heptanol and other minor products. A quantitative estimate of the product distribution, assuming all products of 1-hexene conversion were detected on the columns used, and an estimated 90% conversion of 1-hexene to products, was: trans-2-hexene (0.03 mmol), cw-2-hexene (0.01 mmol), heptaldehyde (0.03 mmol), heptanol (0.03 mmol) and other minor products. The results of the experiments on transfer hydrogenation catalyzed by 3b are summarized in Table 5.3. It is seen that 3b shows significant catalytic activity at 105 °C for the transfer of hydrogenation from methanol to ketones or activated olefins, while cyclohexene and hexene were reduced in only trace amounts under corresponding conditions. The hydrogen transfer to styrene was accelerated by the presence of KOH, but these data will be presented later. Table 5.3 Catalytic hydrogen transfer from methanol, catalyzed by RuCl2(PPh3)(isoPFA), 3b, in neat methanol at 105 °C in the absence of base £Ru] (M x 103) time(h) substrate (M) product(s)a 2.5 3.0 overnight 88 1-hexene (0.20) acetpphenone (0.21) hexaneb-c (3) a-methylbenzyl alcohol0 (26) 3.0 2.4 3.2 88 20 97 cyclohexene (0.24) styrene (0.22) vinylmethylketone cyclohexane0 (3) ethylbenzenec (4) 2-butanolc«d (70) 2-butanonee (50) (0.30) aIn parentheses is the total number of turnovers = moles hydrogenated substrate per mol bAccompanied by 93% isomerization, which corresponds to a turnover of 74. Estimated from GC data. Estimated from *H NMR data. ^us, the turnover number for the hydrogenation of the vinyl group was 120. Reactions of RuC h(PP'hs)(isoPFA), 3b, with methanol in the presence of added base With Proton Sponge® as added base: The spectroscopic and analytical data far the main mmenium-containing products, RuHCl(CO)(PPh3)(isoPFA), 6, and RuHCl(PPh3)(isoPFA), 7, of a typical reaction at ambient temperature are described in section 2.4. The solid state structure of 6 (Figs. 5.9,5.10) contains approximately trans disposed phosphorus atoms, with the hydride located trans to the nitrogen atom of the isoPFA ligand, and the CO trans to Q. Selected bond distances and angles are tabulated in Tables 5.1 and 5.2, respectively, along with parameters for RuHCl(CO)(PPh3)(isoPOF), 5, and 7 (to be discussed below). The lH NMR spectrum (300 MHz, CDC13, under Ru. 171 nitrogen) of 6 is shown in Fig 5.1 la (ambient temperature) and Fig. 5.1 lb (20 0 to -50 °C); detailed assignments are given in section 2.4. The -NMe2 resonances show a temperature dependence (Fig. 5.1 lb), the single peak, 8 2.3 at ambient temperature, being resolved into two singlets at SO °C. Besides the hydride quartet at 8 -15.3 due to 6, an additional hydride at 8 -15.8 (t) is always present in very minor concentration (<10%) and increases with exposure to air. The 31P{ 1H} NMR (121 MHz, ambient temperature, under nitrogen) data for 6 are given in section 2.4, and spectra including resonances due to this complex, prepared using KOH as base, are shown later in Figs. 5.13 and 5.14. These spectra will be discussed in connection with the KOH reaction. A very weak infrared peak at v™1-1852 cm-1 on the shoulder of the carbonyl peak at 1897 cm-1 is always seen and is possibly due to the presence of the same oxidation product detected by the weak hydride triplet in the !H NMR at 8 -15.8. The yellow complex 6 exhibits a weak band in the a) I < • • • I • • • ' , • ' • . | . . . I | . I ! I | • ' I | I I I , | I I • I | I I I I I I I I I | I I . 1 | I I ' • | I I I I | I I I I | I I II | II II | II I I | II II | I I • I | 9!7«S43210 PPM Figure 5.11 a) The 'H NMR (300 MHz, CDCI3, ambient temperature, under nitrogen) spectrum of RuHCl(CO)(PPh3)(isoPFA), 6; peaks for 6 are assigned in section 2.4. Other peak assignments are as follows: 'a', hydride oxidation product impurity which always accompanies 6; T>', CHC13; 'c', complex 5; 'd\ unidentified impurities, b) Variable temperature *H NMR spectra (300 MHz, CDC13, under nitrogen) for 6. The peaks marked with an **' are due to the NMe2 protons; the two NMej resonances are resolved at tow temperature, but coalesce at room temperature. Figure 5.1 lb (see legend on previous page) visible region at ~450 nm. The crystal structure of RuHa(PPh3)(isoPFA), 7, shows that the molecule is five-coordinate and has a structure that is best described as a distorted trigonal bipyramid (Fig. 5.12). Structural parameters are shown in Tables 5.1,5.2, and 5.4 in a manner designed to facilitate comparison of corresponding features of similar complexes. Fortunately, the hydride was located, and a Ru-H bond distance of 1.52 A was obtained. The spectroscopic and analytical data for this complex are given in section 2.4. In addition to the rothenium-containing products described above, other unidentified crystals were isolated on one occasion from a reaction mixture where the Ru:PS ratio was 1:1; the solution of these crystals in CDCI3 revealed a !H NMR triplet at 8 -4.8 with J = 21 Hz. Also, on another occasion, where a 1:1 mole ratio of Ru:PS at a higher temperature of 72 °C, a !H NMR triplet at 8 -13.9 for the isolated solids dissolved in Figure 5.12 The X-ray structure of RuHa(PPh3)(isoPFA), 7. CDCI3 indicated the formation of some 5, in addition to the main product 6 showing a doublet-of-doublets at 8 -15.3 (see sect 2.4). In attempts to identify gaseous and volatile products, in an experiment similar to that described in section 2.4 for the preparation of 7,3b (0.044 g, 0.057 mmol) was allowed to react with dry methanol (4 mL) in the presence of Proton Sponge® at ambient temperature. Reaction occurred slowly and was complete only after four days, at which time there was no rernaining solid starting complex, 3b, which is only sparingly soluble in methanol. An Eegriwe, chronotropic acid test on the distillate from this reaction mixture revealed 0.76 mg (0.026 mmol) of CH2O. Methyl formate could not be detected in the distillate using GC. The gas evolution experiment, using the gas uptake/evolution apparatus, showed that in the reaction of 3b with methanol at 30 °C in the presence of Proton Sponge®, the maximum gas evolution, reached after 3.7 h, corresponded to ~1 mol gas/mol 3b. With KOH as base: A typical reaction of 3b with dry methanol with KOH:Ru = 1.0, as described in section 2.4, yields mainly the complex 6. Attempts to purify 6 by recrystallization from dlchloromethane/methanol were unsuccessful and a satisfactory elemental analysis was never obtained. However, the merely filtered and washed precipitate of 6, which deposited during the 1:1 KOH:Ru reaction, appeared reasonably pure by !H and 31P{1H) NMR spectroscopy. Some of complex 6 was also formed when this KOH reaction was carried out in the presence of styrene. The spectroscopic data far 6 have already been presented (above and in sect 2.4), but more explanation is required for the 31P{1H) NMR spectrum. Figure 5.13 shows the spectrum of a CDCI3 solution which has stood in a NMR tube under nitrogen for a week before the spectrum was recorded (the impurity at 8 29.2 is PlvjPO). The interesting feature is the presence of minor peaks surrounding the two main Figure 5.13 The ''P^H) NMR spectrum (121 MHz, CDC13, ambient temperature, under nitrogen) of 6 after -1 wk; member peaks of the 'two main doublets' mentioned in the text are marked with an '*'. A peak due to PhsPO is marked with an'a'. doublets centred at 8 50.0 and 31.9, with 2Jpp = ~290 Hz; these may be due to geometric isomers or diasteriomers of 6 with similarly trans disposed phosphorus atoms. The results of experiments in neat methanol to show variations in appearance and styrene-hydrogenating activity of solutions of 3b with increasing KOH concentration are summarized in Table 5.4. The colour changes observed depended both on time and KOH concentration, and a light white-yellow precipitate was always deposited. Table 5.4 also conveys some approximate rate data. In all cases, an initial, relatively rapid hydrogenation activity slowed markedly after about the first half-hour, which was approximately the time required for most of the insoluble, green 3b complex to react No inorganic products were characterized from the yellow solutions obtained at higher [KOH] in the presence or absence of styrene, except for KCI (evidenced by an ill-chosen crystal structure determination!!). The wine-coloured supernatants, in appearance the same in the presence or absence of styrene, all exhibited a strong UV-vis band at 542 nm. Figure 5.14 shows the 31P{ lH) NMR spectrum of the supernatant of a KOH/methanol reaction mixture with KOHRu = 2.0, but in other respects similar to that Table 5.4 Colour changes of reaction mixtures* and turnover numbers for hydrogen transfer from methanol to styrene to give ethylbenzene as a function of time and initial [KOH] using 3b as catalyst total turnoversb (with solution colour given in parentheses) [KOH] (M) time: 0.5 h 10 h 20 h 30 h 2d 1 wk 0.2 2.6 5.0 (pink-cherry red) (wine) 6.2 (wine) (red) 0.4 2.6 5.7 (pink-cherry red) (wine) 7.2 (wine) 9.2 (pink) 0.7 (pink-cherry red) 7.7 (wine) 8.8 (wine) 0.9 3.5 (pink-cherry red) 8.7 (wine) 9.1 (yellow) 1.4 3.9 (reddish orange) (pale orange) 1.8 3.4 (yellow orange) •Conditions: 80 °C; neat methanol (1 mL); under nitrogen in a closed Carius tube; RuCl2(PPh3)(isoPFA), 3b, (~6 mg, 0.007 mmol); initial styrene (0.087 mmol). Total turnovers = total number moles ethylbenzene produced per mole Ru. described in section 2.2 for the preparation of 6. The pairs of doublets centred at 8 50.7 and 33.5, and at 8 50.5 and 33.3, both with 2JPP = -290 Hz, if expanded, clearly indicate two equally dominant isomers of 6 along with other products present -4 h after starting the reaction. Singlets at 8 -4.8 and -5.8 are probably due to free isoPFA, also Figure 5.14 The 31P{1H) NMR spectrum (121 MHz, CH3OH, ambient temperature, under nitrogen, unlocked) of the supernatant of a reaction mixture formed by heating RuCl2(PPh3)(isoPFA), 3b, (58 mg, 0.072 mmol) in dry methanol (2 mL) in the presence of KOH (0.1 mL of a 1.8 M solution) at -80 °C for 3 - 4 h under nitrogen. Peak assignments are: 'a', complex 6; V, complex X, formulated in section 5.4; 'c', PPh3; 'd', isoPFA; 'e', unidentified species. isolated by chromatography in a similar reaction (described later, p. 181), and free PPh3, respectively. (The 31P{1H} NMR resonance (109 MHz, CD2CI2) for free isoPFA occurs at 8 -3.3 ; the PPh3 resonance, often used as a 31P NMR standard, occurs at 8 ~-5.6.) Peaks in the range 8 61 to 67 are probably associated with the hydride resonances exhibited by the residue obtained by removal of solvent from this supernatant. The *H NMR (300 MHz, CDCI3) spectrum of the soluble components of this residue is shown in Figure 5.15. The doublet-of-doublets at 8 -15.3 and -15.9 correspond to 6 and an oxidation product, respectively (vide supra); the other hydride resonances in the range 8 -17.5 to -19 were always observed in *H NMR spectra of the wine-coloured supernatants. A visible spectrum of this same CDCI3 solution is shown in Figure 5.16; the wine-coloured solutions always exhibited the intense band shown at 542 nm. The !H NMR (300 MHz, CDCI3) spectrum of the methanol-washed white precipitate from the same reaction (Fig. 5.17) showed the distinctive hydride pattern of RuH2(CO)(PPh3)3: 8 -6.8 (tdd, 2JPH = 16, 30 Hz, 2JHH = 8Hz), -8.8 (dtd, 2JPH = 31,78 Hz, 2JHH = 8 Hz), very similar to the ——i—:—-—• 1 1—i—.—;—i—i—. 1 1 1 1—.—i—i— -5 -10 -15 -20 -2$ P"" Figure 5.15 The *H NMR spectrum (300 MHz, CDC13, ambient temperature, under nitrogen) of the residue obtained from the solution used for the 3IP{1H) NMR spectrum given in Figure 5.14; upfield peak assignments are as follows: 'a', complex 6; V, a hydride oxidation product always accompanying 6; 'c' (a pair of doublets-of-doublets centred at 8 -18.1), complex X, formulated in section 5.4; other upfield resonances are unidentified. 400 500 600 700 800 900 Wavelength (nm) Figure 5.16 The visible spectrum of the CDC13 solution used for the !H NMR spectrum given in Figure 5.15. 