Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Activation of molecular hydrogen and olefins in solution by triphenylphosphine complexes of bivalent… Markham, Larry Derwood 1973

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-UBC_1973_A1 M37.pdf [ 7.74MB ]
JSON: 831-1.0060155.json
JSON-LD: 831-1.0060155-ld.json
RDF/XML (Pretty): 831-1.0060155-rdf.xml
RDF/JSON: 831-1.0060155-rdf.json
Turtle: 831-1.0060155-turtle.txt
N-Triples: 831-1.0060155-rdf-ntriples.txt
Original Record: 831-1.0060155-source.json
Full Text

Full Text

15134 ACTIVATION OF MOLECULAR HYDROGEN AND OLEFINS IN SOLUTION BY TRIPHENYLPHOSPHINE COMPLEXES OF BIVALENT PUTHENIUM BY LARRY DERWOOD MARKHAM B.Sc. (Hons.) The University of Western Ontario, 1969 A THESIS SUBMITTED IN PARTIAL FULFILMENT 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 February, 1973 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 i t 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 representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Chemistry The University of British Columbia Vancouver 8, Canada 20 March, 1973 - i i -ABSTRACT Kinetic and equilibrium studies involving solutions of dichlorotris-(triphenylphosphine)ruthenium(II) and the corresponding hydridochloro complex are described, especially reactions involving molecular hydrogen and olefins. In benzene or in dimethylacetamide (DMA) solution, the dichloro complex dissociates with loss of a triphenylphosphine molecule, RuCl 2(PPh 3) 3 RuCl 2(PPh 3) 2 + PPh3 (1) and in DMA solution, further dissociation of chloride ion from the bisphosphine complex occurs: RuCl 2(PPh 3) 2 — ^ RuCl(PPh 3) 2 + + Cl" (2) DMA solutions of RuCl 2(PPh 3) 3 react rapidly and reversibly with molecular hydrogen at room temperature, for example, RuCl 2(PPh 3) 3 + H2 RuClH(PPh3)3 + HC1 (3) This hydrogenolysis reaction does not occur in benzene solution, but the basic amide solvent promotes hydride formation by effectively stabilizing the released HC1. The relative reactivity toward H2 of the species present in solution is as follows: RuCl(PPh 3) 2 + > RuCl 2(PPh 3) 2 > RuCl2(PPh3)3» For the reverse reaction between hydride complex and HC1, the reactivity order of species present is RuClH(PPh3)2 > RuClH(PPh3)3. - i i i -Thermodynamic and kinetic data are given for equilibria such as (3), and thermodynamic data are presented for equilibria (1) and (2). The hydride complex RuClH(PPh3)3 is an extremely effective catalyst for the homogeneous hydrogenation of olefins in DMA solution at 35°. Unfavorable steric and electronic factors in the olefin both play a major role in reducing the rate of hydrogenation; these effects can be correlated with the proposed reaction mechanism, which involves a predissociation of the catalyst and the formation of a a-alkyl intermediate via a hydrido-olefin species: RuClH(PPh3)3 ^ > RuClH(PPh3)2 + PPh3 (4) RuClH(PPh3)2 + olefin RuClH(PPh3)2(olefin) -—*• RuCl(alkyl) (PPh 3) 2 (5) RuCl(alkyl) (PPh 3) 2 + H 2 —>• RuClH(PPh3)2 + alkane (6) Equilibrium constants for reaction (5) with a variety of olefins, are presented together with rate data for reaction (6). Limited studies have been carried out using the corresponding hydridobromo and hydrido-acetate complexes. The hydride complex RuClH(PPh3)3> isolated as a DMA solvate, is also effective as a catalyst for the polymerization of both ethylene and butadiene in DMA solution. The kinetics of these reactions have been studied and analyzed in terms of a mechanism that involves i n i t i a l formation of a o-alkyl complex, and propagation via insertion of coordinated olefin into the Ru-carbon bond, for example, - iv -Ru-H + C„H. — - Ru-C0HC (7) 2 4 — 2 5 Ru-C2H5 + C2H4 (C2H4)Ru-C2H5 (8) (C2H4)Ru-C2H5 > Ru-C2H4-C2H5 (9) The ethylene and butadiene systems show different kinetic dependences which are accounted for by the stronger complexing of the diene. - v -TABLE OF CONTENTS Page ABSTRACT i i TABLE OF CONTENTS v LIST OF TABLES ix LIST OF FIGURES x i i ABBREVIATIONS xvi ACKNOWLEDGEMENTS xix CHAPTER 1. INTRODUCTION 1 1.1 Aim of Work 1 1.2 Activation of Molecular Hydrogen and Homogeneous Hydrogenation of Olefinic Compounds 3 1.3 Olefin Polymerization by Coordination Catalysis .. 5 CHAPTER II. APPARATUS AND EXPERIMENTAL PROCEDURE 8 2.1 Materials 8 2.1.1 Ruthenium Compounds 8 2.1.2 Gases 9 2.1.3 Olefinic Substrates 9 2.1.4 Other Materials 10 2.2 Apparatus for Constant Pressure Gas-Uptake Measurements 10 2.3 Procedure for a Typical Gas-Uptake Experiment .... 12 2.4 Gas Solubility Measurements 14 2.5 Spectrophotometric Kinetic Measurements 14 2.6 Reaction Product Analysis 15 - vi -Page 2.6.1 Gaseous Products 15 2.6.2 Liquid and Solid Products 16 2.7 Instrumentation 16 CHAPTER III. THE NATURE OF DICHLOROTRIS(TRIPHENYLPHOSPHINE)-RUTHENIUM(II) IN SOLUTION 18 3.1 Introduction 18 3.2 RuCl 2(PPh 3) 3 in Benzene Solution 19 3.3 RuCl 2(PPh 3) 3 in Dimethylacetamide Solution 25 3.4 Discussion 34 CHAPTER IV. THE EQUILIBRIUM REACTION OF DICHLOROTRIS(TRIPHENYL-PHOSPHINE)RUTHENIUM(II) WITH MOLECULAR HYDROGEN IN DMA. 44 4.1 Introduction 44 4.2 Reaction of RuCl2(PPh3>3 with Hydrogen in DMA 46 4.2.1 Preparation of the DMA Solvate RuClH(PPh3)3-DMA 47 4.2.2 Kinetics of the Reaction of RuCl2(PPh3>3 with Hydrogen in DMA 5 0 4.3 Reaction of RuClH(PPh3)3 with HC1 in DMA 62 4.4 The Overall Equilibrium 6 7 4.5 Discussion 71 4.5.1 The Effect of Excess Triphenylphosphine on the Forward Reaction 71 4.5.2 Reaction of RuCl(PPh 3) 2 + with Hydrogen .... 77 4.5.3 Reaction of RuCl 2(PPh 3) 2 with Hydrogen .... 82 4.5.4 Reaction of RuClH(PPh3>3 with HC1 8 8 4.5.5 The Overall Equilibrium 95 - v i i -Page CHAPTER V. HOMOGENEOUS HYDROGENATION OF OLEFINS USING HYDRIDO-CHLOROTRIS(TRIPHENYLPHOSPHINE)RUTHENIUM(II) AS CATALYST IN DMA 99 5.1 Introduction 99 5.2 Catalytic Hydrogenation of Olefins 100 5.2.1 Hydrogenation of Maleic Acid, Fumaric Acid, and Similar Olefins 102 5.2.2 Hydrogenation of Terminal Olefins 108 5.2.3 Hydrogenation of Cyclic Olefins, and trans-2-octene 112 5.2.4 Hydrogenation of Terminal Olefins having Functional Groups H6 5.2.5 Hydrogenation of Ethylene and Butadiene ... 122 5.3 Reaction of RuClH(PPh3)3 with Acetylene 123 5.4 Catalytic Hydrogenation using Other Ruthenium(II) Complexes 124 5.5 Discussion 124 5.5.1 The Reaction Kinetics 129 5.5.2 Other Ruthenium(II) Complexes 151 CHAPTER VI. HOMOGENEOUS POLYMERIZATIONS OF ETHYLENE, AND BUTADIENE, CATALYZED BY HYDRID0CHL0R0TRIS(TRIPHENYL-PHOSPHINE) RUTHENIUM (II) IN DMA 153 6.1 Introduction 153 6.2 Polymerization Reactions in DMA 154 - v i i i -Page 6.2.1 Ethylene Polymerization 154 6.2.2 Butadiene Polymerization 164 6.3 Reaction of RuClH(PPh3)3 with Ethylene, and Butadiene at 25° 170 6.4 Discussion 171 CHAPTER VII. GENERAL CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK 181 REFERENCES I 8 8 - ix -LIST OF TABLES Table Page The nature of RuCl 2(PPh 3) 3 in solution I Dependence of O.D.