180 I—i—I—I—|—I—i—i—I—|—i—i—I—I—|—I—I—I—i—|—I—I—I—I—|—i—I—I—i—|—I—I—I—i—|—r -6.5 -7.0 -7.5 -8.0 -8.5 -9.0 -9.5 PPM Figure 5.17 The upfield JH NMR spectrum (300 MHz, CDC13, ambient temperature, under nitrogen) of RuH2(COXPPh3)3. literature data.68 An X-ray diffraction study on the single crystals of RuH2(CO)0?Ph3)3-CH2Cl2 obtained on recrystallization from dichloromethane/methanol confirmed the structure proposed based on spectroscopic data.90*91 The structure28 is shown in Figure 5.18, and selected bond angles land distances are listed in Table 5.5. The hydrides were not located; CH2CI2 was found to be present in the crystal as a monosolvate. In the chromatography of the wine-coloured supernatants of reaction mixtures with KOH:Ru » 1:1, litres of solvent were required to elute each of the last three bands, and dissolved silica, which is very slightly soluble in methanol, was a contaminant in these fractions. The small sample sizes obtained on removal of methanol gave poor quality NMR spectra because of the low concentration obtained when the residues were dissolved in the NMR solvent However, the *H NMR spectrum of the first fraction (in CDCI3) showed a multiplet at 8 -17.6; the second yellow fraction was a mixture, the lH NMR spectrum exhibiting several hydride signals that were not in the spectrum of Figure 5.15, but 6 was Figure 5.18 The X-ray structure of RuH2(COXPPh3)3. identified by the resonance at Scocb -15.3 (dd). The !H NMR spectrum of the residue from the first red band showed two hydride resonances, the more intense being at 5 -18.7 (dd, J = 36,18 Hz). The second red residue, X, exhibited spectral features (Fig. 5.19: 8 -17.9 (dd, 1H, 2JpH = 36, 20 Hz), -18.2 (dd, 1H, 2JPH = 36, 20 Hz)) which were very similar to certain features of the spectra shown in Figure 5.15 (8 -17.95 (dd, J = 36,20 Hz), -18.25 (dd, J = 36,20 Hz)). Also shown in Fig. 5.19, in the range 8 1.0 - 5.2, are resonances typical of a coordinated isoPFA ligand, perhaps of two types present in equal 182 Table 5,5 Selected molecular structure parameters for RuH2(CO)(PPh.3)3a Bond type OC .Pi H •: >3 H Bond name distance (A) Ru-PPh3 Ru-PPh3 Ru-PPh3 Ru-CO carbonyl C-0 Ru-P2 Ru-P3 Ru-Pi Ru-Ci C1-O1 2.321(7) 2.301(7) 2.419(7) 1.97(3) 1.09(4) Angles (°) P2-Ru-P3 Pi-Ru-P2 P1-RU-P3 CrRu-P2 Ci-Ru-Pi C1-RU-P3 Ru-Ci-Oi 148.23(25) 101.01(23) 103.33(23) 103.3(8) 91.0(7) 96.4(8) 174.9(25) •Parameters arranged in order to compare with corresponding distances for RuHCl(CO)(PPh3)(isoPFA) in Table 5.1. Hydride positions were not crystallographically located; their positions in the diagram are based on spectroscopic data.68 amounts. These resonances are tentatively assigned as follows: 8 5.21 (q, 1H, 3JHH = 7 Hz, FcCJ/Me), 4.80 (q, 1H, 3JHH = 7 HZ, FcCWMe); 4.48, 4.44, 4.39, 4.35, 4.30 (16H, for two C5//3); 423, 4.17 (2s, 10H, for two C5//5); 2.87, 2.76 (2s, 6H, N(C/73)2); 2.28, 2.06 (2s, 6H, N(C//3)2); 1.7 - 1.0 (m, >30H, two P(CH(a73)2)2 ™& two FcCHC/73, 183 Figure 5.19 The 'H NMR spectrum (270 MHz, CDCI3, ambient temperature, under nitrogen) of the residue obtained from the eluted second red band, containing complex X, obtained on chromatography of a worked-up supernatant from a RuG2(PPh3)(isoPFA), 3b/KOH/methanol reaction mixture. Detailed assignments for X are given in the text; other assignments are: 'a', unidentified phenyl-containing species; V, CHC13; 'c', unidentified species. and probably some 'impurity' signals). Other signals around 8 0 (~25H), 4.5 (~6H) and 7 - 8 (~8H), are not assigned. The same solution exhibited a 31P{ lU} NMR (121 MHz) spectrum dominated by two broad singlets of equal intensity at 8 60.7 and 66.1, as well as two minor peaks of insufficient area (5 -10% of the total) for the presence of a PPh3 ligand in the dominant hydride complex present; the low intensity under the !H NMR peaks between 8 7 and 8 supports this. The infrared spectrum in KBr showed two weak peaks (v (Ru-H) 2055,2027 cm1) and one very strong peak (v (CO) 1932), tentatively assigned as indicated; the strong peak had a weak shoulder at 1910 cm1. The third red residue gave *H NMR signals in the range 8 0.8 to 5, possibly due to a coordinated isoPFA ligand; no clear hydride signal was observed. The IR spectrum v™1 showed weak peaks at 2055,2022 and a strong peak at 1940 cm-1 (with a weak shoulder at 1983 cm-1). The fourth red residue yielded a 31P{1H} NMR singlet at SCDCIS (121 MHz) 60.2, and strong IR peaks at vrar 1957 (with a shoulder at 1921) and 1730 cm1. No hydride signal was seen in the !H NMR spectrum In another experiment, in which a wine coloured reaction mixture was generated, but on four times the scale of the previous reaction, a yellow band was eluted in sufficient intensity to be identified as free isoPFA ligand by the *H NMR (300 MHz, CDCI3) spectrum: 5 0.85 - 1.05 (m, PCHC//3); 1.2 (m, PCHC//3) 1.38 (dd, J = 7,14 Hz, POYCH3), 2.18 (m, POYCH3), 3.98 (q (?), FcC//CH3), 4.05 (s, C5//5); 4.10, 4.22, 4.27 (3s, C5//3); the data corresponded roughly to those in the literature.8 Reactions ofRuHCl(CO)(PPh3)(isoPFA)t 6, with methanol in the absence arid presence of base The *H NMR spec turn of the yellow solid mixture obtained in the absence of base showed upfield hydride resonances at 8 -13.9 (t) and -15.3 (dd), indicating the presence of 5 and 6, respectively, in a ratio of ~ 1:9. In the reaction in the presence of KOH, a solution colour change to red and decolouration of the undissolved solid in the tube could be observed during the first hour. .The visible spectrum of the final red solution exhibited an intense 542 nm band, typical of the wine coloured solutions. The *H NMR (300 MHz, CDCI3) spectrum of solids isolated from the red solution by decanting indicated that at least ~50% of the 6 had reacted. No 5 was formed in this case, but RuH2(CO)(PPh3)3 was a major product, as indicated by its distinctive hydride resonances described above. There was also a minor hydride resonance at 8 -17.93 (dd, J = 18,35 Hz), which resembles one observed in Figure 5.15 for products obtained by Ae reaction of 3b with KOH/methanol. The reaction of 6 with methanol and styrene in the absence of base for 5 d at 80 °C resulted in ~3% hydrogenation of the styrene present, as indicated by GC analysis. The ethylbenzene produced is therefore 0.026 mmol, compared to the 0.027 mmol 6 present Experiments on the reaction of RuCl2(PPh3)(isoPFA), 3b, with KOH/methanol, in THF in the presence or absence of styrene In an experiment to determine the reactivity of THF with methanolic KOH, a KOH/MeOH/THF/C6f>6 solution, after 2 h of standing at ambient temperature, showed *H NMR signals due only to QjDe (8 723), THF (8 3.62 (t, 1H), 1.48 (quintet, 1H)) and MeOH (8 3.25 (s, 3H), 1.64 (s, 1H)). (For a MeOH/CeDe solution: C6D6(8 7.23), MeOH (3.19 (d, 3H), 2.03 (q, 1H)), and adding KOH to this solution gave rise to a spectrum showing MeOH singlets with 8 values identical to those given above; for a THF/CfiDe solution: C^)6 (8 7.23), THF (8 3.64 (t, 1H), 1.49 (quintet, 1H).) Figure 5.20 shows the 31P{1H) NMR spectra acquired during various time intervals for the reaction mixture formed from 3b, KOH, and styrene, using the method described in section 5.2 for methanol/THF reactions, at ambient temperature and other conditions shown in the figure legend. Spectrum i (Fig. 5.20) reveals some unreacted 3b (8 95.2 (d), 36.0 (d)), as well as a relatively strong peak due to PPI13 (8 -5.6). Two pairs of AB doublets are seen at 8 52.01 (d), 42.43 (d) (both with 2JPP = 318 Hz), and 8 51.88 (d), 42.22 (d) (both with 2JPP = 318 Hz), and these intensify with increased reaction time (ii - iv, Fig. 5.20) although the spectrum does not change very much after 1 h. The PPh3 resonance decreases in intensity with time relative to that in spectrum i, and this decrease is roughly paralleled by a decrease in a resonance at 8 69; decreases in the PPI13 and 8 69 resonances roughly parallel the increase in the AB pattern. Also, a weak resonance at 8 ~58.5 persists for about the first 45 min and then decreases, while weak resonances at 8 56,60.2,64.2, and 66 persist throughout the experiment; after ~2 h a weak resonance at 8 ~29.5 due to PlvjPO appears. Figure 5.21 shows the effect on the spectrum, after it had become tirne-independent, of lowering the temperature. The AB pattern shifts downfield by almost 1 ppm, on lowering the temperature from ambient to -18 °C (spectra i - iii, Fig. 5.21), and then back upfield on warming back to about 20 °C (spectrum iv). Figure 5.21 also shows that, as the temperature is lowered, the integration ratio of the AB pattern to that i)30s to 18 min f OC BO 60 40 20 0 I I -20 ii) 18 min to 29 min i | JI 1 1 1 iii) 29 min b i to 44 min I \oO 80 60 40 ' ' 200' I -20 iv) (with expansion) 46 min to 66 min ' i" " i"1 t i"1 11111111111111111' •' i' "' i' •" i11" '1 54 52 50 4» 46 44 42 PPU " '!1 11111111111 j 111.111 II 111 II 140 120 too so I'11•111111'I I 1 1 1 1 I I III 11111111111" 11111111'1' 'I' (0 -40 20 0 -20 -40 PPU Figure 5.20 The 31P{1H) NMR spectrum (121 MHz, CH3OH/THF (1:3), ambient temperature, under nitrogen, unlocked) of the yellow reaction solution formed from mixing a RuCl^Pt^XisoPFA), 3b/styrene/THF solution with a KOH/methanol solution at ambient temperature under nitrogen. Initial conditions on mixing: [Rua2(PPh3)(isoPFA)] (0.0095 M), [KOH] (0.075 M), [styrene] (0.43 M). Peak assignments are: 'a', complex 3b; V, complex 6', discussed later in the text; 'c', PPh3; other peaks are unidentified. Data acquisition periods are given at the left 187 60 40 I 1 • 1 1 ' • • • • i • • • 20 0 iii) -18 °c S2 1 I 11 1 1 I 1 1 I ' 48 46 •Wn*«f»(in»'^ii«i ^iimiytiaiiii.niMWwWW'M^f ' 11111111 I 11 140 120 I ' i i i I i i i i | i i i i | i i i i | 80 60 40 T 20 M M I I | I | I I I I | I I I I | I I il I I 0 -20 -40 PPU Figure 5.21 Variable temperature spectra for the same solution described for Figure 5.20, recorded after the spectrum had become time-independent. See Figure 520 for peak assignments. Expanded plots of the V AB quartet are shown above each spectrum. 188 iv) 20 °c ' • 1 | M | i i i i | ; i i i i | i i I i | i III | i i i I | i i i i | i i i i | i i i i I i i i ' | ' ' ' ' i i i ' ' | ' i ' ' I 11 1 1 I 1 1 11 i 1 '•1 1 140 110 100 80 bO ,40 20 0 -20 -40 PPU Figure 5.21 continued. of the PPh3 resonance clearly increases, and then decreases again on warming back to 20 °C. In another ambient temperature reaction for 2 d, when the solution had turned bright red, the 31P{1H} NMR spectrum (Fig. 5.22) was dominated by peaks at 8 66, 62.4, 61, and the PPI13 peak at 8 -5.6. The peaks in the range 8 61 to 66 are similar to those seen in Figure 5.14 for a purely methanolic system in the absence of styrene. Weak resonances at 8 61 (obscured?) and 47 may be due to RuH2(CO)(PPh3)3 (see above). In none of these reactions in THF/MeOH were resonances due to 6, per se, observed. When these solutions, with styrene either absent or present, and with KOH:Ru = 1.0, were worked up (sect 5.2), 6 was found to be present, as evidenced by *H and 31P{1H} NMR spectroscopy. It should be noted that the presence of ~1 equivalent of residual styrene after work-up did not affect the 31P{ lH] NMR resonances due to 6. Sometimes decomposition of 6 occurred because of exposure to air, the styrene solution of this complex being more air-sensitive than the solid precipitated sample. A colour change to greenish from the yellow colour of 6 in the worked-up residue accompanied aerobic a a Figure 522 The "PpH) NMR spectrum (121 MHz, CHjOHATHF (1:3), ambient temperature, under nitrogen) spectrum of the red reaction solution formed, after 2 d, in the same way as that described for Figure 5.20, but with the following concentrations: RuCl2(PPh3)(isoPFA), 3b (0.01 M), KOH (0.09 M) and styrene (0.86 M). Peak assignments are: 'a', complex X, formulated in section 5.4; V, complex 6', described later, in section 5.4; 'c', isoPFA; 'd', PPh3; 'e', RuH2(CO)(PPh3)3; other peaks are unidentified. decomposition, along with growth of a *H NMR hydride resonance at 8 -15.8 (t, 2JPH = 24 Hz). In the H2 evolution experiment, described in section 5.2, at 27 °C in the absence of styrene in THF (6 mL)/MeOH(2 mL) with KOH:Ru » 1 (KOH, 0.096 mmol and 3b, 0.084 mmol), the total gas evolved was 0.075 mmoL Ruthenium complexes were isolated as described in section 5.2 for such THF/MeOH solutions and spectra of this isolated material showed that about 0.064 mmol 3b had reacted to give mainly 6 which showed characteristic 31PpH} NMR resonances at 8 49.8 (d), 31.7 (d), (2jpP = 290 Hz) and a XH NMR upfield hydride resonance at 8 -15.3 (dd); other weak 31P resonances observed at 8 98.5 (d), 61 (d), (2JPP - 40 Hz) are assigned to the product RuHCl(PPh3)(isoPFA), 7, and other weak resonances at 8 51.0 (d), 31.7 (d), (2Jpp = 290 Hz), which increased on exposure to air, are assigned to an aerobic decomposition product of 6. A 31P singlet at 5 29.1 is attributed to PI13PO, while another weak resonance at 8 71 is unassigned. Of importance, the gas production corresponds to roughly 1 mol H2 per mol Ru, the evolution being a first-order process (Figs. 5.23, 5.24), with kobs = 0.008 s1. g_ , 1 • • 6- • 1 "3 2-l af - • op • 1 • 1 • I 1 I 0 200 400 600 800 time (s) Figure 5.23 Gas evolution plot for a reaction between a solution of 3b in THF and a KOH/MeOH solution: Reaction conditions: 3b (0.0105 M), KOH (0.012 M), THF (6 mL), MeOH (2 mL), 27 °C, under nitrogen. time (s) Figure 5.24 Straight-line semi-log plot for Figure 523 indicative of first-order gas evolution, based on total mol gas evolved. In a related experiment, where the KOH/Ru ratio was 0.84, and styrene was present (styrene:Ru =11), there was no gas evolution. In an experiment at 27 °C with high base and styrene concentrations ([3b]o = 0.0035 M, [KOH]o = 0-38 M, [styrene]o = 0.87 M, in THF (6.5 mL)/MeOH (2.5 mL), total volume: 10 mL), gas evolution was initially rapid and had virtually ceased at 40 min to give a total of -0.12 mmol gas, i.e., 3 mol gas per mol Ru. Gas chromatography indicated the formation of ethylbenzene (0.007 M) at this time. Formaldehyde (0.0027 M) was determined in this reaction solution using the Hantzsen reaction (sect 5.2); this concentration corresponds to roughly 1 mol CH2O per mol Ru. Monitoring the THF/MeOH reaction by visible spectrophotometry: The changes in the visible spectrum during a reaction of 3b with basic methanol in THF are shown in Figure 5.25. The visible bands of 3b decrease rapidly (a - d, t1/2 = 400 s), with an accompanying change in the solution colour from green to yellow. A slower secondary colour change from yellow to red then occurs, with the growth of an intense band at 542 nm (d - g, \m = -14 h). Figure 5.26 shows the changes in visible spectrum during the secondary stage (intermediate [KOH], yellow - red) of a reaction under the same conditions as those shown for Figure 5.25, but with styrene absent Similar general features are observed in both the presence and absence of styrene, although in the latter there are differences in 400 500 600 700 Wavelength (nm) 800 900 Figure 5.25 The changes in the visible spectrum during a 20 °C reaction of 3b with basic methanol in THF under nitrogen: [3b]0 = 4.6 x lO3 M, [KOH]0 = 0.0041 M, [styrene]0 = 0.44 M; THF (15.2 mL), MeOH (4.0 mL). Spectra are labelled as follows: 'a', recorded at 3 min after starting the reaction, shows mainly the presence of 3b, with 455,605 nm; V, at 8 min; 'c\ at 20 min; 'd', recorded at 1.5 h, shows mainly the presence of 6', with = -450 nm; 'g', recorded at -4 d, shows mainly the presence of red species with = 542 nm. absorbance in the region around 450 nm. Thus, in the first stage in the absence of styrene, the 3b bands at 455 and 605 nm decrease rapidly, as the solution turns yellow. Then, in the second stage, shown in Figure 5.26, the 6' absorbance at 450 nm decreases slightly at first, even though at this stage, nothing remains of the 3b 605 nm band; the absorbance at 450 nm then gradually increases along with the red species band at 542 nm The first reaction, loss of 3b to give 6\ was monitored by following the loss of absorbance at 605 nm, the Amu of 3b. Typical continuous plots of the change in absorbance with time are shown in Figures 5.27 ([KOH] = 0.066 M) and 5.28 ([KOH] = 0.21 M); the derived plots of log(At -A») vs. t, where At = absorbance at time t and A« is the final absorbance, are linear (Figs. 529 and 5.30), indicating that the reaction is first-400 500 600 700 800 Wavelength (nm) Figure 5.26 Changes in the visible spectrum during the second stage (intermediate [KOH], yellow "-red) of a reaction between 3b and methanol in THF in the presence of KOH Conditions were the same as for the reaction described in Figure 5.25, but with styrene absent The spectra are: 'a', at 1 h 50 min, showing mainly a band at -450 nm due to complex 6', discussed in section 5.4; "b', at -3 h, showing a decrease at -450 nm and the slight growth of the band at 541 nm due to the red species (including complex X, discussed in sect 5.4); 'c', at 6.5 h; 'd', at 14 h; 'e', at 3.8 d, showing the continued growth of bands at -400 and 542 nm. The entire reaction was monitored at 605 nm, and based on an estimated A. at about 8 d, the spectrum 'e' represents -80 % completion of the reaction. Figure 5.27 Plot of the change in absorbance (605 nm) with time during a reaction of 3b with KOH/MeOH in THF at intermediate [KOH] (reaction 8, Table 5.6). Figure 5.28 Plot of the change in absorbance (605 nm) with time during a reaction of 3b with KOH/MeOH in THF at high [KOH] (reaction 10, Table 5.6). -i 1 • r 200 400 600 time(s) 800 1000 Figure 5.29 Plot of log (A, - A.) vs time (t) for the data of Figure 527, showing behaviour that is first-order in [3b]. 0 100 200 300 400 time(s) Figure 5.30 Plot of tog (A, - A«) vs time (t) for the data of Figure 528, showing behaviour that is first-order in [3b]. order in 3b. Similar plots at other [KOH] values were also linear over at least two half-lives, and the addition of styrene had little influence on the reaction rates. The kinetic data are summarized in Tables 5.6 and 5.7. The variations in kobs with [KOH] and with [styrene] are shown in Figures 5.31 and 5.32, respectively. The reaction was tested for a possible salt effect, in the absence of styrene, by carrying out two further reactions at low [KOH] (0.001 M), but with added UNO3 at concentrations that corresponded to two of the higher [KOH] values used to obtain the data of Figure 5.31, namely, 0.056 M and 0.18 M (reactions 13 and 14, respectively, Table 5.7). The data are also shown in Figure 5.33, in a standard salt-effect plot, where ki=o is taken as the value of kobs at [KOH] = 0.001 M. There is only a slight positive salt effect (corresponding to a rate increase by a factor of ~ 1.3 over the range 0.00 - 0.18 M KOH M) and thus the shape of Figure 5.31 is largely due to a specific kinetic KOH-dependence. Table 5,6 Some kinetic data for the reaction" of RuCl2(PPh3)(isoPFA), 3b, with KOH/MeOH in the absence of styrene Reaction [3b] [KOH] tl/2 kobS (MxlO4) (M) (s) (s-1 x 103) 1 3.2 0.0017 570 1.21 2 4.7 0.0040 540 1.28 3 3.4 0.0048 520 1.33 4 5.1 0.0075 510 1.35 5 1.8 0.017 420 1.64 6 2.5 0.034 430 1.60 7 2.0 0.051 400 1.73 8 2.6 0.066 350 1.97 9 2.4 0.10 260 2.60 10 1.8 0.21 120 5.8 11 3.5 0.23 100 6.9 12 2.2 0.25 80 8.6 13 2.0 0.0010b 510 1.35 14 2.8 0.0008c 430 1.60 15 3.6 0.0035d 520 1.32 16 3.4 0.0048e 520 1.31 17 2.6 0.0048f 520 1.33 18 2.6 0.0048f 500 1.38 19 1.0 -0.036 380 1.8h 20 0.8 -0.03 260 2.7h 8In THF at 21 °C under nitrogen. b[LiN03] = 0.056 M. c[LiNO3]=0.18M. dAdded [isoPFA] = 8.0 x 104 M. eAdded [PPh3] = 19.8 x 104 M fAdded [PPh3] = 6.9 x 1(H M. sAdded KCI (0.4 mg, 0.006 mmol). •There were thennostatting problems; however, the temperature drift was estimated to be -1 °C/250 s, and therefore the difference between at least the initial rates of these reactions is considered significant. Table 5.7 Some kinetic data for the reaction8 of Rua2(PPh3)(isoPFA), 3b, with KOH/MeOH showing dependence on [styrene] Reaction [3b] (Mx 104) [styrene] (M) [KOH] <M) tl/2 (S) (s-1 x 103) obs 1 2 3 4 5 6 7 8 9 3.0 2.6 3.1 4.7 3.8 3.0 4.0 2.2 1.8 0.088 0.26 0.55 0 0 0.44 0.22 0.44 0.66 0 0.060 0.060 0.060 0.060 0.0040 0.0035 0.0035 0.0035 0.25 0.25 360 350 310 320 540 525 520 500 80 80 1.90b 1.97 2.23 2.16 1.28 1.31 1.33 1.38 8.6 8.6 aIn THF at 21 °C under nitrogen. bInterpolated from Figure 5.31. Salt effect corrections for the data of Figure 5.31 were afforded by the line drawn in Figure 5.33. After these corrections were applied, a curve of the form shown in equation 5.24 was fitted to the [KOH] dependence data within experimental error and Figure 5.34 shows the plot of kobs. corrected for salt effect, vs. [KOH]2. The rates are essentially independent of the styrene (added as a hydrogen acceptor), Figure 5.32. The rates were also unaffected by addition of either isoPFA or PPI13 (Table 5.6). Experiments on the effect of added KCI were carried out, although solubility is limited. The amount used (0.44 mg KCI dissolved in 16 mL of ~0.03 M KOH/(MeOH/THF) solution, i.e., 0.0008 M), though not enough to cloud the solution, exceeded the solubility limit; as mentioned above, KQ precipitated out during the MeOH/THF reactions even in the absence of added KCI. The data are summarized in Table kobs = 1.3 x lO"3 + 0.082[KOH]2 (5.24) 0.010 0.008 -0.006 0.004 -0.002 - • • Q 0.000 0.0 —T~ 0.1 I 0.2 0.3 [KOH](M) Figure 5.31 Dependence of k^, from visible spectrophotometric measurements, on [KOH] for the reaction of 3b with KOH/MeOH in THF (Table 5.6). 0.0025 0.0000 D[KOH] 0.060 M • [KOH] 0.004 M 0.2 0.4 0.6 Istyrene] Figure 532 Plot of k^, from spectrophoionjetric measurements, versus [styrene] at two different [KOH] values (Table 5.7). 0.12 0.0 0.1 0.2 0.3 0.4 0.5 I1/2(M1/2) Figure 5.33 Conventional salt effect plot, showing the effect of LiN03 on the for the first step of the reaction of 3b with KOH/MeOH in THF (1^ values are for reactions 13,14, Table 5.6; kM is taken as the value of kob, for reaction 1, Table 5.6). 201 0.001 -o.ooo -j 1 1 1 1 • 1 1 0.00 0.02 0.04 0.06 0.08 [KOHpCM2) Figure 5.34 Plot of k^, from spectrophotometric measurements, and corrected for salt effect, versus [KOH]2, for the reaction of 3b with KOH/MeOH in THF. 5.6. The difference in at least initial rates in the absence and presence of added KCI is significant, and the data are tentatively taken as indicating that added KCI suppresses the reaction rate. In one of the kinetic experiments, the spectrophotometer light beam was blocked off for approximately half of one half-life after the reaction had commenced; there was no disruption in the exponential profile of the reaction, showing that light is, at least, not required to maintain reactivity. 202 Gas chromatographic analysis of styrene/ethylbenzene concentrations for these 3b/KOH/MeOH reactions was carried out in the presence of styrene as a hydrogen acceptor. An example of a plot of {[styreneMstyreneJo} x 100% vs. time is shown in Figure 5.35 (reaction 8, Table 5.8) and the corresponding semi-log plot is shown in Figure 5.36. The observed first-order process occurs with Kbs= 2.23 x 103 s4. The decrease in hydrogenation activity of the system parallels well the disappearance of RuCl2(PPh3)(isoPFA), 3b, as seen by comparison with the data of a similar reaction at almost the same [KOH], monitored spectrophotometrically, where kobs = 1.97 x 103 s1 (reaction 8, Table 5.6). Of note, the total amount of styrene hydrogenated, at the point where hydrogenation activity has virtually ceased, is 0.0032 M (-2%), equal to the amount 100 -f 8 X EI o EJ B B B B B 98 0 1000 2000 3000 time(s) 4000 5000 6000 Figure 5.35 Plot of the percentage of styrene, {[styrene]y[styrene]0} x 100%, present at time (t), during reaction 8, Table 5.8. The value of [styreneJo was taken as the sum of [styrene]t and [ethylbenzene];, the concentrations at time (t), which were obtained from the GC printout 1000 2000 3000 time (s) Figure 5.36 Plot of log F vs. time (t), derived from the data of Figure 5.35; „ .[styrene], [styrene]., , [styrene]t F = {rr r - 7~ T~J x 100%, where TZ r = the fracbon of the original styrene remaining at l[styrene]0 [styrene]0 [styrene]0 ^ J e rstvTcnd time (t), and = the final fraction of the original styrene remaining. of 3b used. Thus, the hydrogenation of styrene is stoichiometric, with one mole of styrene hydrogenated per mole of 3b. If the initial [KOH] is increased above -0.060 M, a secondary hydrogenation process increases in importance and this is readily seen in the hydrogenation curves. Figures S.37 and 5.38 show curves for reactions 6 and 7 (Table 5.8), respectively. In Figure 5.37, the rate decreases exponentially to about 2000 s, and then a second hydrogenation phase is evident until hydrogenation ceases after about 8000 s. This interpretation, of course, depends heavily on the positions of the two points at 1000 and 2000 s, but two other experiments, using the same concentrations and temperature, also exhibited the same 'inflection' in the hydrogenation curve after about 10 half-lives for the rate of disappearance of 3b (t1/2 = 90 s, from spectrophotometric 204 Table 5.8 Summary of kinetic data from the GC experiments used to follow styrene hydrogenation during the reaction* of 3b with KOH/MeOH in the presence of styreneb>c Reaction [3b] (MxlO3) [styrene] (M) [KOH] (M) ti* 00 k'obi1 (s-1 x 103) 1 6.6 0.088 0.030 450 1.53 2 6.7 0.088 0.06 460 1.50 3 6.6 0.088 0.12 250 2.76 4 4.5 0.088 0.24 140 4.93 5 6.5 0.088 0.24 125 5.52 6 6.4 0.088 0.24c 165 4.18 7 6.1 0.088 0.30 100 6.9 8 3.2 0.18 0.060 310 2.23 9 3.0 0.35 0.060 210 3.29 10 6.6 0.35 0.060 230 3.00 11 3.6* 0.33 0.060 180 3.83 12 3.6f.s 0.33 0.060 210 3.29 13 3.6f-h 0.33 0.060 170 4.06 •In THF at 21 °C under nitrogen. b3b recrystallized from THF; samples recrystallized from CH2CI2 gave variable hydrogenation rates. cko'bs is defined in the text, and refers to the primary hydrogenation process. aLtss frequent sampling (e.g., taking the first sample after -10 min) consistently gave ethylbenzene concentrations -20 - 30% higher than those detected using sampling every -2 min. •Technique in this run improved over that for other 0.24 M experiments. Qb recrystallized from CH2CI2. «Added [PPh3] = 1.4 x l(r2 M. hAdded [isoPFA] = 1.1 x l(r2 M. 100 98 • 5? 96' 5 94. 92-90' I 1000 —I 2000 time(s) 3000 4000 Figure 537 Plot of the percentage of styrene, {[styrene]^lsryrene]0) x 100%, present at time (t), during reaction 6, Table 5.8. The value of [styrenelo was taken as the sum of [styrene], and [ethylbenzene]t, the concentrations at time (t), which were obtained from the GC printout. 100 1000 2000 3000 4000 timed) 5000 6000 Figure 538 Plot of the percentage of styrene, {[styrene]^[styrene]0) x 100%, present at time (t). during reaction 7, Table 5.8. The value of [styrenelo was taken as the sum of [styrenejt and [ethylbenzene]t, the concentrations at nine (t), which were obtained from the GC printout. measurements, Table 5.6). The ratio of erhylbenzene/Ru is 0.9 at 10 half-lives in this case. Further, the curve in Figure 5.38, at an even higher [KOH], shows a clearer demonstration of this two stage process, with the hydrogenation rate decreasing exponentially to about 1000 s, followed by relatively increased activity which slows markedly on approaching 6000 s. (Qouding of solutions at higher [KOH] interfered with visible absorption measurements, and no kobs values for decay of 3b were obtained above 0.25 M KOH; however, xm for this reaction at [KOH] = 0.3 M is probably less than 60 s, based on data from Figure 5.34, including a correction for a salt effect estimated from Figure 5.33.) The stoichiometric hydrogenation associated with the primary reaction of 3b with KOH/MeOH would be virtually over after 10 half-lives; in this case, the ratio of ethylbenzene/Ru is 0.7 at t = lOtia by which time less than 0.1% of the activity resulting from the presence of 3b would be remaining. This is further reason to believe that the observed two-stage hydrogenation at higher [KOH] is real, since the first stage occurs within 10 half-lives for loss of 3b. Thus, for semi-log plots used to estimate kobs for these hydrogenations at higher [KOH] (Figs. 5.39,5.40), the value for [styrene]«J[styrene]o was taken as the value at time (t) = 10tl/2, tw being estimated from syjectrophotometric measurements; [styrene] o was taken as the sum of the [styrene] and [ethylbenzene] which were obtained from the GC printout The variation of kobs with [KOH] and with [styrene], along with experimental conditions used, are summarized in Figs. 5.41 and 5.42, respectively, and in Table 5.8. The high value of kobs at 0.3 M (Fig. 5.41), is probably due to a contribution by the secondary hydrogenation process. The hydrogenation rates very approximately follow a first-order dependence on [KOH], with about the same intercept as obtained for the dependence on [KOH] for the decay of 3b. Figure 5.42 indicates a dependence on [styrene] that is at least close to first-order. The y-intercept of Figure 5.42 1.0 •OA I i | i | i |i i | i | i 0 100 200 300 400 500 600 timc(s) Figure 5.39 Semi-log plot, showing the first-order behaviour of the primary hydrogenation process, derived from the data of Figure 5.37 as described in the text. The function F is the same as defined for Figure 5.36. 