^ gg and O.D.75Q on [PPh3] in benzene solution at 25° 24 II Temperature dependence of in benzene solution.. 26 III Dependence of O.T).^^ on [LiCl] in DMA solution at 25° 3 1 IV Temperature dependence of in DMA solution 35 V Dependence of O.D.^ gQ and O.D.75Q on [PPh3] in DMA solution at 25° 3 6 VI Temperature dependence of in DMA solution 38 VII Thermodynamic parameters 1^ Equilibrium reaction of RuCl 2(PPh 3) 3 with hydrogen in DMA solution VIII Dependence of k' on temperature and [H2] 53 IX Dependence of k" on [PPh3] at 25° 56 X Dependence of k" on temperature and [H2] at high phosphine concentration 57 XI Dependence of k'" on [H^] and [LiCl] 60 XII Dependence of k 1 V on [HCl] and [PPh3] at 25° 6Z> iv c.c. XIII Temperature dependence of k 0 0 XIV Dependence of O.D.52Q on [HCl] and [H2] at 25° ... 72 XV Temperature dependence of K Ik XVI Temperature dependence of k^ 9^ XVII Temperature dependence of k^ a x 84 XVIII Dependence of [V]/[IV] on [PPh ] at 25° 91 - X -Table Page RuClH(PPh.j)3-catalyzed hydrogenation of olefins in DMA solution XIX Dependence of i n i t i a l rate on [Ru*1] and [olefin] at 35° (M.A. and similar olefins) 101 XX Dependence of i n i t i a l rate on [olefin] at 35° (F.A. and similar olefins) 106 XXI Dependence of i n i t i a l rate on temperature (M.A. and F.A.) 109 XXII Dependence of i n i t i a l rate on [H^ ] and [olefin] at 35° (terminal-olefins) HO XXIII Dependence of i n i t i a l rate on temperature (terminal olefins) H3 XXIV Dependence of i n i t i a l rate on [H,,] and [olefin] at 35° (cyclic olefins and trans-2-octene) XXV Temperature dependence of k (cyclooctene) H 8 XXVI Dependence of i n i t i a l rate on [olefin] at 35° (terminal olefins having functional groups) ...... H9 XXVII Temperature dependence of k (terminal olefins having functional groups) II XXVIII Initial rates at 35° using different Ru complexes as catalysts 125 XXIX Effect of added PPh3 on i n i t i a l rates at 35° (9 olefins) 130 XXX Values of k and K' calculated at 35° (12 olefins). 138 XXXI Activation parameters (8 olefins) 144 XXXII Temperature dependence of (for F.A.) and K' (for 1-octene) 149 - xi -Table RuClH(PPh3)3-catalyzed polymerization of ethylene and butadiene in DMA solution XXXIII Dependence of linear rate of C^B.^ polymerization on [Ru11] and [ C ^ ] at 80° XXXIV Temperature dependence of XXXV Dependence of i n i t i a l rate of C^ H^  polymerization on [Ru11] and [ C ^ ] at 65° XXXVI Temperature dependence of k^ Page 157 160 166 169 - x i i -LIST OF FIGURES Figure Page 1 Apparatus for constant pressure gas-uptake measurement 11 The nature of RuCl 2(PPh 3) 3 in solution 2 Absorption spectra of benzene solutions of RuCl 2(PPh 3) 3 at 25° 20 3 Dependence of optical density on [PPh3] in benzene and DMA' at 25° 23 4 Absorption spectrum of a DMA solution of RuCl 2(PPh 3) 3 at 25° 28 5 Absorption spectra of DMA solutions of RuCl 2(PPh 3) 3 containing LiCl at 25° 29 6 Dependence of optical density on [LiCl] in DMA at 25°.. 32 7 Inverse dependence of [III]/[II] on [LiCl] in DMA at 25° 33 8 Inverse dependence of [II]/[I] on [PPh3] in DMA at 25°. 37 9 Van't Hoff plot for (benzene solution) 39 10 Van't Hoff plot for R (DMA solution) 39 11 Van't Hoff plot for K2 (DMA solution) 40 Equilibrium reaction of RuCl 2(PPh 3) 3 with H 2 in DMA solution 12 Absorption spectra of DMA solutions of RuClH(PPh3)3 and RuBrH(PPh3)3 at 25° 49 13 First-order rate plots (forward reaction) 52 14 Dependence of k' on [H2] 54 15 Dependence of k" on [PPh3] at 25° 58 16 Dependence of k" on [Hj at 25° 58 - x i i i -Figure Page 17 First-order rate plot (forward reaction, no added chloride) at 15° 59 18 Dependence of k"' on [H ] at 10° 61 19 Dependence of k'" on [LiCl] at 25° 6 1 20 First-order rate plots (reverse reaction) at 25° 63 21 Dependence of k ± V on [HCl] at 25° 6 5 22 Dependence of k l v/[HCl] on [PPh3] at 25° 6 5 23 Absorption spectra of DMA solutions of RuCl 2(PPh 3) 3 containing H2 at 25° * 69 24 Absorption spectra of DMA solutions of RuCl 2(PPh 3) 3 containing H2 and PPh3 at 25° 7 0 25 Plot of [Ru-H]/[Ru-Cl] vs. [H2] at 25° 7 3 26 Plot of [Ru-H]/[Ru-Cl] vs. [HCl]" 1 at 25° 7 3 27 Plot of [II]/[I] vs. [PPh 3J _ 1 at 25° 7 6 28 Arrhenius plot 8 0 29 Plot of (k') _ 1 vs. [ H 2 ] _ 1 83 30 Arrhenius plot 85 31 Plot of [V]/[IV] vs. [PPh 3] _ 1 at 25° 9 2 32 Plot of log k 3K 3 vs. T _ 1 9 6 33 Van't Hoff plot 9 6 34 Van't Hoff plot 9 7 RuClH(PPh3)3~catalyzed hydrogenation of olefins in DMA solution 35 Rate plots for hydrogenation of olefins at 35° l ^ 3 36 Dependence of M.A. hydrogenation rate on [Ru11] at 35°. 107 37 Dependence of M.A. and F.A. hydrogenation rates on [olefin] at 35° I 107 - x i v -Figure Page 38 Dependence of 1-octene and 1-decene hydrogenation rates on [olefin] at 35° m 39 Dependence of 1-hexene and 1-nonene hydrogenation rates on [olefin] at 35° HI 40 Rate plot for hydrogenation of a mixture of 1-octene and 2-octene at 35° H5 41 Dependence of 2-octene and cyclohexene hydrogenation rates on [olefin] at 35° H7 42 Dependence of 4-methoxystyrene and cyclooctene hydrogenation rates on [olefin] at 35° H7 43 Plot of (initial rate) ^ vs. [olefin] ^ for M.A. and F.A. at 35° 133 44 Plot of (initial rate) ^ vs. [olefin] ^ for 1-octene at 35° 133 45 Plot of (initial rate) ^ vs. [olefin] for 1-nonene at 35 134 46 Plot of (initial rate) ^ vs. [olefin] ^ for 1-hexene and 1-decene at 35° i 3 4 47 Plot of (initial rate) vs. [olefin] "" for cyclooctene and 4-methoxystyrene at 35° 135 48 Plot of (initial rate) vs. [olefin] ^ for cyclohexene at 35° 135 49 Plot of (initial rate) vs. [olefin] ^ for acrylamide at 35° 1 3 6 50 Arrhenius plot (hydrogenation of M.A.) 1^ 1 - XV -Figure Page 51 Arrhenius plots (hydrogenations of 1-nonene, 1-hexene, cyclooctene, methyl vinyl ketone) 1^ 2 52 Arrhenius plots (hydrogenations of 1-decene, 4-methoxy-styrene, acrylamide) l ^ 3 RuClH(PPh3)3-catalyzed polymerization of ethylene and butadiene in DMA solution 53 Linear rate plots for polymerization at 80° 155 54 Dependence of C"^^ polymerization rate on I^H^] at 80° 158 55 Dependence of C^R^ polymerization rate on [Ru11] at 80° 159 56 Arrhenius plot (polymerization of C^i^ 1^1 57 Rate plot for C^ H^  polymerization at 