1 . -1 1 ' i ' i ' i 1 i ' 0 100 200 300 400 500 time(s) Figure 5.40 Semi-log plot, showing the first-order behaviour of the primary hydrogenation process, derived from the data of Figure 538 as described in the text The function F is the same as defined for Figure 536. 0.010 0.008" 0.006-0.004 0.002 0.000 • • • 0.0 0.1 I 02 [KOH] 03 • for decay of 3b D k'ot6 for ethylbenzene formation 0.4 Figure S.41 Plot showing [KOH] dependence of k^, for styrene hydrogenation, in the reaction of 3b with methanol in the presence of styrene and KOH in THF (Table S.8); data from Figure S31 are included for comparison. 0.004 0.000 0.4 [styrene] Figure 5.42 Plot showing [styrene] dependence of Icjt, for styrene hydrogenation, in the reaction of 3b with methanol in the presence of styrene and KOH in THF (Table 5.8). (0.0011 s'1) is slightly lower than that (0.0017 s1) obtained by interpolation of the data of Figure 5.31 at this 0.060 M [KOH] where the [styrene] = 0. (The data of Figure 5.32, which shows the [styrene] dependence of kobs determined from spectrophotometrie measurements, give a kobs value of 0.002 S'1 at corresponding conditions.) Values of kobs at higher [styrene] (> 0.18 M) are greater than kobs at corresponding [styrene]. In reaction 8 within Table 5.8, at 0.060 M KOH, the data show stoichiometric hydrogenation of styrene based on [3b], as was generally the case in experiments with [KOH] below 02 M. Effects of adding excess PPh3 or isoPFA in these GC kinetic experiments were relatively small (Table 5.8), and it should be noted that these experiments were carried out using 3b recrystallized from CH2CI2 and the measured styrene hydrogenations were some 25 - 30% more than stoichiometric based on [Ru]; but the calculation of kobsvalues for the three experiments (reactions 11-13, Table 5.8) was based on a stoichiometric limiting amount of hydrogenation for comparison with the previous results. (Use of CH2CI2-recrystallized 3b, instead of the THF-recrystallized complex, did not significandy affect the kobs values obtained from spectrophotometric experiments.) Experiments were conducted for the detection of KCI, which deposited during the course of the primary reactions between RuCl2(PPh3)(isoPFA), 3b, and KOH/MeOH/THF. When the white precipitate was dissolved in water and mixed with AgN03 solution, a white solid precipitated which was insoluble in dilute HNO3 and presumed to be AgCl. The precipitate deposited during the Ru reaction must be KCI. Product KCI was unambiguously identified when CH2CI2 was used in place of THF. In the CH2CI2 'KCI detection experiment' described in section 5.2, the isolated white precipitate formed in the first minute of the reaction of 3b with KOH/MeOH analyzed correctly for KCI; and atomic emission analysis of an aqueous solution of the white solid (16 mg/100 mL) gave a reading intermediate between those obtained for 10 and 20 mg/100 mL KCI solution standards, thus indicating the expected K content. The residue derived 210 from the yellow supernatant decanted off from the precipitate of KCI exhibited a 31P{ !H} NMR spectrum which showed the residue was a mixture, including free PPh3 (5 -5.6) and some RuHCl(CO)(PPh3)(isoPFA), 6, (8 50.0 (d), 31.9 (d), 2JPP = 290 Hz) with associated isomers, as noted elsewhere. The lH NMR showed the associated upfield hydride resonance at 8 —15. Other unidentified species were present, but the reaction seems similar to that in THF, and KQ is considered to be formed in the basic methanol reaction in both THF and CH2Q2. In the experiment to detemrihe the formaldehyde content of a distillate from a reaction of 3b with KOH/MeOH in THF, a trace of CH20 (0.00023 mmol) was detected, which corresponds to -1% of a stoichiometric amount In the experiment to (ietermine the noil-volatile formaldehyde content in the reaction mixture produced in the high [styrene], high [KOH] experiment for H2 evolution, visible absorbance measurements on the yellow Hantzsch reaction product solution indicated a CH^ concentration of -19 |ig/mL in the 40 mL of worked up aqueous solution; thus, a total of -750 |ig (0.025 mmol) was produced in the reaction. Sampling of the headspace gas above the 3b/KOH/MeOH/THF reaction mixture in the 'CO detection experiment' revealed nitrogen but no CO. Deuterium transfer experiments were conducted on reactions of 3b with KOH/MeOH-d* and MeOH-di in the presence of styrene. The extents of hydrogenation in the MeOH-dt and MeOH-di reactions were 10 and 15% (by GC), respectively; i.e., -5 and 7.5 turnovers based on Ru. The *D NMR spectra of the distillates of the reaction mixtures obtained after reaction are shown in Figures 5.43a and b, respectively. The reference spectra are shown in Figure 5.44a (styrene/acetone) and Figure 5.44b (ethylbenzene/acetone). All the spectra shown in Figures 5.43 and 5.44 show a peak due to deuterium in unenriched acetone at 8 2.03, which provided a useful reference peak Both spectra in Figure 5.43 show peaks due to the deuteriated ethyl group of ethylbenzene, at 8 1.15 and 2.58, which are proof that methanol D atoms are transferred to styrene in the a) .[ • i • •. • i •. • • i n b) b bD J T T T Figure 5.43 *D NMR (46 MHz, ambient temperature) spectra of deuteriated ethylbenzene obtained by hydrogen transfer from (a) MeOH-cLj and (b) MeOH-di to styrene during the reaction of RuCl2(PPh3)(isoPFA), 3b, with MeOH-d4 and MeOH-d,, respectively, in the presence of KOH in THF. Peaks are assigned as follows: 'a', acetone: V, styrene; 'c', ethylbenzene; 'd', an unidentified impurity. reaction, and also weak peaks due to the deuteriated vinyl group of styrene at 5 52, 5.8 and 6.8, indicative of considerable exchange of vinyl hydrogen atoms with deuterium, because the NMR spectrum represents a GC fraction that contained perhaps 10% of the styrene present because of peak overlap with the ethylbenzene fraction. Both spectra also show an unidentified, strong peak at 8 2.9. Figure 5.44 *D NMR (46 MHz, ambient temperature) spectrum of (a) a 1:1 unenriched acetone/styrene mixture and, (b) a 1:1 unenriched acetone/ethylbenzene mixture. Peaks are marked as described for Figure 5.43. An experiment to determine what would happen to CH2O generated in a solution similar to those used for the reactions of 3b with KOH/MeOH When the foimalirvlCOH/MeOH/nff solution described in section 5.2 for the title experiment was neutralized, a white solid precipitated which was insoluble in HClaq, water, diethylether, acetone, methanol, benzene and dichloromethane, but was somewhat soluble in methanolic or aqueous KOH solutions. The mass spectrum showed mainly peaks at m/z = 30,29, and 28, probably due to HCHO+ CHO+ and CO+. Anal. Calcd. for H(CH20)„OH (n = 20 - 40): C, 38.83 - 39.41; H, 6.80 - 6.73. Found: C, 39.11; H, 6.90. Thus, the white solid is a polymerized form of formaldehyde, with an average of 20 - 40 carbons per chain, a range based on experimental errors in the analysis. 5.4 Discussion Solution behaviour ofRuCl2(PPh3)(isoPFA), 3b There is no evidence for the dissociation of PPh3 from 3b, as the visible spectrum is unaffected by the presence of added PPh3; nor is there any indication of the coordination of MeOH or THF to 3b, since the visible spectrum is the same in THF, THF/MeOH or the non-coordinating solvent, CHQ3. It appears that there is no observable coordination of styrene to 3b, since the 31P{1H) NMR spectrum of the starting complex, 3b, in CDC13 was unchanged on addition of up to a 10-fold excess of styrene. Reactions of RuChiPPhsXisoPFA), 3b, and derivatives formed from methanol in the absence of added base The RuHCl(CO)(PPh3)(isoPOF), 5,^PpHJ NMR coupling constants (2jPP = 188 Hz) are in the range typical for coordinated trans phosphorus ligands.91 The *H NMR values of 2JPH = 19 HZ for the upfield hydride triplet at 8 -13.9 are in the range typical of a hydride cis to phosphorus atoms coordinated to ruthenium(n), and the position of the shift (8 upfield of -10) is typical of a hydride trans to a ligand of low trans influence, such as CI.89 The downfield shift (8 6.52) of the proton on the carbon a to the ferrocene ring is typical of values (8 5.0 - 6.5) obtained for coordinated isoPFA or PPFA (sect. 2.2), suggesting that the OMe group on the isoPOF ligand is coordinated in solution. The spectroscopic data are thus consistent with the crystal structure shown in Figure 5.6. In this structure, the arrangement of ligands is very close to octahedral, with a P1-RU1-P2 axis that is almost linear, ZP1-RU1-P2 = 177.55(6). The ZCl-Ru-CO angle is 101.4(2), and the other three angles that are equatorial to the P1-RU1-P2 axis are slightly reduced from 90°: ZHi-RurCO = 82.5, ZO2-RU1-H1 = 87.7, Z02-Rui-Cl! = 88.4. The hydrido-carbonyl H-C distance is 2.2 A, precluding any C-H interaction (there is precedent for intramolecular hydride transfer to a carbonyl on ruthenium).155 The Ru-P distances of 2.358(2) and 2.354(2) are close to the Ru-P distances, ranging from 2.301 to 2.383(2), 215 obtained in the present work for other ruthenium(II) hydrido-carbonyls where Ru-P is approximately trans to another Ru-P bond (Tables 5.1,5.2,5.5). For the hydrido-chloro-phosphines without carbonyl groups (cf. Tables 5.1,3.7 and 3.8), there is a tendency for the Ru-P distances to be slightly shorter, possibly because of greater electron density on the ruthenium atom due to the absence of back-bonding to a carbonyl, and consequendy increased back-bonding to phosphines. The Ru-H distance of 1.57 A for 5 is generally close to those of other ruthenium(IT) terminal hydrides of complexes synthesized in the present work, which are in the range: 1.47(4) to 1.53(6) A (Tables 5.1 and 3.8). The formulation of 5 is analogous to that of the hydrido-carbonyls RuHCl(CO)(PR3)3, formed from the reaction of other ruthenium phosphine complexes with methanol and reported in the literature (see sect. 5.1). A possible stoichiometry for the formation of the complex 5 is: RuCl2(PPh3)(isoPFA) + 2CH3OH H2NMe2+Cl- + H2 + RuHCl(CO)(PPh3)(isoPOF) 5 (5.25) No attempt was made to detect H2NMe2+Cl* or H2 in the product mixture. A suggested approach to isolate H2NMe2+Q_ would be to induce its crystallization by addition of benzene and hexane, as was done in the isolation of H2NMe2+Cl- from the reaction of 3b with H2, where the same NMe2 cleavage was noted (sect 4.3). The mechanism for the formation of 5 could involve protonation of the -NMe2 by coordinated CH3OH, or from free proton following methoxide formation, to give non-ccordinated -NMe2H+ and a coordinated methoxo group, followed by substitution, at the carbon atom a to the ferrocene, of the NMe2H by the OMe group. This could occur after loss of Cl and formation of the hydrido-carbonyL because 6, RuHa(CO)(PPh3)(isoPFA), can also react in neat methanol to give 5 (see below). More details on a possible mechanism are given below in connection with the mechanistic study on the reaction of 3b with methanol in the presence of added base to give 6. 216 Decarbonylation of alcohols by low valent phosphine complexes has been proposed to occur via alkoxide formation.90,133'156'157 It was stated in 197790 that this proposal had not been proved, and this is probably still the case. The present reaction of 3b with methanol perhaps presents some mdirect evidence for the proposal, since 5 may represent an "arrested methoxo intermediate" in which the methoxo ligand, before further dehydrogenation can take place, is transferred to the ferrocene ligand via substitution at the a-carbon position. The lH NMR spectrum of 5 prepared from methanol-cU, with ortholmeta, para and hydride/ferrocene peak ratios reduced from the non-deuteriated values, indicates that exchange between hydride and orr/io-phenyl hydrogens occurs in solution, although it is not clear whether this occurs in some intermediate, or in the final product, or both. There is clearly no exchange involving the methoxy group. The extent of deuteriation observed at the ortho-phenyl position shows that other processes are also involved in this system. Detection of trace formaldehyde and methyl formate in the product mixture, along with other volatile components, is further evidence that other dehydrogenation processes, that can feed more deuterium atoms into the exchanging hydride/orr/w-phenyl system, are also occurring in solution. The observed hydride/orf/io-phenyl exchange is a further example of the well-established orthometallation reaction of ruthenium(H) phosphine complexes.91'158-159'160'161'162*163'164 The extent of catalytic methanol dehydrogenation, based on the preliminary search described above for organic products in the system, appears to be small. In the Ru acetate methanol dehydrogenation system investigated by Saito's group and described in section 5.1, precipitation of the hydrido-carbonyl RuQ(CO)(PEtPh2)3 marked the deactivation of the catalytic methanol dehydrogenation. Thus, in the isoPFA system, deactivation by hydrido-carbonyl formation may just be more rapid than in the Ru acetate systems, perhaps owing to more facile Ru-C bond formation and, therefore, faster decarbonylation to give the hydrido-carbonyl. It should be noted that, although hydrido-carbonyl formation has been associated, by Saito's group, with catalyst 217 deactivation (sect S.l and references 115 and 145), Smith and Maitlis came to the opposite conclusion: they observed no organic products until carbonyl formation had occurred.128 Morton and Cole-Hamilton123 saw somewhat decreased activity upon formation of RuHCl(CO)(PPh3)3 in their methanol dehydrogenation system. The apparent contradictions may well be partly due to the differences in temperature used by the different groups. Both Smith and Maitlis128 and Morton and Cole-Hamilton123 carried out their reactions at 150 °C, at which temperature the hydrido-carbonyls do exhibit dehydrogenation activity. The experiments of Saito's group were done at 66 °C, which may just be too low for any significant activity from the hydxido-carbonyls in their system. Also, the RuH2(PPh3)3-based systems (Morton and Cole-Hamilton), and the Ru(OAc)Cl(PR3)3 systems (Saito's group) are, in fact, more active than the respective hydrido-carbonyls formed in the systems. The RUCl2(PPh3)3 complex as claimed may be less active than the hydrido-carbonyls produced in methanol.128 But it must be noted that the solubility of RuCl2(PPh3)3 in methanol is low, and the bulk of the ruthenium remains suspended as solid (hence incapable of homogeneous dehydrogenation catalysis) until the soluble hydrido-carbonyl [(CO)(PPh3)2Ru(n-H)(ii-a)2Ru(CO)(PPh3)2]+is formed.122 Thus, the lack of significant activity prior to carbonyl formation, observed by Smith and Maitlis, may be due partly to a lack of significant solubility prior to carbonyl formation. The single hydroformylation experiment described in section 5.2 was carried out partly because of the observation that many hydroformylation catalysts are complexes containing carbonyl groups,lb«165 and the recent success of a hydroformylation catalyst containing a P-O ligandlb gives hope that 5 might also be effective for this purpose. The strong smell of hydroformylation products, coupled with the GC evidence, show that the complex 5 is an active hydroformylation catalyst, albeit a not very selective one. The transfer-hydrogenating activity of 3b generally appears comparable to that of RuCl2(PPh3)3 as reported by Maitlis' group.122 Unfortunately, these workers reported only the number of moles of reactants and did not give volumes or concentrations. Also, 218 the reaction conditions are different, Maitlis' group using higher temperature (150 °C) and shorter times. Nevertheless, RuCl2(PPh3)3 also catalyzed the reduction of cyclohexene in only trace amounts, while the reduction of vinylmethylketone was much more efficient. As is common for hydrogen transfer from alcohols (sect 5.1), reduction of the ketone acetophenone, and the oc,p"-unsaturated ketone vinylmethylketone, was more effective than reduction of olefins. Reactions ofRuCl2(PPh3)(isoPFA), 3b, with methanol in the presence of added base With Proton Sponge® as added base: This investigation of the reaction of 3b with basic methanol was actually prompted by the observation, by Thorburn,133 of a colour change, from green to yellowish, on stirring RuCl2(PPh3)(PPFA), 3a, under hydrogen in toluene/methanol in the presence of Proton Sponge® (see sect 4.1); he had also discovered the reaction of Ru2CU(dppb)2 with methanol to give Ru2H(^-H)(u,-Cl)2(CO)(dppb)2. One mechanism considered133 involved formation of both 'RuHCl(dppb)' and 'RuHCl(CO)(dppb)' in sufficient amounts to be able to react together to give the observed yield of product: Ru2Cl4(dppb)2 + 2MeOH 2'RuHCl(dppb)' + 2CH20 + 2HC1 (5.26) Ru2Cl4(dppb)2 + 2MeOH - 2'RuHCl(CO)(dppb)' + H2 + 2HC1 (5.27) 'RuHCl(CO)(dppb)' + 'RuHCUdppb)' - Ru2H(n-H)(n-Cl)2(CO)(dppb)2 (5.28) In the present reaction between RuCl2(PPh3)(isoPFA) and methanol in the presence of Proton Sponge®, analogous stoichiometrics could be proposed for the formation of 7 and 6, respectively: pc Rua2(PPh3)(isoPFA) •+ MeOH » RuHCl(PPh3)(isoPFA) + CH20 + HC1 3b 7 (5.29) PS RuCl2(PPh3)(isoPFA) + MeOH —RuHCl(CX))(PPh3)(isoPFA) + H2 + HCI 3 b 6 (5.30) These two hydrides appeared to be the major ruthenium products, but spectroscopic determination of the product distribution was not carried out Formaldehyde was produced, and, where the reaction appeared to have gone to completion, the amount was 0.5 mol CH2O per mol of 3b. Gas evolution of -1 mol/mol Ru was also detected and, considering that hydrogen is evolved in the reaction between 3b and KOH/methanol (also, with the formation of 6), the gas evolved in the present case is almost certainly H2 as well. The alternative CO gaseous product is highly unlikely. Thus, there is some support for the stoichiometrics of equations 5.29 and 5.30, although only 0.5 mol H2 is expected if there is no other inechanism for H2 production. In reaction 5.29, the formaldehyde is dissociated before it is decarbonylated, and therefore no CO ligand results; in reaction 5.30, decarbonylation of CH2O would give 6 and H2. The mechanisms for these reactions probably involve alkoxide, TI2-CH20 and, in the case of equation 5.30, formyl /mtermediates similar to those suggested for other examples of methanol decarbonylation.90'133 As noted in section 5.1, examples of such intermediate complexes have been characterized. More details on a probable mechanism are given later in connection with the mechanistic study on the reaction of 3b with KOH/MeOH A reaction between 3b and CH2O such as: PS RuCl2(PPh3)(isoPFA) + CH20 —RuHCl(CO)(PPh3)(isoPFA) + HCI 3b 6 (5.31) is also a possibility. Reactions of ruthenium complexes with CH2O have been reported.90'145 The present reaction is similar in some respects to the reaction of RuH(C^H4PPh2)(PPh3)(Et20) with methanol in diethylether at 0 °C to give a rnixture of RuH2(CO)(PPh3)3 and RuH2(PPh3)4, as reported by Cole-Hamilton and Wilkinson.91 220 The structure of 6 should be compared particularly with that of 5, RuHCl(CO)(PPh3)(isoPOF), because, as discussed later in this section, the former complex reacts with methanol, in the absence of base, to give the latter. In the crystal structures of both complexes, the coordinated P atoms are mutually trans. It seems clear from the NMR data that, in fact, all the observed geometric isomers of 6 have mutually trans phosphines. There appears to be only one isomer of 5 in solution; however, it should be noted that the 31P{1H) NMR resonances for this complex are ambiguous, and could be interpreted as two overlapping doublets with J larger than the difference in 8, or two close doublets with J smaller than the difference in 8. In addition to the possibility of geometric isomers, there is also the possibility of diasteriomers for the isoPOF complex, 5, as is also the case for the isoPFA complex, 6; it should be noted that the isoPOF ligand may be R.S-, as in Figure 5.3, or SJZ-, and the ruthenium atom is also a centre of chirality in the molecule. The temperature dependence of the JH NMR spectrum of 6 also has a bearing on the structural comparison with the isoPOF analogue. The most likely interpretation of -this temperature dependence data is that the -NMe2 functionality of the isoPFA ligand rapidly binds to, and dissociates from, the ruthenium at about 20 °C, but remains coordinated below 0 °C. Such lability of the -NMe2 could provide a mechanism for the formation of other geometrical isomers. Thus, while in the crystal structure of 6 the hydride is trans to -NMe2 and the CO is trans to Cl, dissociation of the -NMe2 could allow the chloride to swing away from the hydride, to give mutually trans hydride and chloride (as observed in the crystal structure of 5 (Fig. 5.3)) upon re-ccordination of the -NMe2 of the isoPFA. The structure of 7 is very sinailar to that of RuHCl(PPh3)3.166 Both are best described as strongly distorted trigonal bipyramidal. In the case of RuHCl(PPh3)3, two phosphines are at the apices of this "trigonal bipyramid", with a "ZP3-RU-P2" angle of 153.1°, with the other phosphine, the hydride and the chloride approximately equatorial. In the case of 7, the PPh3 and the nitrogen of isoPFA are at the apices ("ZN1-RU1-P2" = 157.1°) with the P atom of isoPFA, the chloride and the hydride approximately equatorial. Other parameters for these two structures are compared in Table 5.9, by listing side-by-side corresponding angles and distances. The angle of 121.4° between "the equatorial" P and CI atoms in the case of RuHCl(PPh3)3 is close to the corresponding angle of 124.9° in 7. Skeletal bond distances in the two complexes also correspond quite closely. Slightly shorter Ru-P distances in 7 can be partly attributed to higher electron density on Ru due to the presence of more basic ligands, combined with fewer ^back-bonding ligands. Comparison of the structure of 7 with that of its CO adduct, 6, reveals the effect of the carbonyl on the various bond distances. Table 5.1 lists the corresponding distances of these two structures. The generally longer skeletal bond distances in the case of 6 can be Comparison of structural parameters of RuHCl(PPh3)(isoPFA), 7, and RuHCl(PPh3)3166 Table 5,9 . P; 7 RuHCl(PPh3)3 Bond distances (A): Rui-Cli Rui-Hi RurPi RurP2 Rui-Ni 2.441(1) 1.52 Ru-Cl Ru-H Ru-Pi Ru-P2 Ru-P3 2.421 1.70(15) 2.206 2.361 2.329 2.178(1) 2.250(1) 2.265(5) Angles (°) Ni-RurP2 Pi-Rui-Ch 157.1(1) 124.9(1) P3-Ru-P2 PrRu-Cl P,.-Ru-H 153.1 121.4 -89 222 attributed to decreased electron density on the ruthenium, and consequent decreased back-bonding to other ligands, caused by the high degree of back-bonding to CO. With KOH as added base: Reactions in the 3b/KOH/methanol system are complex but some elucidation has been possible. The 6 complex is at least the main ruthenium product of the reaction when the KOH:Ru mole ratio is 1:1, and a possible stoichiometry is: RuCl2(PPh3)(isoPFA) + KOH + CH3OH RuHCl(CO)(PPh3)(isoPFA) + KCI + H2 + H20 (5.32) At higher KOH:Ru mole ratios, more complex product mixtures are obtained. In the experiments summarized in Table 5.4, no red solution was observed at any time during the reaction of 3b with methanol with 1.8 M KOH (KOH:Ru = 240), whereas wine-coloured solutions persisted at KOH:Ru ratios below -120. In the preparative scale experiments, wine-coloured solutions were produced when KOH:Ru mole ratios ranged from 10 to 50. The white dihydride RuH2(CO)(PPh3)3 is a major product present in these wine-coloured .reaction mixtures. Redistribution of the coordinated phosphine ligands thus occurs in this system, as well as in the reactions of 3a and 3b with H2 (chapter 4). The formation of a ruthenium complex with three PPh3 ligands requires the dissociation of both isoPFA and PPh3 ligands from other complexes present in the system; in fact, the presence of these ligands in the KOH/methanol system is indicated by NMR data, and, in one experiment, free isoPFA was isolated by column chromatography. Other partially characterized complexes are also present in these wine-coloured reaction mixtures, which certainly contain Ru complexes with hydride, carbonyl, and isoPFA ligands. Column chromatography on silica results in some decomposition of the complexes, as judged by differences between the NMR spectra of the residues of some fractions and those of unseparated residues of supernatants of these KOH reactions. However, certain complexes eluted from the column, mcluding a yellow hydride with H&CDCb -18.7 (dd), and a red hydrido-carbonyl, X, containing two isoPFA ligands and with h8CDCI3 17.9 (dd), -18.2 (dd), survived the chromatography. The latter complex contains no PPI13 ligands; no v (OH) was observed in the IR, and the intense red colour suggests a five-ccordinate species.90 The lH NMR integrations indicate that two hydrides are present A tentative formulation of this complex, X, is: The hydride resonances of the two complexes (yellow and red species) resemble those observed (Fig. 5.15) for the residue of wine-coloured reaction mixtures after solvent removal only. Also, the 31P{ 1H) NMR spectrum of the red hydrido-carbonyl exhibits peaks resembling those of a wine-coloured 3b/KOH/methanol in situ reaction mixture (Fig. 5.14). Formation of RuH2(CO)(PPh3)3 and a red species (Xmax = 542 nm) from the reaction of 6 with KOH/methanol does occur, and this intermediate presumably provides the pathway for the secondary reactions observed when 3b reacts with KOH/methanol. Experiments on the reaction ofRuCh(PPhz)(isoPFA), 3b, with KOH/methanol, in THF in the presence or absence of styrene The reaction of 3b with KOH/MeOH was also conducted in 3:1 v/v THF/MeOH, since all ruthenium species in the system were soluble in this solvent, thus facilitating kinetic studies; fortunately, *H NMR data show that no reaction occurs between KOH/MeOH and THF at ambient temperature, in the absence of styrene and 3b. In the 3b/KOH/MeOH/THF reaction, no resonances due to the product 6, RuHCl(CO)(PPh3)(isoPFA), were observed in thein situ 31P{1H) NMR spectra, although an AB pattern due to a species 6' was observed. The 31P{1H) NMR spectra of residues + cr from these reactions, however, when stopped after the 'primary reaction' had occurred, showed the presence of 