80° 163 58 Rate, plots for C^ Hg polymerization at 80° 165 59 Dependence of C^ H^  polymerization rate on [Ru11] at 65° 167 60 Arrhenius plot (polymerization of C^ H^ ) 167 61 Rate plot for reaction of RuClH(PPh3)3 in DMA with C.H. at 25° 172 2 4 62 First-order rate plot for reaction of RuClH(PPh3)3 in DMA with C„H. at 25° 172 - xv i -ABBREVIATIONS AND COMMON NAMES The following l i s t of abbreviations and common names, most of which are commonly adopted in chemical research literature,will be employed in this thesis. All temperatures are in °C unless specifically denoted °K. atm acrylamide ally l alcohol calc cps atmosphere propenamide, CH2=CH-C0NH2 2-propen-l-ol, CH2=CH-CH20H calculated cycles per second crotonic acid diethylf umarate diethylmaleate trans-2-butenoic acid, H .CO Et \ . / Et0 2C' H „ Et0 2C^ NC0 2Et H C02H CH3 X H DMA DMF DMSO Et fumaramide F.A. N,N-dimethylacetamide, CHACON(CH3)2 dimethylformamide, HCON(CH3)2 dimethylsulfoxide, (CH^SO ethyl, ~C2H5 H CONH„ >\ H2NOC^ NH fumaric acid, trans-butenedioic acid, H02C H ,C00H C c - xvii -glc ir J L log ln M gas-liquid chromatography infrared coupling constant, cps ligand common logarithm natural logarithm molar M.A. maleic acid monoamide Me methyl vinyl ketone monoethy1f umarate monomethylfumarate nmr O . D . maleic acid, cis-butenedioic acid, ^ = C H O CO H H ^ H 1 maleamic acid, ^0=0 H O 2 C T C O N H 2 methyl, - C H 3 4-methoxystryene p-vinylanisole, C H ^ O — \ \ A—CH=CH2 methyl-fumaric acid mesaconic acid, H C02H > \ H02C N C H 3 3-butene-2-one, C H 3 " C 0 - C H = C H 2 H .C0_Et \ / R02C Nl H yC0„Me \ / 2 / C = \ E02C H nuclear magnetic resonance optical density, absorbance x v i i i -P Ph PPh, ppm R p a r t i a l pressure phenyl, -C,RC o j triphenylphosphine p a r t s per m i l l i o n a l k y l or a r y l group trans-cinnamic a c i d t r a n s - s t i l b e n e ^COOH trans-3-phenylpropenoic a c i d , ,C=C " " / \ Ph't r a n s - l , 2 - d i p h e n y l e t h y l e n e , C=C / \ Ph' X £ A v T halogen, unless s t a t e d otherwise molar e x t i n c t i o n c o e f f i c i e n t -1 2 -1 e q u i v a l e n t conductance, ohm cm mole frequency, cm ^ chemical s h i f t , ppm Note: The word . " o l e f i n " i s used to describe a l l d e r i v a t i v e s of o l e f i n s . - x i x -ACKNOWLEDGEMENTS I w i s h t o thank Dr. B.R. James f o r h i s gu i d a n c e and encouragement th r o u g h o u t t h e c o u r s e of t h i s work. F i n a n c i a l s u p p o r t o f t h i s r e s e a r c h and t h e award o f a "1967 S c i e n c e S c h o l a r s h i p " by the N a t i o n a l R e s e a r c h C o u n c i l o f Canada i s g r a t e f u l l y acknowledged. CHAPTER I INTRODUCTION 1.1 Aim of Work The first report of homogeneous catalytic hydrogenation, reduction of an organic substrate using a cupric acetate solution, appeared in 1938,1 although heterogeneous catalysts had already been studied for many years. Interest in both homogeneous and heterogeneous catalysis is currently at a high level, especially because of the possible 2 industrial applications of these types of investigations. Similarity 3 to enzymic systems is another feature of some catalytic processes which has recently served to encourage their study. Although hetero-geneous catalysts are currently more prevalent than homogeneous in industrial use, the actual mechanism of reaction is often much better understood in the case of homogeneous catalysts because of their greater susceptibility to investigation, especially by spectroscopic methods. In fact, there appear to be many similarities between homogeneous and heterogeneous reaction mechanisms, although the homo-geneous catalysts can in some cases be far more specific in their 4 activity. Reports of homogeneous catalysis by ruthenium salts and complexes appeared in 1961 and concerned studies on Ru(II) chloride complexes in - 2 -acid solution.^''' These solutions were found to be catalytically active for the hydrogenation of certain substituted ethylenes containing an activated double bond, such as maleic acid, fumaric acid, and acrylic acid. More recently, triphenylphosphine complexes of Ru(II) have been investigated, in particular the complexes RuCl2(PPh3>3 and RuClH(PPh3> . 7 The latter complex was discovered by Wilkinson's group to be extremely effective in benzene for the homo-8—10 geneous catalytic hydrogenation of terminal alkenes. However, experimental difficulties were encountered in the study of these hydrogenation reactions, so that detailed kinetics were not obtained. More recently, the hydrogenation of maleic acid using this catalyst in dimethylacetamide (DMA) solution has been thoroughly investigated. 7 The success of this study in DMA suggested the extension of the work in this solvent to other olefins, in order to investigate the effect of steric and electronic effects in the substrate on the hydrogenation rate and mechanism. Because of some uncertainty regarding the nature of these Ru(II) phosphine complexes in solution, a detailed study was first undertaken to investigate the catalyst solutions. Chapter III describes the solution studies on RuCl 2(PPh 3) 3, and Chapter IV is concerned with the species RuClH(PPh3)3. The catalytic hydrogenation of a variety of olefins is described in Chapter V. During the course of this work, the hydride complex RuClH(PPh3)3 was discovered to be effective for the catalytic polymerization of ethylene and butadiene, and these studies are described in Chapter VI. The use of ruthenium complexes for homogeneous catalysis has recently been reviewed.7 The wide range of reactions discussed - 3 -(including hydrogenation, hydrogen migration, hydration, oxidation, polymerization, arylation, alkylation, hydroformylation and decarbonyla-tion) attests to the versatility of ruthenium catalysts. In view of the importance to this thesis of hydrogenation and polymerization reactions, a short discussion of these two aspects of catalysis is included here. 1.2 Activation of Molecular Hydrogen, and Homogeneous Hydrogenation  of Olefinic Compounds For hydrogenation of olefins to occur, both the hydrogen and olefin molecules are usually considered to be first activated by the 13 catalyst. The hydrogen molecule is activated through the formation of a reactive transition metal hydride complex. As a result „of the high reactivity of these hydride intermediates, the rate-determining step in the overall hydrogenation reaction is usually the formation of the hydride. Hydrogen activation has been recognized thus far as 13—16 occurring in