6. (The work-up described in section 5.2 for the THF/MeOH styrene-containing reaction solutions is not recommended for isolation for Ru products, but it produced conclusive evidence that 6 can be recovered from these reaction mixtures.) Further evidence that 6 is present in some form 6' (possibly with coordinated THF, see below) is obtained from the observation mentioned above (p. 162) that the reaction proceeds on to the formation of the red species (^PpH) NMR 8 61,62.5,66) which are also formed in the absence of THF and in the absence or presence of styrene. It should be remembered that the coordinated N atom of the isoPFA is labile (see the discussion of the temperature dependence of the *H NMR spectrum of 6, above), and hence could be readily replaced by a ligand, L, such as THF, which, however, may be lost during work-up, when its concentration becomes low. Since the 31P{1H} NMR spectrum of a solution of 6 is unaffected by the presence of ~ 1 mole of styrene, L cannot be styrene. The data could be accounted for by an equilibrium of the type: P—RuHCl(CO)(PPh3) + L = P—RuHCl(CO)(PPh3) (5 33) 6 L = THF? 6' Thus, the reaction in the presence or absence of THF may well proceed along the same path, viz.: RuCl2(PPh3)(isoPFA) + KOH + CH3OH RuHCl(CO)(PPh3)(isoPFA) + H2 + H20 (5.32) but the form assumed by 6 is different in the 3:1 v/v THF/methanol solution. The temperature dependence of the ratio of the integration of the 6' AB pattern to that of the PPh3 resonance may represent an equilibrium involving loss of PPh3 from the 6' species: 1 r\y P—RuHa(CO)(PPh3) — P—RuHCKCO) + PPh3 (5 34) The complex 6 in the form of 6' (and its phosphine-dissociated equilibrium species) is probably the main product detected in the spectra of Figures 5.20 and 5.21. The higher yield of RuH2(CO)(PPh3)3, from a secondary reaction in the neat methanol case could be rationalized in terms of low solubility in this solvent, and shifts in various equilibria. The visible spectrophotometric experiment giving the spectra shown in Figure 5.25 is comparable to that giving the 31P{1H) NMR spectra shown in Figures 5.20 and 5.21; the [3b] and [KOH] are much lower in the 'visible' experiment, but the [KOH]/[Ru] ratio is about the same (9:1), and the [styrene] in the two experiments is about the same (-0.44 M). Thus, the main product at the end of the first stage of the 3b/KOH/MeOH/THF reaction, indicated by the visible absorption spectrum'd' in Figure 5.25, is perhaps the THF complex 6'. The yellow complex, RuHCl(CO)(PPh3)(isoPFA), 6, exhibits a weak visible band at 450 nm, and it is likely that the THF complex is also yellow and mainly responsible for the yellow colour observed. At intermediate [KOH], generally <0.1 M, a -much slower secondary colour change to red occurs, as exemplified in Figure 5.25 (curves e - j). The growth in red colour is accompanied by an increase in an intense band at 542 nm (t1/2 = ~14h), as observed for species X and possibly other red ruthenium hydrido-carbonyl species are formed which are also produced in the reaction in neat methanol. It seems clear from NMR spectra (Figs. 5.14,5.15,5.22) that there are more than one species produced in secondary reactions in basic methanol, or basic methanol/THF. The KOH ratio is important in (ktermining which secondary processes occur (see also Table 5.4, for the neat methanol reaction). At [KOH] higher than 0.2 M, for the THF/MeOH reactions, the formation of the red species is less readily observed, and instead, formation of a second yellow species is seen. This latter change is accompanied by a small decrease in absorbance at 605 nm (yellow -~yellow); this reaction is also slower than the formation of 6', and both these secondary colour changes, at intermediate [KOH] (yellow -red), or at high [KOH] (yellow • yellow), can be monitored at 605 nm, independently of the first reaction that gives 6'. The behaviour of the spectrum at 450 nm during the yellow "-red process with styrene absent (Fig. 5.26) is similar to that with styrene present; but with styrene absent there is an unexplained gradual increase (Fig. 5.26) at 450 nm that accompanies the growth of the 542 nm band. Kinetic studies on the secondary reactions were not attempted. The formation of -1 mol gas during the primary reaction of 3b to give 6' in the absence of styrene, and the lack of gas evolution, at a KOH:Ru ratio of ~1, coupled with the formation of ~1 mol of ethylbenzene in the presence of styrene at [KOH] below -0.06 M, indicate that -1 mol Hymol Ru is produced during this primary reaction. This, together with the detection of KCI produced during this stage, support the stoichiometry shown in equation 5.32 for the reaction in the absence of styrene; but in the THF case, 6 presumably further reacts to give 6', as in equation 5.33. The production of 6' and KCI occurs whether styrene is present or absent, suggesting that the pathways for the styrene hydrogenation and uncoupled methanol dehydrogenation are very similar, perhaps differing only in a couple of steps, and having the same stoichiometry, except that the H2 is transferred to styrene if styrene is present It is noteworthy that this is perhaps the first time that KCI has been positively identified in a reaction of a platinum metal chloro complex with alcoholic KOH, although KCI formation has been proposed.143 This reaction of 3b to give 6, or 6', is at least the dorninant reaction in the primary process (sect 5.3); this is in contrast to the relative complexity of reactions during the second stage (vide supra). Therefore, a kinetic study of the primary process both in the presence of styrene (Fig. 5.25, curves a - d) and in the absence of styrene, has been carried out The rate of decay of 3b, measured in terms of kobs, is essentially independent of styrene, and essentially independent of added PPh3 and isoPFA in the absence of styrene; the rate of styrene hydrogenation is essentially independent of added isoPFA but inhibited by -10% in the presence of 1.4 x 10"2 M added PPh3 (i.e., PPh3:Ru = 4). Results on chloride dependence of the rate of decay of 3b, hampered by insolubility of KCI, tentatively indicate an inverse dependence of kobs on added chloride (sect. 5.3). Spectrophotometric studies (sect 52,5.3) monitoring the disappearance of 3b in the presence of base showed that there is a pathway with a second-order dependence on [KOH] for this first step in the reaction of 3b with KOH/MeOH, as well as a base-independent pathway: kobs = 1.3 x lt>3 + 0.082[KOH]2 (5.24) Since KOH can react with MeOH to give OMe* (eq. 5.35), attack on 3b by OMe" must be considered in proposing a mechanism for the second-order base dependent pathway:. For the equilibrium in equation 5.35: OH- + MeOH OMe- + H20 (5.35) a value for the equilibrium constant, K, of 18 has been reported for the NaOMe/neat MeOH system;167 this value yields a plot of [OMe-] against [OH-] that is close to linear (Fig. 5.45), over the range of [OH-] used to determine the second-order [KOH] dependence in THF/MeOH in the present work. I. 0.3 0.2 -0.1 -0.0 1 • Tl D D D B i l 0.0 0.1 0.2 0.3 0.4 [OH] Figure 5.45 Plot of [OMe*] versus [OH-] based on the equilibrium constant, K (18), for the reaction of NaOH with MeOH at 20 °C.167 Determination of this equilibrium constant is discussed further in reference 168. 228 Measurements of K for equilibrium 5.35 in the THF/MeOH system are not available; however, based on the data for the NaOMe/MeOH system, a linear dependence of [OMe"] on [OH-] is assumed in the present work on the THF/MeOH reaction. Therefore, a second-order dependence on [KOH] is accounted for by the mechanistic proposal shown RuCl2(PPh3)(isoPFA), 3b rapid step |[?cif RuCl(OMe)(PPh3)(isoPFA), i slow step OH" | - PPh3 [RuCl(OH)(OMe)(isoPFA)]\ ii j-OH" RuHCl(T|2-CH20)(isoPFA), iii I RuIVH2Cl(CHO)(isoPFA), iv styrene, RuwHCl(CHO)(ethylbenzyl)(isoPFA), vi PPh3 H2J PPh, ethylbenzene RuCl(CHO)(PPh3)(isoPFA), v RuHCl(CO)(PPh3)(isoPFA), 6 Figure 5.46 Proposed mechanism for the second-order base dependent pathway for reaction 5.30 via methoxide attack on 3b. in Fig. 5.46 for the reaction of 3b with KOH/MeOH in THF. To account for the second-order dependence on KOH, and for the inhibition of the rate by chloride, the initial step in the mechanism (Fig. 5.46) is proposed to be a rapid step involving replacement of CI" by OMe-: K RuCl2(PPh3)(isoPFA) + OMe" RuCl(OMe)(PPh3)(isoPFA) + Cl" 3b i (5.36) followed by the slow replacement of PPh3 by OH': RuCl(OMe)(PPh3)(isoPFA) + OH" —^—^ [RuCl(OH)(OMe)(isoPFA)]- + PPh3 i ii (5.37) This step is supported by the observation, by 31P{1H) NMR spectroscopy (sect. 5.3), of PPh3 dissociation early during the overall reaction (5.32). Hydride transfer to the vacant site, accompanied by dissociation of OH- is next: [RuCl(OH)(OMe)(isoPFA)]- RuHCl(ri2-CH20)(isoPFA) + OH" ii iii (5.38) The coordinated formaldehyde is likely to be dihapto-coordinated, because it is further dehydrogenated (see discussion in sect. 5.1) and also there is structural precedent for rj2-CH20 ligands on members of the iron triad (sect. 5.1). Step 5.38 is followed by further migration of hydride from CH20 to give a Ru(IY) dihydride with a carbon-bonded formyl ' ligand; in proposing this intermediate, consideration is given to the structural precedent for formyl complexes in ruthenium chemistry (sect 5.1): RuHCl(ri2-CH20)(isoPFA) RuIVH2Cl(CHO)(isoPFA) (5.39) • • • • in iv Reaction with available PPh3 could then lead to a reductive ehmination of hydrogen to give a RufJJ) species: RurvH2Cl(CHO)(isoPFA) + PPh3 — RuCl(CHO)(PPh3)(isoPFA) + H2 iv v (5.40) Finally, hydride transfer from the five-coordinated formyl intermediate gives the product, 6: RuCl(CHO)(PPh3)(isoPFA) RuHCl(CO)(PPh3)(isoPFA) (5.41) In the presence of styrene, step 5.40 could be circumvented by reaction of the 16-electron intermediate iv with styrene (Fig. 5.46), followed by concerted reductive elimination of ethylbenzene and coordination of PPh3. The slightly inhibitive effect of added PPh3 on the rate of styrene hydrogenation could be due to competition of PPh3 with styrene for intermediate iv. The parallel decreases in hydrogenation rate and [3b] decay rate during reactions at these KOH concentrations below 0.06 M, together with the chloride dependence data, show that the rate determining step for the formation of 6' from 3b occurs early in the mechanism of the styrene-hydrogenation-coupled pathway as well as in the methanol-dehydrogenation pathway in the absence of styrene. With the second step (eq. 5.37) in the mechanism 0?ig. 5.46) slow and rate-determining, the reaction rate, = k2 [OH-][i] (5.42) As the value of [i] depends on equilibrium 5.36: [i] = K'3y (5.43) and because [OMe-] is approximately proportional to [OH-]: [OMe-] = C[OH-] (5.44) where C is a constant, equation 5.43 becomes: [i)=KC[Mp£l (5.45, Thus the net rate of decay of 3b is given by: _d[3bj _ k2KC[3birOH-]2 dt " [C1-] This rate law is thus consistent with the observed second-order dependence on [KOH], as well as the first-order dependence on [RU]T and the inhibitive effect of added chloride; also, the mechanism allows for the observed phosphine-independent rate, and is consistent with the observed PPI13 dissociation. 231 A possible mechanism for the base-independent pathway is shown in Figure 5.47. To account for the inhibitive effect of Cl' on the rate, and the fact that 3b shows little evidence of reaction with methanol in the absence of base at ambient temperature, the first step (3b » i, Fig. 5.47) is taken as a reversible one with a small K value, involving replacement of Cl" by MeOH: RuCl2(PPh3)(isoPFA) + MeOH [RuCl(MeOH)(PPh3)(isoPFA)]+ + Cl-k-i 3b i (5.47) RuCl2(PPh3)(isoPFA), 3b , •• il MeOH slow step 1L . Q-[RuCl(MeOH)(PPh3)(isoPFA)]+, i rapid step j ^ - H* [RuCl(OMe)(PPh3)(isoPFA)], ii 1 RuHCl(ri2-CH20)(PPh3)(isoPFA), iii I Ru(ri2-H2)Cl(CHO)(PPh3)(isoPFA), iv Styrcne H2 j RuIVHa(CHO)(ethylbenzyl)(PPh3)(isoPFA), vi RuCl(CHO)(PPh3)(isoPFA), v — ' • ethylbenzene RuHCl(CO)(PPh3)(isoPFA), 6 Figure 5.47 Proposed base-independent pathway for the reaction of 3b with KOH/MeOH in THF. The second step is reversible loss of a proton from coordinated MeOH (i —ii, Fig. 5.47): [RuCl(MeOH)(PPh3)(isoPFA)]+ :==^r [RuCl(OMe)(PPh3)(isoPFA)] + H+ k-2 i ii (5.48) Subsequent steps in Figure 5.47 describe the stepwise hydride migration and liberation of hydrogen associated with the dehydrogenation of the OMe' ligand. Added base would rapidly neutralize the protons produced in the second reversible step (eq. 5.48), and thus shift the system toward formation of ii. Thus, if the neutralization is rapid, then the scheme allows for the rate of the overall process to be independent of [OH'], although at least 1 equivalent of added base (based on Ru) is required for formation of ii and final product formation. The rate for this base-independent pathway can be written as: = k![3b][MeOH] - k.![i][a-] (5.49) A steady state approximation for [i] gives: k![3b][MeOH] + k_2[ii][H+] = k_1[Cl][i] + k2[i] (5.50) while in the presence of OH*: k.2[ii][H+] « 0 (5.51and therefore equation 5.50 becomes: ki [3b] [MeOH] = k_i[Cl-][i] + k2[i] (5.52) Rearranging gives: _ ktflbHMeOH] l,J ~ k.![Cl] + k2 C5-33) and substituting into equation 5.49 and rearranging finally gives: -^ = k1[3b][MeOH]{l-ki[^jf]C1"] (5.54) This rate law is consistent with the observed first-order dependence of the rate of decay of 3b, as well as being independent of base, although the rate law only holds if at least 1 equivalent of base is present mtermediate iv is presented as an H2-H2 complex in view of the propensity of 18-electron Ru(H) complexes to stabilize this ligand (sect 3.1). The step involving equilibrium loss of H2 (the iv - v step, Fig. 5.47) is consistent with the typical reversibility of coordination of Ti2-H2 ligands (sect 3.1). Styrene, if present, could react with iv to give the Ru(TV) hydrido-alkyl complex, vi, with reductive elimination of ethylbenzene giving v (Fig. 5.47). Organic and gaseous products from high [KOH] reactions The polymerized formaldehyde formed in the farmalin/KOH/methanol/IHF mixture is regarded as high grade.117 Such polymerization would account for the lack of formaldehyde detected in distillates from reactions of 3b with KOH/MeOH in THF. Thus, at least in the presence of a large excess of KOH, these reactions generate approximately stoichiometric amounts of paraformaldehyde. The dehydrogenation of methanol in these reactions of 3b at high [KOH] appears to be 'catalytic1, and the three moles of gas produced in such an experiment at [KOH] = 0.38, is probably H2, as CO was not detected by GC in a comparable experiment The Figure 5.47 mechanism and the reaction ofRuCtyPPhsXisoPFA), 3b with methanol in the presence of Proton Sponge® Some discussion of a mechanism for the reaction of 3b with methanol in the presence of Proton Sponge® was presented earlier (p. 219); the reaction scheme shown in Figure 5.47, proposed for the KOH reaction, provides details of a probable mechanism for the PS reaction. No kinetic studies were carried out on this reaction, but products RuHCl(CO)(PPh3)(isoPFA), 6, and RuHa(PPh3)(isoPFA), 7, were isolated and -0.5 mol CH20/mol 3b was produced. The reaction scheme in Figure 5.47 accounts for the 234 formation of 6, while dissociation of CH2O from intermediate iii would account for the formation of 7 and CH2O. The Figure 5.47 mechanism and the reaction ofRuCl2{PPh3)(isoPFA), 3b, with methanol in the absence of base As discussed earlier, the complex 3b reacts with methanol in the absence of base to give RuHCl(CO)(PPh3)(isoPOF), 5, with complex 6 not being formed in detectable amounts; but as 6 reacts with methanol in the absence of base to give 5,6 is a possible intermediate in the 3b *• 5 reaction (5.25). Possible stoichiometrics for these steps are: RuCl2(PPh3)(isoPFA) + MeOH - RuHa(CO)(PPh3)(isoPFA) + H2 + HC1 3 b 6 (5.54) proceeding via the mechanism in Figure 5.47, and: RuHCl(CO)(PPh3)(isoPFA) + MeOH » RuHCl(CO)(PPh3)(isoPOF) + HNMe2 6 5 (5.55) proceeding via steps suggested on page 215. Reaction 5.55 would be followed by a neutralization: HC1 + HNMe2 • H2NMe2+Cl' (5.56) Because 6 is not observed as a product in the absence of base, this proposal involving 6 as an intermediate requires step 5.54 to be slow and step 5.55 to be relatively fast It should be remembered that the reaction of 3b with methanol was very slow, requiring refluxing 3b in methanol for several days (sect 2.4); and the reaction scheme proposed in Figure 5.47 is expected to generate 6 (and therefore 5) slowly if base is not present to shift the first two reversible reactions (eqs. 5.47 and 5.48) toward products. 235 Chapter 6 Summary and suggestions for further work The previously known23 complexes, Ru2H4Cl2(PR3)4, have now been correctly reformulated as the rj2-H2 species (Ti2-H2)(PR3)2Ru(p-Cl)2(p-H)RuH(PR3)2 (R = Ph, ptol), la and lb, and it is confirmed that in solution they are dimeric and undergo no ligand dissociation. The complexes la and lb decompose under nitrogen or vacuum to give the corresponding RuHCl(PR3)3 complex, and, at least in the case of lb, a species with a four-spin 31P NMR spectrum which had been noted previously by Dekleva.21 Also, a new analogue of complexes of type 1 is reported: the complex (T|2-H2)(isoPFA)Ru(p-Cl)2(p-H)RuH(PPh3)2,4, is formed from the reaction of RuCl2(PPh3)(isoPFA), 3b, with H2 in methanol/benzene, and a crystal structure of 4 . shows the T|2-H2 ligand. Complexes la, lb and 4 all react with 1-hexene to give hexane; the main ruthenium phosphine product in the case of 1 is the corresponding RuHCl(PR3)3 . complex, while 4 reacts to give a complex mixture of ruthenium phosphine complexes, including 3b. The amount of hexane formed from the reaction of 4 with hexene is quantified as 2 mol/mol 4. The results summarized above are of particular interest and significance for the following reasons: a) Dinuclear dihydrogen complexes are rare; besides the work described here, there is only one other report on such species.96 Further, the ^ NMR resonances for the hydrogen ligands are resolved only in the present work, fast exchange precluding direct Ti measurements in the other system.96 b) Crystal structures with a located TJ2-!^ ligand are rare (sect 3.1). The crystal structure for 4, presented in chapter 3, is the first such structure for a dinuclear complex. c) There have been several previous studies23 on systems involving these T|2-H2 complexes, when their true formulation was not recognized, or a dihydrogen complex went unnoticed; thus, more detailed interpretation and, in some cases, reinterpretation of earlier 236 work is required. In the reactions of complexes of type 1 with the corresponding triarylphosphine, the demonstrated initial H2 loss23 is almost certainly the loss of the T|2-H2, i.e.: (ri2-H2)(PR3)2Ru(lt-Cl)2ai-H)RuH(PR3)2 :=" H2 + (PR3)2Ru(n-a)20i-H)RuH(PR3)2 (3.12) (PR3)2Ru(u.-a)20i-H)RuH(PR3)2 + 2PR3 » 2RuHQ(PR3)3 (3.13) The hydrogenation of 1-hexene catalyzed by la is now interpreted as occurring via the mechanism: (T|2-H2)(PPh3)2Ru(n-Cl)2(li-H)RuH(PPh3)2 + hexene —^—- (PPh3)2Ru(^-a)2(^-H)RuH(PPh3)2 + hexane (4.10) (PPh3)2Ru(n-Cl)2(^-H)RuH(PPh3)2 + H2 =~ (Ti2-H2)(PPh3)2Ru(n-Cl)2(^-H)RuH(PPh3)2 (4.11) Wang's kinetic studies on the hydrogenation of acrylamide catalyzed by la in DMA78 have yet to be re-interpreted based on the re-formulation of la given in the present work. d) It is possible that complex la is involved in hydrogenations catalyzed by the much studied2,3-66-68 complex, RuHCl(PPh3)3, which remains one of the most active catalysts known for the hydrogenation of terminal olefins.66"68 Thus, a rein terpre tation of some unpublished data obtained by Hui103and Markham72 on hydrogenation catalyzed by RuHClOPPh^ needs to be carried out Reactions of Rua2(PPh3)(PPFA), 3a, and Rua2(PPh3)(isoPFA), 3b, with H2 have been further studied, in connection with earlier mechanistic studies on the hydrogenation of organic substrates catalyzed by complex 3a.25 The complex 3a reacts with 2-8 atm H2 in n-butanol to give ruthenium phosphine products including la and a species A, possibly of the structure shown below, the species A is formed reversibly from 3a as follows: H RuCl2(PPh3)(PPFA) + H2 BuOH Ph3R 3a A (4.20) The complex 3b reacts with H2 in methanol/benzene to give 4, as mentioned above, as well as a number of unidentified hydrides; in DMA, the reaction of 3b with H2 gives la, 4, 7, RuHCl(PPh3)3 and other unidentified ruthenium phosphine complexes. The compound H2NMe2+Cl_ was another product isolated from the methanol/benzene reaction mixture, and, together with the formation of complex A from the 3a reaction with H2, provides evidence that the arnine functionality of the P-N ligands is involved in the promotion of the heterolytic cleavage of dihydrogen to give a proton and a hydride 0H2 determine if such a reaction might occur.25 Another goal25 was to deterrriine the accessibility of the Ru(TV) state during reactions of Ru(H) with H2, and the demonstrated propensity of complexes 3 for rj2-H2 complex formation suggests the possible participation of Ru(H)T|2-H2 complexes in catalytic hydrogenation. Kinetic studies on the hydrogenation of 1-hexene catalyzed by 3a,25 and by 3b in the present work, are now interpreted according to the mechanism: H+ + H-); in fact, one of the goals of the original P-N ligand work was to Ru(H)H(PPh3) Ru(H)H + PPh3 (4.21) Ki Ru(U)H + H2 Ru(IT)H(ri2-H2) K2 hexene Ru(n)(hexyl) k2 H2 hexene Ru(H)H + hexane (4.22) Reactions involving 3b and methanol have also been studied, and 3b is also active for the transfer hydrogenation (from methanol) of ketones and activated olefins. The reaction of 3b with methanol in the absence of base is proposed to occur with the stoichiometry: RuCl2(PPh3)(isoPFA) + 2MeOH H2NMe2+Cl- + H2 + 3 b RuHCl(CO)(PPh3)(isoPOF), 5 (5.25) via the following steps: RuCl2(PPh3)(isoPFA) + MeOH RuHCl(CO)(PPh3)(isoPFA), 6 3 b + H2 + HCI (5.54) RuHCl(CO)(PPh3)(isoPFA) + MeOH RuHCl(CO)(PPh3)(isoPOF), 5 + HNMe2 (5.55) HCI + HNMe2 H2NMe2+Cl- (5.56A mechanism for reaction 5.54 is presented (Fig. 5.47) and invokes reversible attack by MeOH, replacing CI", followed by reversible deprotonation of coordinated MeOH to give successively methoxo, formaldehyde and formyl mterrnediates, for which there is precedent in the literature,90,147-153 and finally the hydrido-carbonyl, 6. 239 The reaction of 3b with methanol in the presence of KOH is proposed to occur according to the stoichiometry: RuCl2(PPh3)(isoPFA) + KOH + CH3OH RuHCl(CO)(PPh3)(isoPFA) + KCI + H2 + H20 (5.32) and two pathways have been identified, one base-independent, identical to that proposed for reaction 5.54 (Fig. 5.47), and one showing a second-order dependence on KOH (Fig. 5.46). The latter pathway invokes initial reversible attack on RuCl2(PPh3)(isoPFA), 3b, by MeO", replacing Cl" to give RuCl(OMe)(PPh3)(isoPFA), and subsequent reversible replacement of PPh3 by OH', followed by concerted loss of OH' and hydride transfer from coordinated OMe' to give a hydrido-formaldehyde complex RuHCl(rj2-CH20)(isoPFA). A subsequently formed formyl intermediate reacts via intramolecular hydride transfer from the formyl to the metal, H2 loss and phosphine coordination to give the hydrido-carbonyl 6. The complex, RuHCl(PPh3)(isoPFA), 7, as well as 6, is formed when Proton Sponge® rather than KOH is used as base; the Proton Sponge® reaction is tentatively accounted for by the stoichiometry: PS RuCl2(PPh3)(isoPFA) + MeOH —RuHCl(PPh3)(isoPFA) + CH20 + HC1 (5.29) PS RuCl2(PPh3)(isoPFA) + MeOH —RuHCl(CO)(PPh3)(isoPFA) + H2 + HC1 (5.30) Both reactions could occur via the pathway suggested for reaction 5.54, with Proton Sponge® promoting the deprotonation of coordinated MeOH. Unlike OH', Proton Sponge® does not coordinate to Ru to displace PPh3; thus, in the PS reaction, an intermediate RuHCl(Ti2-CH20)(PPh3)(isoPFA) (Fig. 5.47) dissociates CH20 to give 7. Thus, both KOH and Proton Sponge® act as proton acceptors, but differ in that, while OH' displaces coordinated PPh3 (Fig. 5.46), Proton Sponge® does not In summary, the roles of KOH in reactions of 3b win. methanolic KOH solution are: i) deprotonation of MeOH by OH- to give H20, leading to methoxide attack on 3b, ii) deprotonation of cc>OTdinated MeOH to give H20 and a methoxo complex (Fig. 5.47), iii) supplying OH* for the displacement of PPh3 from an intermediate in the reaction (Fig. 5.46), and iv) supplying K+ for the formation of KCI, which is proven, perhaps for the first time, to be formed in a reaction involving a platinum metal chloro complex and alcoholic KOH. The decarbonylation of methanol or other alcohols must involve several steps, even merely to break the four bonds associated with the CO group. In general, using platinum metal complexes, if a chloride is to be dissociated, and perhaps a phosphine as well, and also if two hydrogen atoms are to be transferred to a substrate, or alternately H2 is eliminated, the total number of steps required for a mechanistic proposal becomes imposing. It is therefore desirable that the solution behaviour of the starting complex be simple, and RuCl2(PPh3)(isoPFA), 3b, satisfies this requirement, as it is undissociated in solution, although dissociation is observed for analogous RuCl2(PPh3)(P-P) complexes.169 241 Bibliography 1 Fieser, M.; Fieser, LJF. "Reagents for Organic Synthesis," Wiley, New York, vol. 1 - 12, 1967-. lb Collman, J.P.; Hegedus, L.S.; Norton, J.R.; Finke, R.G. "Principles and AppUcations of Organotransirion Metal Chemistry", University Science Books, Mill Valley, CA, 1987. lc James, B.R. Chem. Can. 1978,30, 13. 2 James, B.R. Adv. Organomet. Chem. 1979,17, 319. 3 James, B.R. in "Comprehensive Organometallic Chemistry," vol. 8. Wilkinson, G., Stone, F.P.A., Abel, E.W., eds. Pergamon, Oxford, 1982. Chap. 51. 4 James, B.R. "Homogeneous Hydrogenation," J. Wiley and Sons, New York, 1973. 5 Ozin, G.A.; Garcia-Prieto, J. /. Am. Chem. Soc. 1986,108, 3099. 6 Eisenschmid, T.C.; Kirss, R.V.; Deutsch, P.P.; Hommeltoft, S.I.; Eisenberg, R. J. Am. Chem Soc. 1987,109, 8089. 7 Eisenberg, R.; Kirss, R.U.; Eisenschmid, T. 3rd North American Chem. Congress, Toronto, 1988. Abstract No. 294. 7b Kirss, R.U.; Eisenschmid, T.C.; Eisenberg, R. /. Am. Chem. Soc. 1988,110, 8564. 8 Butler, I.R.; Cullen, W.R.; Kim, T.-J. Synth. React. Inorg. Met.-Org. Chem. 1985,15, 109. 9 Hayashi, T.; Mise, T.; Fukushima, M.; Kagotani, M.; Nagashima, N.; Hamada, Y.; Matsumoto, A.; Kawakami, S.; Konishi, M.; Yamamoto, K.; Kumada, M. Bull. Chem. Soc. Jpn. 1980,53, 1138. 242 10 Yamamoto, K.; Wakatsuki, J.; Sugimoto, R. Bull. Chem. Soc. Jpn. 1980,53, 1132. 11 Bishop, J.J.; Davison, A.; Katcher, MJL; Lichtenberg, D.W.; Merrill, R.E.; Smart, J.C. J. Organomet. Chem. 1971,27, 241. 12 Cullen, W.R.; Kim, T.-J.; Einstein, F.W.B.; Jones, T. Organometallics 1983,2, 714. 13 McCulloch, B.; Ward, DX..; Woolins, J.D.; Brubaker, Jr., CH. Organometallics 1985, 4, 1425. 14 Aratani, T.; Gonda, T.; Nuzaki, H. Tetrahedron Lett. 1969,2265. 