three ways, depending on the metal complex being used. Overall heterolytic splitting, exemplified by reaction (1.1),"^ I l l 3- III 3 - 4 -Ru CI, + H„ — - Ru HCl- + H + CI (1.1) involves substitution of a hydride for another ligand, with no change in the oxidation state of the metal. This mechanism occurs readily in the presence of a suitable base to stabilize the released proton. An olefin molecule which has become coordinated to the metal may then undergo insertion into the Ru-hydride bond, yielding a a-alkyl complex. - 4 -18 For example, a Ru(II) system has been described in terms of Equation (1.2): Ru || —>- —Ru_ I  —> — Ru — fT (1.2) / | C /| / | /\ / \ / \ Alkane release and regeneration of the Ru(II) catalyst can occur via 18 electrophilic attack by a proton at the carbon attached to the metal. In the case of activation by homolytic splitting, the oxidation state and generally the coordination number of the metal both increase 13 by one. The reaction therefore depends on the susceptibility of the metal to oxidation and the ability to expand its coordination shell. An example is the reversible uptake of by aqueous solutions of lCo(CN) 5 J 3 ~: 1 9 2[Co(CN)5]3~ + H2 2[HCo(CN)5]3 (1.3) In this particular case, an alkyl complex formed via olefin insertion may react subsequently with another hydride complex to produce the 20 saturated product: [(NC) 5Co-alkyl] 3" + [HCo(CN)5]3" —»- 2[Co(CN)5]3~ + alkane (1.4) The third method of hydrogen activation involves oxidative addition of H2 to the metal, increasing the oxidation state and coordination 21 number of the metal by two, for example: - 5 -IrCl(CO)(PPh3)2 + H2 ^ H2IrCl(CO)(PPh3)2 (1.5) An olefin coordinated to the dihydride may be reduced by consecutive 13 transfer of the two hydrogen atoms, via a a-alkyl hydride intermediate. A recent text on homogeneous hydrogenation covers the literature to 1972, and includes a survey of the reduction of both organic and 22 inorganic substrates. Less extensive reviews have appeared on 23 hydrogenation of carbon-carbon multiple bonds and selective hydro-24 genation of dienes and polyenes to monoenes. Two recent and highly significant developments in the field of homogeneous hydrogenation are the asymmetric hydrogenation of olefins accomplished using optically 25—27 28 active catalyst or solvent, and the use of catalysts supported 29 on resins rather than in a conventional homogeneous system. 1.3 Olefin Polymerization by Coordination Catalysis Polymerization of olefins using the Ziegler-Natta type of catalyst 30 31 has been studied extensively, ' although most investigations have been concerned with the nature of the product, rather than the reaction mechanism. These catalysts are formed from a transition metal compound (Group IV-VIII) and an organometallic compound of a non-transition element (Group I-III),for example TiCl^ and an aluminum trialkyl. A transition metal alkyl complex is thought to be the active catalytic species. The olefin is thought to coordinate to a vacant site on the metal complex, and to become inserted into the growing polymer chain R 32 which is attached to an adjacent position on the metal ion: - 6 -R 1/ C 1/ ^ C M II > — M — C T \ (1.6) / | C / I / x ' /\ ' Chain termination can occur via a reverse insertion reaction, with formation of a metal hydride: H R H \ | / R H ' _ R H . C H \ / = V 1/ I 1/ \ |/ C — M—C >• —M > — M— + (1.7) Although the original Ziegler catalysts were used as heterogeneous 30 catalysts, some have more recently been studied in solution, as have 33-35 other transition metal complexes. A detailed investigation of ethylene dimerization catalyzed by Rh(III) chloride in ethanol confirmed that the increase in chain length occurs 36 via an olefin insertion reaction. The proposed mechanism, as illustrated in Scheme I (L = Cl or solvent), involves protonation of a bis(ethylene)rhodium(I) complex (species A) to an ethylrhodium(III) compound (species B_),a rate-determining rearrangement to a butyl-rhodium(III) complex (species C), release of 1-butene and regeneration of species A. A literature survey of homogeneous polymerization reactions catalyzed by ruthenium complexes is included in a recent review article. 7 - 7 -(A) L2Rh (CH2=CH2)2 >- L3RhI:[IH(CH2=CH2)2 +2C 2H 4 -L -CH2=CH-C2H5 +L L.Rh (CH = CH-C0H,.) J 2 ^ 3 -H -L + L4RhIIi:C2H5(CH2=CH2) (B) +L L4RhIITH(CH2=CH-C2H5) < L • L ^ ^ C a ^ - C ^ (C) Scheme I - 8 -CHAPTER II APPARATUS AND EXPERIMENTAL PROCEDURE 2.1 Materials 2.1.1 Ruthenium Compounds Ruthenium was obtained as the trihydrate Rud^^^O, which is 6 37 38 actually a mixture of Ru(III) and Ru(IV) species, ' ' from Johnson-Matthey Co. The complex RuC^CPPh^^ was prepared by refluxing under nitrogen a methanol solution of RuCl-j^^O containing a six-fold excess of triphenylphosphine according to the method reported by Stephenson 39 and Wilkinson. Dark brown crystals of the complex separated from the refluxing solution in good yield (%70%), and were washed with methanol and ether. Microanalysis confirmed the purity of the product (Calculated for RuC^P-jC^H^,.: C, 67.6; H, 4.7. Found: C, 67.7; H, 4.6%). The solid complex is apparently stable in air, but the crystals were nevertheless stored in a vacuum desiccator in order to prevent any slow oxidation which might occur. The analogous complex RuB^(PPh^).j was prepared using the method as for the dichloro complex, except that a large excess of LiBr was present in the refluxing 39 solution (Analysis: Calculated for RuBr^P^C^^H^^: C, 61.8; H, 4.3. Found: C, 61.9; H, 4.3%). The acetato complex RuH(CH3C02) (PPh^ which was kindly given to us by Dr. G.L. Rempel, can be prepared by passing - 9 -hydrogen gas through a suspension of RuCl 2(PPh 3) 3 in methanol 40 containing two equivalents of sodium acetate. The mono- and dicarbonyl complexes RuC^(CO)(PPh^)2 a™l RuCl2(CO) (PPh 3) 2 were prepared by reaction of carbon monoxide gas with DMA and DMF solutions 41 of RuCl 2(PPh 3) 3. The new hydride complex RuClH(PPh3)3.DMA was prepared by reaction of molecular hydrogen with RuCl 2(PPh 3) 3in DMA solution (Section 4.2.1). 2.1.2. Gases All gases were obtained from Matheson Co., except argon and nitrogen (Canadian Liquid Air Co.). Prepurified grade hydrogen was passed through a Deoxo catalytic purifier before use to remove traces of oxygen, as was deuterium (CP. grade). Ethylene and butadiene (CP. grade), hydrogen chloride (anhydrous) and vinyl fluoride were purified by freezing to liquid nitrogen temperature followed by pumping to remove the non-condensible gases. 