15 Marquarding, D.; Kusacek, H.; Gokel, G.; Hoffmann, P.; Ugi, I. /. Am. Chem. Soc. 1970, 92, 5389. 16 Gokel, G.; Ugi, I. /. Chem. Educ. 1972,49,294. 17 Cullen, W.R.; Woolins, J.D. Coord. Chem. Rev. 1981, 39, 1. 18 Appleton, T.G.; Cullen, W.R.; Evans, S.V.; Kim, T.-J.; Trotter, J. J. Organomet. Chem. 1985,279,5. 19 Chalk, A.J.; Halpern, J. /. Am. Chem. Soc. 1959, 81, 5846. 20 Kim, T.-J. Ph.D. Thesis, University of British Columbia, Vancouver, 1984. 21 Dekleva, T.W. Ph.D. Thesis, University of British Columbia, Vancouver, 1983. 22 Cullen, W.R.; Einstein, F.W.B.; Huang, C.-H.; Willis, A.C.; Yeh, E.-S. /. Am. Chem. Soc. 1980,102, 988. 23 Dekleva, T.W.; Thorburn, I.S.; James, B.R. Inorg. Chim. Acta 1985,100,49. 24 James, B.R.; Thompson, L.K.; Wang, D.K.W. Inorg. Chim. Acta 1978,29, L237. 25 Rodgers, G.E.; Cullen, W.R.; James, B.R. Can. J. Chem. 1984, 61, 1314. 26 Hoffman, P.R.; Caulton, K.G. /. Am. Chem. Soc. 1975, 97, 4221. 27 Stephenson, T.A.; Wilkinson, G. /. Inorg. Nucl. Chem. 1966,28, 945. 243 28 Charland, J.-P. Personal communication. 29 Kubas, G.J. Acc. Chem. Res. 1988, 21, 120. 30 Ashworth, T.V.; Singleton, E. /. Chem. Soc, Chem. Commun. 1976, 705. 31 Kubas, G.J. /. Chem. Soc, Chem. Commun. 1980, 61. 32 Kubas, G.J.; Ryan, R.R.; Verganuni, P.J.; Wasserman, H. 185th Am. Chem. Soc. National Meeting, Seattle, 1983. Abstract INOR 229. Cited in reference 21. 33 Chem. Eng. News, 1983, 61(13), 4. 34 Crabtree, R.H.; Lavin, M, /. Chem. Soc, Chem. Commun. 1985, 794. 35 Morris, R.H.; Sawyer, J.F.; Shiralian, M.; Subkowski, J.D. J.Am. Chem. Soc. 1985,707, 5581. 36 Hamilton, D.G.; Crabtree, R.H. /. Am. Chem. Soc 1988,110, 4126. 37 Bautista, M.T.; Earl, K.A.; Maltby, P.A.; Morris, R.H. /. Am. Chem. Soc. 1988, 110, 4056. 38 Crabtree, R.H.; Hamilton, D.G. /. Am. Chem. Soc. 1986,108, 3124. 39 Henderson, R.A. /. Chem. Soc, Chem. Commun. 1987, 1670. 40 Jackson, S.A.; Upmacis, R.K.; Poliakoff, M.; Turner, J.J.; Burdett, J.K.; Grevels, F.-W. /. Chem. Soc, Chem. Commun. 1987, 678. 41 Bianchini, C; Mealli, C; Peruzzini, M.; Zanobini, F. J. Am. Chem. Soc 1987, 109, 5548. 42 Sweany, R.L.; Russell, F.N. submitted, quoted in reference 29. 43 Sweany, R.L.; Polito, MA. 194th Am. Chem. Soc. National Meeting, New Orleans, LA; Am. Chem. Soc., Washington, D.C, 1987; quoted in reference 29. 44 Conroy-Lewis, F.M.; Simpson, SJ. /. Chem. Soc, Chem. Commun. 1987, 1675. 45 Hay, P.J. /. Am. Chem. Soc. 1987, 109, 705. 244 46 Jean, Y.; Eisenstein, O.; Volatron, F.; Maouche, B.; Sefta, F. /. Am. Chem. Soc. 1986,108, 6587. 47 Burdett, J.K.; Phillips, J.R.; Pourian, M.R.; Poliakoff, M.; Turner, J.J.; Upmacis, R. Inorg. Chem. 1987, 26, 3054. 48 Hampton, C; Cullen, W.R.; James, B.R.; Charland, J.-P. /. Am. Chem. Soc. mi, 110, 6918. 49 Kubas, G.J.; Ryan, R.R.; Swanson, B.I.; Vergamini, P.J.; Wasserman, H.J. /. Am. Chem. Soc. 1984,106, 451. 50 Vergamini, P.J.; Wasserman, HJ.; Koetzle, T.; Kubas, G.J., manuscript in preparation, quoted in reference 51. 51 Kubas, G.J.; Unkefer, C.J.; Swanson, B.I.; Fukushima, E. J. Am. Chem. Soc. 1986,108, 7000. 52 Bautista, M.T.; Earl, K.A.; Maltby, P.A.; Morris, R.H.; Shweitzer, C.T.; Sella, A.J. Am. Chem. Soc. 1988,110, 7031. 53 Ricci, J.S.; Koetzle, T.F.; Bautista, M.T.; Hofstede, T; Morris, R.H.; Sawyer, J.F. In preparation, quoted in reference 52.. 54 Bianchini, C; Mealli, C; Peruzzini, M.; Zanobini, F. J. Am. Chem. Soc. 1987, 109, 5548. 55 Simpson, J. Personal communication. 56 Hampton, C; Dekleva, T.W.; James, B.R.; Cullen, W.R. Inorg. Chim. Acta 1988,145, 165. 57 Kubas, G.J.; Ryan, R.R.; Wrobleski, D.A.; /. Am. Chem. Soc. 1986,108, 1339. 58 Aresta, M.; Giannoccaro, P.; Rossi, M.; Sacco, A. Inorg. Chim. Acta 1971,5, 115. 59 Chinn, M.S.; Heinekey, D.M. /. Am. Chem. Soc. 1987,109, 5865. 245 60 Conroy-Lewis, F.M.; Simpson, S.J. /. Chem. Soc, Chem. Commun. 1987, 1675. 61 Bautista, M.; Earl, K.A.; Morris, R.H.; Sella, A. /. Am. Chem. Soc. 1987,109, 3780. . 62 Crabtree, R.H.; Lavin, M.; Bonneviot, L. /. Am. Chem. Soc. 1986,108, 4032. 63 Bautista, M.T.; Earl, K.A.; Morris, R.H. Inorg. Chem. 1988, 27, 1126. 64 Butler, I.R. Personal communication. 65 Thorburn, I.S. Personal communication. 66 Evans, D.; Osborn, J.A.; Jardine, F.A.; Wilkinson, G. Nature 1965,208, 1203. 67 Hallman, P.S.; Evans, D.; Osborn, J.A.; Wilkinson, G. /. Chem. Soc, Chem. Commun. 1967, 305. 68 Hallman, P.S.; McGarvey, B.R.; Wilkinson, G. /. Chem. Soc. (A) 1968, 3143. 69 James, B.R.; Rattray, A.D.; Wang, D.K."\ J. Chem. Soc, Chem. Commun. 1976,792. 70 Wang, D.K.W. Ph.D. Thesis, University of British Columbia, Vancouver, 1978. 71 Thorburn, I.S. M. Sc. Thesis, University of British Columbia, Vancouver, 1980. 72 Markham, L.D. Ph. D. Thesis, University of British Columbia, Vancouver, 1973. 73 James, B.R.; Markham, L.D. Inorg. Chem. 1974,13, 97. 74 Hoffman, P.R.; Caulton, K.G. /. Am. Chem. Soc 1975, 97, 4221. 75 Caulton, K.G. /. Am. Chem. Soc 1974,96, 3005. 76 Sasson, Y.; Blum, J. /. Org. Chem. 1975,40, 1887. 77 Poddar, R.K.; Agarwala, U. Ind. J. Chem. 1971, 9, 1971; /. Inorg. Nucl. Chem. 1973,55, 567. 78 James, B.R.; Wang, D.K.W. /. Chem. Soc, Chem. Commun. 1977, 550. 79 Old, M. "Applications of Dynamic NMR Spectroscopy to Organic Chemistry", vol. 4 of the series, "Methods in Stereochemical Analysis", Marchand, A.P., ed. VCH 246 Publishers, Deerfield Beach, FA, 1985. p. 407. The Eyring equation is given in the form: ln-^- = In(K-^p) - AHtj^r + , where k is the rate constant, Kis the transmission coefficient, usually taken as unity, kg is the Boltzmann constant, 1.381 x lO23 J K"1, h is Planck's constant, 6.626 x lO"34 Js, Ris the gas constant, 8.314 J moHK-1 and T is absolute temperature. 79b Pople, J.A.; Schneider, W.G.; Bernstein, HJ. "High Resolution NMR", McGraw-Hill, New York, 1959. 80 Morris, R.H. Personal communication. 81 For RuHCl(PPh3)3, from reference 68, 0CDCI3 (25 °C) -17.44 (q, 2JPH = 26 Hz). From reference 21, 0CDCI3 (30 °C) -17.78 (q, 2JPH = 26 Hz). 82 For RuHCl(P(p-tol)3)3, from reference 21, JH NMR OCDCI3(30 °C) -17.78 (q, 2jPH = 26 Hz), 3lp{ lH} NMR OCD2C12(30 °C) 55.5. From references 21,23, *H NMR 8C6D6(30 °C) -16.84. 83 Dekleva has noted a similar spectrum for a species he refers to as an "ABCD" complex. From reference 21, the 31P{ *H} NMR spectrum for this complex was 5DMA/toluene-d8(-60 °C) 58.9 (d, JPP = 22.5 Hz), 51.9 (d, JPP = 30 Hz), 41.7 (dd), 19.8 (dd). 84 Yawney, D.B.W.; Doedens, RJ. Inorg. Chem. 1972,11, 838. 85 Mason, R.; Thomas, K.M.; Gill, D.F.; Shaw, B.L. /. Organometal. Chem. 1972, 40, C67. 86 Bennett, M.J.; Caulton, K.G.; Cotton, F.A. Inorg. Chem. 1969, 8,1. 87 Mattson, B.M.; Heiman, J.R.; Pignolet, L.H. Inorg. Chem. 1976,15, 564. 88 Dahl, L.F.; Rodulfo de Gil, E.; Feltham, RX>. /. Am. Chem. Soc. 1969,91, 1653. 89 Kaesz, H.D.; Saillant, R.B. Chem. Rev. 1972, 72, 231. 247 90 Chaudret, B.N.; Cole-Hamilton, D.J.; Nohr, R.S.; Wilkinson, G. /. Chem. Soc, Dalton Trans. 1977,1546. 91 Cole-Hamilton, D.J.; Wilkinson, G. Nouv. J. Chim. 1977,1, 142. 92 Heinekey, D.M.; Payne, N.G.; Schulte, G.K. /. Am. Chem. Soc 1988,110, 2303. 93 Antinolo, A.; Chaudret, B.; Commenges, G.; Fajardo, M.; Jalon, F.; Morris, R.H.; Otero, A.; Schwetzer, CT. /. Chem. Soc, Chem. Commun. 1988, 1210. 94 Chaudret, B.; Commenges, G.; Jalon, F.; Otero, A. /. Chem. Soc, Chem. Commun. 1989, 210. 95 Suzuki, H.; Omori, H.; Lee, D.H.; Yoshida, Y.; Moro-oka, Y. Proc. 6th Intern. Symp. on Homog. Catal., Vancouver, Canada, 1988. P-115. 95b A re-interpretation of the crystallographic data for this complex shows that all four bridging hydride ligands are, in fact, classical hydrides (see Marsh, R.E. Organometallics 1989, 8, 1583). 96 Arliguie, T; Chaudret, B.; Morris, R.H.; Sella, A. Inorg. Chem. 1988,27, 598. 97 Hauge, R.H.; Kafafi, Z.H.; Margrave, J.L. Proceedings, The Physics and Chemistry of Small Clusters, Richmond, VA, Oct 1986. quoted in ref. 29. 98 Joshi, A. Personal communication. 99 Calvin, M. /. Chem. Soc, Trans. Faraday Soc. 1938,34,1181. Quoted in reference 3. 100 Roelen, O. U.S.Pat. 2 327 066 (1943). Chem. Abstr. 1944,38, 550. Quoted in references 3,101. 101 Halpern, J. /. Organomet. Chem. 1980,200, 133. 102 Halpern, J.; Harrod, J.F.; James, B.R. /. Am. Chem. Soc. 1961, 83, 753. 103 Hui, B. C-Y. Ph. D. Thesis, University of British Columbia, Vancouver, 1969. 104 Strathdee, G.; Given, R. Can. J. Chem. 1975,53, 106. 105 Stolzenberg, A.M.; Muetterties, E.L. Organometallics 1985,4, 1739. 106 Mestroni, G.; Zassinovich, G.; Camus, A. /. Organomet. Chem. 1977,140, 63. 107 James, B.R.; Mahajan, D. Can. J. Chem. 1979,57, 180. 108 Butler, I.; Hampton, C; Cullen, W.R.; James, B.R.; Charland, J.-P. To be published. 109 LaPlaca, S.J.; Ibers, J.A. Inorg. Chem. 1965,4, 778. 110 Fryzuk, M.D.; MacNeil, P.A. Organometallics 1983,2, 682. 111 Charman, H.B.; Nature 1966,212, 278. 112 Charman, H.B. J. Chem. Soc, B 1967, 629. 113 Dobson, A.; Robinson, S.D. Inorg. Chem. 1977,16, 137. 114 Shinoda, S.; Itagaki, FL; Saito, Y. /. Chem. Soc, Chem. Commun. 1985, 860. 115 Yamamoto, H.; Shinoda, S.; Saito, Y. /. Mol. Catal. 1985,30,259. 116 Doughty, D.H.; Pignolet, L.H., in "Homogeneous Catalysis with Metal Phosphine Complexes" Pignolet, L.H., ed. Plenum Press, New York, 1983. Chap. 11. 117 Chatt, J.; Shaw, B.L.; Field, A.E. /. Chem. Soc. 1964, 3466. 118 Walker, J.F. "Formaldehyde," Reinhold, New York, 1964. 119 Noller, C.R. "Chemistry of Organic Compounds," 2nd ed. W.B. Saunders, London, 1957. p. 169. 120 Mannich, C; Geilman, W. Ber. 1916,49, 586. 121 Takahashi, T.; Shinoda, S.; Saito, Y. /. Mol. Catal. 1985,31, 301. 122 Smith, T.A.; Aplin, R.P.; Maitlis, P.M /. Organomet. Chem. 1985,291, C13. 123 Morton, D.; Cole-Hamilton, D.J. /. Chem. Soc, Chem. Commun. 1988, 1154. 124 Linn, D.E.; Halpern, J. /. Am. Chem. Soc 1987,109, 2969. 125 Bailar, J.C; Itatani, H. /. Am. Chem. Soc 1967, 89, 1592. Ibid., p. 1600. 126 Milstein, D. /. Mol. Catal. 1986,36, 387. 127 Smith, T.A.; Maitlis, P.M. /. Organomet. Chem. 1984,269, C7, and references therein. 128 Smith, T.A.; Maitlis, P.M. /. Organomet. Chem. 1985,259, 385. 129 Imai, R; Nishiguchi, T.; Fukuzumi, K. /. Org. Chem. 1976,41, 665. 249 130 Chaudret, B.; Poilblanc, R. /. Organomet. Chem. 1979,174, C51. 131 Speier, G.; Marko, L. /. Organomet. Chem. 1981,210, 253. 132 Speier, G.; Marko, L. "Proc. Int. Conf. Coord. Chem., 16th" Dublin, Ire. 1974, 4.37. 133 Thorburn, I.S. Ph.D. Thesis, University of British Columbia, Vancouver, 1985. 134 Beaupere, D.; Nadjo, L.; Uzan, R. /. Mol. Catal. 1983,20, 195, and references therein. 135 Beaupere, D.; Massoui, M.; Rdamirina, R.; Uzan, R. Nouv. J. Chim. 1986,10, 493. 136 Massoui, M.; Beaupere, D.; Goethals, G.; Uzan, R. 1985,33, 209. 137 Beaupere, D.; Nadjo, L.; Uzan, R.; Bauer, P. /. Mol. Catal. 1983,20, 185. 138 Beaupere, D.; Nadjo, L.; Uzan, R.; Bauer, P. /. Mol. Catal. 1982,14, 129. 139 Ohki, A.; Nishiguchi, T.; Fukuzumi, K. Tetrahedron 1979,35, 1737. 140 Ohki, A.; Nishiguchi, T.; Fukuzumi, K. /. Org. Chem. 1979,44, 766. 141 Beaupere, D.; Bauer, P.; Nadjo, L.; Uzan, R. /. Organomet. Chem. 1982,238, C12. 142 Spogliarich, R.; Zassinovich, G.; Mestroni, G.; Graziani, M. /. Organomet. Chem. 1979,179, C45. 143 Kvintovics, P.; James, B.R.; Heil, B. /. Chem. Soc., Chem. Commun. 1986, 1810. 144 Sasson, Y.; Blum, J. J. Org. Chem. 1975,40, 1887. 145 Itagaki, H.; Saito, Y. J. Mol. Catal. 1987,41, 209. 146 Dobson, A.; Moore, D.S.; Robinson, S.D.; Hawthorne, M.B.; New, L. Polyhedron 1985,4, 1119. 147 Brown, K.L.; Clark, G.R.; Headford, C.E.L.; Marsden, K.; Roper, W.R. /. Am. Chem. Soc. 1979, 101, 503. 250 148 Collman, J.P.; Winter, S.R. /. Am. Chem. Soc. 1973, 95, 4089. 149 Nagao, H.; Aohagi, K.; Yukawa, Y.; Howell, F.S.; Mukada, M.; Kakihana, H. Bull. Chem. Soc. Jpn. 1987, 60, 3247. 150 Stasunik, A.; Malisch, W. J. Organomet. Chem. 1984,270, G56. 151 Gillard, R.D.; Kaue-Maguire, L.A.P.; Williams, P.A. /. Chem. Soc, Dalton Trans. 1977, 1039. 152 Cole-Hamilton, D.J.; Stephenson, T.A. /. Chem. Soc, Dalton Trans. 1976, 2396. 153 Hayashi, Y.; Komiya, S.; Yamamoto, T.; Yamamoto, A. Chem. Lett. 1984, (8), 1363. 154 Jones, R.; Einstein, F.; Hampton, C; Cullen, W.; James, B. To be published. 155 Dombek, B.D. Ann. N.Y. Acad. Sci. 1983, 415, 176. 156 Chatt, J.; Shaw, B.L.; Field, A.E. /. Chem. Soc. 1964, 3466. 157 Benedetti, E.; Bracci, G.; Shrana, G.; Solvetti, F.; Grassi, B. /. Organomet. Chem. 1972,57, 361. 158 Bennett, M.; Milner, D.L.; /. Am. Chem. Soc. 1969, 91, 6983. 159 Parshall, G.W.; Knoth, W.H.; Schunn, R.A. /. Am. Chem. Soc. 1969, 91, 4990. 160 Cole-Hamilton, D.J.; Wilkinson, G. J. Chem. Soc, Dalton. Trans. 1977, 797. 161 Hata, G.; Kondo, H.; Miyake, A. /. Am. Chem. Soc 1968,90, 2278. 162 Ainscough, E.W.; Robinson, S.D. J. Chem. Soc. (D) 1970, 863. 163 Valentine, J.S. /. Chem. Soc, Chem. Commun. 1973, 857. 164 Gosser, L.W. Inorg. Chem. 1975,14, 1453. 165 Suess-Fink, G.; Reiner, J. /. Mol. Catal. 1982,16, 231. 166 Skapski, A.C.; Troughton, P.G.H. /. Chem. Soc, Chem. Commun. 1968, 1230. 167 Tarasov, B.P.; Vergunova, N.G.; Belovodskii, VP.; Blazhin, Yu.M.; Ogorodnikov, S.K. Zh. Prikl. Khim. 1986, 59, 1795. 251 168 Pavelek, Z. Chem. Papers 1988,42, 299. 169 James, B.R.; Pacheco, A.; Rettig, S.J.; Thorburn, IS. /. Mol. Catal. 1987,41, 147. 

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 9 0
Germany 5 1
China 5 16
Ukraine 3 0
France 2 0
Ireland 2 0
Czech Republic 1 0
Poland 1 0
Zimbabwe 1 0
Japan 1 0
City Views Downloads
Unknown 13 1
Ashburn 4 0
Beijing 3 0
Changchun 2 0
Dublin 2 0
Federal Way 1 0
Wurzburg 1 0
Harare 1 0
Mountain View 1 0
University Park 1 0
Tokyo 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059788/manifest

Comment

Related Items