2.1.3. Olefinic Substrates Organic substrates were purchased as CP. grade. From K and K Laboratories were obtained 1-hexene, cyclohexene, 1-octene, 2-octene, cyclooctene, 1-decene, diethylfumarate, diethylmaleate, monomethyl-fumarate, monoethylfumarate, methyl-fumaric acid, fumaramide, maleic acid monoamide, acrylamide, trans-cinnamic acid and 4-methoxystyrene. Obtained from Eastman Organic Chemicals were a l l y l alcohol, maleic acid, fumaric acid and trans-stilbene. Methyl vinyl ketone was from Aldrich Chemical Co. and crotonic acid from B.D.H. Ltd. Liquid olefins were - 10 -purified by passage through a 30 cm alumina column under nitrogen, and were used immediately after purification. The absence of peroxide impurity was verified by addition of a few drops of purified olefin to an aqueous solution of ferrous ammonium sulfate and potassium thiocyanate, a red color being indicative of peroxide. The purity of liquid olefins was also checked by gas chromatographic analysis on a suitable column, and by nmr and i r in some cases. The purity of solid olefins was verified by melting point determinations. Maleic acid was recrystallized from water, and fumaric acid from water-ethanol before use. 2.1.4. Other Materials Certified N,N-dimethylacetamide, obtained from Fisher Scientific Co., was purified by distillation from calcium hydride under nitrogen, and stored on Linde 4A molecular sieve under a nitrogen atmosphere. Reagent grade triphenylphosphine (Strem Chemicals) was recrystallized from benzene-ethanol before use. All other solvents and chemicals used were of reagent grade. 2.2. Apparatus for Constant Pressure Gas-Uptake Measurements Gas-uptake measurements were performed using the constant pressure apparatus shown in Figure 1. A flexible glass spiral tube connected a capillary manometer D at tap C to the pyrex reaction flask A; two-necked reaction flasks could also be used, allowing attachment of a sampling tube for collection of gaseous reaction products. The flask was immersed m a thermostated bath B containing silicone o i l (Dow Corning 550), and clipped to a piston-rod and wheel driven by a Welch Figure 1. Apparatus for constant pressure gas-uptake measurements. - I n -variable speed electric motor for shaking purposes. The capillary manometer D containing n-butyl phthalate (a liquid of negligible vapor pressure) was connected to the gas-measuring burette consisting of a mercury reservoir E and a precision bored tube N of known diameter. The gas burette was in turn connected, via an Edwards high vacuum needle valve M, to the gas-handling part of the apparatus; this consisted of a mercury manometer F, gas inlet Y, and connections to the Welch Duo-Seal rotary vacuum pump G. The capillary manometer and gas burette were contained in a transparent perspex water tank thermo-stated at 25°. The silicone o i l bath consisted of a four-litre glass beaker insulated by polystyrene foam on a l l sides and enclosed by a wooden box, with the top also covered by polystyrene. Both thermostat units were controlled by Jumo thermo-regulators and Merc-to-Merc relay control circuits, with heating provided by a 40 watt elongated light bulb. With mechanical stirring and good insulation, the temperature could be maintained to within +0.05°C. A vertically mounted cathetometer was used to follow the gas uptake, and time was recorded during kinetic experiments using a Lab-Chron 1400 timer. 2.3. Procedure for a Typical Gas-Uptake Experiment For each experiment, known amounts of the reactants (e.g. ruthenium complex, substrate) were added to the solvent in the reaction flask A. Solid substrates and catalyst were added in glass buckets, and liquid substrates via syringe. The solvent was always degassed before the addition of catalyst, either by repeated freezing followed by - 13 -warming under vacuum, or simply by shaking under vacuum at room temperature and pumping off the evolved gas (the latter method being used for a solvent of very low vapor pressure, such as DMA). The reaction flask and spiral were f i l l e d with the reactant gas, at a pressure somewhat less than that required for the experiment, by connection to the gas-handling part of the apparatus at 0. The taps C and P were then closed and the reaction flask complete with spiral was disconnected from 0 and attached to the motor driven shaker I. The whole system up to tap C was then pumped down with taps H, K, L, J, and M open. Reactant gas was admitted to the rest of the apparatus beyond C to a pressure slightly less than that desired for reaction. Tap C was then opened and pressure adjusted to the desired reaction pressure by introduction of gas through Y. The vapor pressure of solutions containing liquid olefins was calculated using Raoult's Law. A run was started by closing taps K and L, and simultaneously starting the shaker and timer. Gas uptake was indicated by the difference in the o i l levels of the manometer D. The manometer was balanced by allowing gas to enter the burette through the needle valve, and a constant pressure was thus maintained in the reaction flask. The resulting rise in the mercury level in N was measured at appropriate intervals of time. Since the diameter of the tube N was known, the volume of gas consumed could be calculated and expressed as moles of uptake per li t r e of solution. A rapid shaking rate together with the use of small volumes of solution (2-5 ml) in a relatively large indented reaction flask (30 ml) - 14 -ensured absence of diffusion control on the reaction due to slow dissolution of the reactant gas. 2.4 Gas Solubility Measurements The solubility of a gas in DMA under specific temperature and pressure conditions could be determined using the gas-uptake apparatus and a reaction flask containing a stopcock in its neck. The entire system including the reaction flask containing a measured amount of DMA was evacuated at room temperature. The tap on the flask was then closed, and the flask was placed in the o i l bath at the desired temperature. The system was then evacuated to the flask tap and fill e d with gas to the approximate pressure desired. The flask tap was then opened and the pressure adjusted immediately to that required. Taps K and L were closed, the shaker started and the immediate uptake measured as described in the previous section, allowing calculation of the gas solubility. 2.5. Spectrophotometric Kinetic Measurements Spectrophotometric measurements in the visible range were used to study the kinetics of two reactions in DMA solution. The first case involved the reaction of hydrogen with a ruthenium complex, and the other involved the reverse reaction, evolution of hydrogen from the reaction of HC1 with the hydrido complex, for example: RuCl0(PPh0)„ + H_ ^ RuClH(PPhQ)„ + HC1 (2.1) Because of the extreme air-sensitivity of the ruthenium species in - 15 -solution, i t was necessary to perform the reactions in a strictly anaerobic system. In the reaction involving hydrogen uptake, a relatively large volume of gas (equal to the solution volume) was contained above the reacting solution in a 10 mm anaerobic cell. Because -4 of the low ruthenium concentrations used (% 8 x 10 M), the partial pressure of over the solution remained essentially constant throughout the reaction. In the gas evolution experiments, the optical cell was pumped down at frequent intervals, to ensure that no buildup of evolved hydrogen occurred. In both reactions, the optical cell was contained in a detachable thermostated cell compartment; between optical density measurements the cell compartment was continuously agitated to ensure thermal and physical equilibrium of the solution. Because of the unfortunate instantaneous reaction of hydrogen 42 chloride gas with DMA to produce the white solid DMA(HCl), standard HCl solutions used for the back reaction were prepared instead from LiCl and p-toluenesulfonic acid. Measured aliquots were added to the frozen solutions of hydrido complex prepared in situ, as in reaction (2.1). The solutions were then thawed and brought to the desired temperature in a water bath before transfer of the optical cell to the cell compartment. 2.6. Reaction Product Analysis 2.6.1. Gaseous Products For collection of gaseous reaction products a double-necked reaction flask was employed, one neck being connected to an evacuated sample bulb having a stopcock. At the completion of the reaction, the - 16 -stopcock on the bulb was opened momentarily and a gas sample collected. The sample was then subjected to either gas chromatographic or mass spectrographic analysis. 2.6.2. Liquid and Solid Products Distillation under vacuum (i.e. pumping through a liquid nitrogen cold trap) was used to separate the solvent plus liquid organic products from the catalyst and non-volatile products. The distillate could then be investigated by gas chromatography using a suitable column, or by i r or nmr spectroscopy. The residue could also be studied using ir or by nmr when dissolved in an appropriate solvent. 2.7 Instrumentation Visible and ultraviolet absorption spectra were recorded using a Perkin-Elmer 202 spectrophotometer, fitted when necessary with a thermostated cell compartment. Matched silica cells of 1 mm or 10 mm path length were used. Infrared spectra were recorded on a Perkin-Elmer 457 grating instrument, solution spectra using 0.1 mm NaCI cells,and mull spectra on Csl or KBr plates. Nujol or hexachlorobutadiene were used as mulling media. Proton nmr spectra were obtained using a Varian T-60 instrument, and mass spectra on an Associated Electrical Industries MS9 mass spectrometer. A Beckman GC-2A unit with thermal conductivity type detector was used for gas chromatographic separations with Poropak, dinonyl-- 17 -phthalate, AgNO^-triethyleneglycol or carbowax columns. Conductivity measurements were carried out using a Thomas Serfass conductivity bridge and dip-type conductivity cell. Melting points were recorded using Superior Electric or Fisher-Johns apparatus. - 18 -CHAPTER III THE NATURE OF DICHLOROTRIS(TRIPHENYLPHOSPHINE)RUTHENIUM(II) IN SOLUTION 3.1 Introduction The preparation of the complex RuCl^PPh^)^ was first reported 39 in 1966 by Stephenson and Wilkinson. The five-coordinate nature of this spin-paired d complex is rather unique, considering the high 43 tendency of such a configuration to form octahedral complexes. X-Ray 44 diffraction analysis revealed that the complex is pseudo-octahedral, with a hydrogen atom of a phenyl ring effectively blocking the sixth coordination site. Acetone and benzene solutions of the complex were found to be non-conducting and very air-sensitive, rapidly turning 39 green on exposure to air. Molecular weight measurements in benzene under nitrogen atmosphere suggested dissociation of the complex in solution with loss of a phosphine molecule, K l RuCl 2(PPh 3) 3 —-*» RuCl 2(PPh 3) 2 + PPh3 (3.1) I II 39 since the molecular weight obtained was about half the expected value. 45 However, i t has recently been demonstrated that original molecular - 19 -46 weight measurements on solutions of the related rhodium complex RhClCPPh^)^ xjere unreliable, and in fact a much higher molecular weight (indicating l i t t l e dissociation of the complex) was obtained by ensuring exclusion of oxygen from the rhodium system. The presence of even a trace of oxygen was found to cause a large decrease in the 45 apparent molecular weight. The conclusion of Shriver and coworkers that l i t t l e dissociation of RhCl(PPh3)3 occurs in benzene or chlorinated 47-49 hydrocarbon solutions is in agreement with some nmr data obtained in dichloromethane solution. In view of the degree of uncertainty surrounding the reported molecular weight measurements on solutions of RuCl 2(PPh 3) 3, a spectro-photometry study was carried out on benzene and DMA solutions of the complex in order to determine accurate values for the equilibrium constant K^ . Preliminary conductivity studies on RuCl 2(PPh 3) 3 in DMA solution indicated that a significant portion of the complex was present as an ionic species. For this reason, the complex was examined first in benzene solution, where the non-polar nature of the solvent would preclude the formation of ions. 3.2 RuCl 2(PPh 3) 3 in Benzene Solution The visible spectrum of RuCl 2(PPh 3) 3 was recorded in benzene in the presence of varying concentrations of added triphenylphosphine, the complex and phosphine being dissolved under vacuum in an anaerobic optical cell. Addition of PPh3 was observed to result in a distinct alteration of the spectrum (Figure 2), indicating that the position of .D. 0.0 400 500 600 700 X , nm -3 II Figure 2. Absorption spectra of benzene solutions of RuCl2(PPh3)3 at 25°. (1.04 x 10 M Ru ), {PPh ] : (A) 0.0 M; (B) 5.21 x 10~3 M; (C) 9.24 x 10~2 M. 3 a - 21 -the equilibrium shown in Equation (3.1) was being shifted to the left. An isosbestic was observed at 442 nm. Extinction coefficient (e) values for the trisphosphine species RuCl 2(PPh 3) 3 could be calculated from the spectrum of a solution containing a large excess of phosphine; approximate E values for RuC^(PPh^)2 were estimated from the spectrum of a solution with no added phosphine, initiall y assuming that the complex is completely dissociated at the ruthenium concentration used. At any particular wavelength, the absorbance of a solution containing species I and II is given by the expression: O.D. = e ] [[I] + £T1[II] (3.2) where O.D. is the optical density of the solution measured in a 10 mm optical cell, and E ^ and are the molar extinction coefficients of species I and II, respectively, at the given wavelength. Also, [Bxill]T = [I] + [II] (3.3) [PPh 3] T = [PPh 3] a + [II] (3.4) where [Ru**^ and [PPh3],j, are the total ruthenium and total free phosphine concentrations, respectively, and [PPh»] is the concentration J a of phosphine added to the solution. Combining Equations (3.2) and (3.3) gives the result: O.D. - E [Ru11] [I] = — ~ — (3.5) £ I " £ I I - 22 -Knowing the extinction coefficients of the two species, their concentra-tions may be calculated from Equations (3.5) and (3.3), which allows K^  to be calculated using the relationship: [II][PPh ] Ki = — n r ^ (3-6) Using the estimated extinction coefficients, a value of K^  was calculated for each of the intermediate phosphine concentrations used. Using the median of these K^  values, the concentrations of species I and II were then calculated for the two limiting solutions, the one having maximum triphenylphosphine and the one having no added phosphine. Then, by considering the fact that the limiting solutions may not contain entirely species I or species II, more accurate and e values were calculated; this allowed a second computation of K^  values. Self-consistency could easily be achieved by repetition of this successive approximation procedure two or three times. Accurate extinction coefficients were determined for two different wavelengths, corresponding to the absorption maxima of species I: £ T = 1350 (480 nm) and 515 (750 nm); = 370 (480 nm) and 0 (750 nm). The dependence of solution absorbance on phosphine concentration (Figure 3) is summarized in Table I, along with the calculated K^  values; good agreement is observed between the values obtained at the two different wavelengths. Using the mean value of K^  (2.75 x 10~3 M at 25°), a 10~3 M solution of RuCl 2(PPh 3) 3 is calculated to be 80% dissociated. O.D. to CO 0.0 [PPhJ x 10, M 3 a ' Figure 3, Dependence of optical density of solutions of RuCl 2(PPh 3) 3 on [PPh^] at 25' (O) DMA solution, 2.0 x 10~3 M LiCl (Table V); (•) benzene solution (Table I) (1.04 x 10 3 M Ru 1 1). - 24 -Table I. Spectrophotometric study of equilibrium reaction (3.1) in benzene solution. Dependence of optical density on triphenylphosphine concentration at 25°. [Ru11] = 1.04 x 10~3 M [pph.l j a x 102, M °-D'480a °-D'750a [H]/[I] b [PPh 3] T C x 102, M x 103, M 0 0.60 0.10 3.62 0.08 3.0 (4.36)e (0.085)6 (3.7)e 0.14 0.84 0.23 1.22 0.19 2.4 (1.33) (0.195) (2.6) 0.38 1.00 0.31 0.65 0.42 2.7 (0.73) (0.43) (3.1) 0.52 1.04 0.34 0.54 0.56 3.0 (0.57) (0.56) (3.2) 0.76 1.21 0.43 0.23 0.77 1.8 (0.24) (0.775) (1.9) 9.24 1.37 0.52 0.03 9.24 2.8 (0.03) (9.24) (2.8) Measured in a 10 mm optical cell. Calculated using Equations (3.3) and (3.5). Calculated using Equation (3.4). Calculated using Equation (3.6). Values in parenthesis are those calculated using the O.D. values. a b c d e - 25 -Temperature variation on benzene solutions of RuCl2(PPh^)^ resulted in significant and reproducible change in the visible spectrum, consistent with a shift in the position of equilibrium (3.1). At two of the intermediate phosphine concentrations, the value of was determined at four different temperatures, and the mean value for each temperature is reported in Table II. Extinction coefficients were assumed to be independent of temperature. 3.3 RuCl2 (PPh^-j in Dimethylacetamide Solution Conductivity measurements at 25° on solutions of RuCl2(PPh^)^ in DMA indicated the presence of ionic species. Because of the extreme air-sensitivity of the solutions, an air-tight glass conductivity cell was used, and the solvent was degassed using purified nitrogen or argon before the addition of the compound from a suspended glass bucket. The stream of inert gas was maintained while the solid was dissolving, and as long as measurements were being recorded. Even after the complex appeared to be completely dissolved (^ 30 min), the conductivity of the pale orange solutions continued to increase, and a slow increase was s t i l l observed even after many hours. In addition, some of the solutions exhibited a definite green tinge, indicating that some oxidation of the complex was taking place. This is not -4 surprising, since the solutions were Very dilute (3 x 10 M) and therefore sensitive to traces of oxygen in the gas stream or residual oxygen in the solvent. In contrast, visible absorption spectra recorded on these DMA solutions contained in an anaerobic cell under vacuum remained unchanged from the time of dissolution of the complex, - 26 -Table II. Spectrophotometries study of equilibrium reaction (3.1) in benzene solution. Temperature dependence of K^ . [Ru11] = 1.04 x 10~3 M, [PPhJ = 0.52 x 10~ 2 M and 3 a 0.76 x 10" 2 M. * Temp. K l °C x 10 3, M 25.0 2.6 33.2 3.2 40.8 3.8 49.0 4.4 Mean value (calculated using Equation (3.6) at two PPh„ concentrations). - 27 -confirming that the slow changes in solution color and conductivity were a result of destruction of the complex and not due to a slowly attained equilibrium. Conductivity measurements recorded after the solid appeared to be completely dissolved gave an equivalent conductance -1 2 -1 (A) value of about 10 ohm cm moi . However, consistent results could not be obtained, presumably due to the oxygen sensitivity of the complex. Since strong 1:1 electrolytes in DMA solution were observed -1 2 -1 to exhibit conductivity in the range A = 40-60 ohm cm moi , the ruthenium complex was probably only partially ionized. The addition of LiCl to DMA solutions of RuC^CPPh-j) 3 resulted in a distinct change in the visible spectrum (Figures 4,5A). The presence of a slight excess of chloride resulted in a spectrum (Figure 5, Spectrum A) very similar to that of RuC^^Ph^^ in benzene solution; addition of larger amounts of LiCl caused no further change in the spectrum. Addition of triphenylphosphine to a DMA solution of RuC^^Ph^), containing excess chloride ion resulted in similar spectral changes (Figure 5) to those which characterized an increase in the ratio [I]/[II] in benzene solution, indicating that the dissociation equilibrium (3.1) also applies to DMA solution. A good isosbestic was observed at 444 nm. The greater magnitude of in DMA than in benzene (see later) was evident from the much larger concentrations of added PPh^ required in DMA for an equivalent spectral change. No conductivity increase occurred on addition of phosphine (PPh^Ru = 30) to DMA solutions of RuC^tPPh^) 3 > suggesting no displacement of ionic chloride from the complex by PPh^. The results obtained in DMA solution thus indicated complete loss of one PPh^ ligand, followed by a second (ionic) dissociation: 500 X , nm -3 II Figure 4. Absorption spectrum of RuCl2(PPh3>3 in DMA solution at 25°. (1.04 x 10 M Ru ), - 29 -o o o d cu M 3 60 - 30 -K + RuCl 2(PPh 3) 2 RuCl(PPh3)2 + Cl (3.7) II III In order to determine the magnitude of K2, visible spectra were recorded for DMA solutions of RuCl2(PPh3>3, containing in fact bis-phosphine species, with various concentrations of added chloride ion. An isosbestic was observed at 439 nm. Extinction coefficients for species II were obtained from the spectrum of a solution containing excess LiCl. Because the majority of complex was present in solution as the neutral species II, even in the absence of added chloride, i t was more difficult to estimate E values. These were calculated using the same successive approximation method employed to obtain more accurate values of and in benzene solution. Extinction coefficients were calculated at the two absorption maxima for species III in the visible spectrum: = 460 (476 nm) and 71 (670 nm) ; cm = 2360 (476 nm) and 940 (670 nm). Optical density measurements at varying chloride concentration are tabulated in Table III and shown graphically in Figure 6. The concentrations of species II and III were calculated (cf. Equation 3.5), and a K2 value of 7.3 x 10 ^ M at 25° was determined from the slope of a plot of [III]/[II] against fCl "^ (Figure 7), where the total free chloride ion concentration is given by: IC1 ] T = [LiCl] + [III] (3.8) and [III][C1 ] T (3.9) [II] - 31 -Table III. Spectrophotometry study of equilibrium reaction (3.7) in DMA solution. Dependence of optical density on LiCl concentration at 25°. [Ru11] = 9.75 x 10~4 M [LiCl] x 10*, M. O.D.4?6a [UI]/[II] b x 10 [ci ] T C x 104, M [ c r ] T _ 1 -3 -1 x 10 ,M K d  K2 x 105, M 0.00 0.89 3.09 2.30 4.35 7.1 0.72 0.84 2.70 2.79 3.58 7.5 1.44 0.81 2.40 3.33 3.00 8.0 2.16 0.77 2.13 3.87 2.58 8.2 3.60 0.68 1.46 4.84 2.07 7.1 5.04 0.64 1.16 6.05 1.65 7.0 7.20 0.59 0.86 7.97 1.26 6.8 14.40 0.52 0.43 14.80 0.68 6.3 Measured in a 10 k Calculated using mm cell. equations analogous to (3.5) and (3.3). Calculated using Equation (3.8). Calculated using Equation (3.9). 0.95 0.80 O.D. 476 0.65 0.50 6.0 0.5 1.0 1.5 [LiGl] x 10" M Figure 6. Dependence of optical density of DMA solutions of PoiCl,, (PPh^ on [LiCl] at 25°. (9.75 x 10~4 M Ru 1 1). (Table III). - 34 -Using the estimated value shows that the ionic species III makes -3 II up 24% of the total ruthenium concentration at 10 M Ru and 25°. The temperature variation of K2 was determined with no added chloride, and the results are summarized in Table IV. The magnitude of was determined in DMA solutions containing excess chloride ion to suppress reaction (3.7). Molar extinction coefficients, determined by successive approximations, were = 1370 (480 nm) and 530 (750 nm); = 540 (480 nm) and 40 (750 nm). These values are very similar to those found in benzene solution. The variation in solution absorbance with changing phosphine concentration at 25° is summarized in Table V and is illustrated graphically in -2 Figure 5. A value of 4.6 x 10 M (at 25°) was determined from the slope of the straight line obtained by plotting [II]/[I] against [PPh 3] T - 1 (Figure 8). Using this value of R^ , RuCl 2(PPh 3) 3 is calculated to be completely dissociated to RuCl 2(PPh 3) 2 in DMA -3 II solution at 10 M Ru , as compared to 80% dissociation in benzene at the same concentration of complex. The temperature dependence of in DMA was determined at one intermediate phosphine concentration, the results being summarized in Table VI. 3.4 Discussion Using the temperature dependence data for in benzene (Table II) and DMA (Table VI), and for K2 in DMA (Table IV), van't Hoff plots were constructed (Figures 9-11) and the values of AH° and AS0 determined for the three reactions (Table VII). If a solvent molecule is considered as an entering ligand in these reactions, for example, - 35 -Table IV. Spectrophotometric study of equilibrium reaction (3.7) in DMA solution. Temperature dependence of K^. [Ru11] = 9.75 x 10~4 M (no added LiCl). * Temp. K2 °C 5 x 10 , M 10.0 5.7 20.0 6.7 30.0 7.3 40.0 8.3 50.0 9.4 * Calculated using Equation (3.9). - 36 -Table V. Spectrophotometric study of equilibrium reaction (3.1) in DMA solution. Dependence of optical density on triphenyl-phosphine concentration at 25°. [Ru11] = 1.04 x 10~3 M, [LiCl] = 2.0 x 10~3 M j a x 102,M O' D-480 a °-D'750a [H]/[I] b [PPh 3] T C x 102,M [PPh 3] T- 1 x l O - 1 ^ " 1 x 102,M 0.00 0.56 0.04 0.76 0.68 0.11 6.64 0.85 11.7 5.7 (6.32)e (0.85)6 (11.7) 6 (5.4)6 1.22 0.76 0.15 3.47 1.30 7.69 4.5 (3.66) (1.30 (7.69) (4.8) 1.90 0.79 0.19 2.87 1.98 5.05 5.7 (2.41) (1.97) (5.08) (4.8) 1.98 0.86 0.20 1.95 2.05 4.88 4.0 (2.20) (2.05) (4.88) (4.5) 2.82 0.92 0.24 1.45 2.88 3.48 4.2 (1.55) (2.88) (3.48) (4.5) 3.69 0.96 0.27 1.22 3.75 2.67 4.6 (1.27) (3.75) (2.67) (4.8) 4.19 0.99 0.29 1.09 4.24 2.36 4.6 (1.09) (4.24) (2.36) (4.6) 8.56 1.11 0.38 0.59 8.60 1.16 5.1 (0.50) (8.59) (1.17) (4.3) 14.20 1.19 0.41 0.39 14.23 0.70 5.5 (0.38) (14.23) (0.70) (5.4) 18.40 1.21 0.44 0.34 18.43 0.54 6.3 (0.27) (18.42) (0.54) (5.1) Measured in a 10 mm optical ce l l . Calculated using Equations (3.3) and (3.5). Calculated using Equation (3.4). Calculated using Equation (3.6). Values in parentheses are those calculated using the O.D.7C.n values. t l ] / [ l ] - 38 -Table VI. Spectrophotometry study of equilibrium reaction ( 3 . 1 ) in DMA solution. Temperature dependence of K^ . I R U 1 1 ] = 1 . 0 4 x 1 0 ~ 3 M, [LiCl] = 2 . 0 x 1 0 ~ 3 M, [PPh_] = 4 . 1 9 x 1 0 ~ 2 M. Temp. . °C A K l x 102, M 25.0 4.4 33.0 5.0 41.0 5.7 48.5 6.4 * Calculated using Equation (3.6). - 39 --2.35 h -2.45 log Kx -2.55 h 3.10 3.20 T _ 1 x 103, °K _ 1 3.30 3740 Figure 9. Van't Hoff plot for equilibrium (3.1) in benzene solution. (Table II) -1.20 log Kx -1.30 3.40 Figure 10. Van't Hoff plot for equilibrium (3.1) in DMA solution. (Table VI). - 40 -Figure 11. Van't Hoff plot for equilibrium (3.7) in DMA solution. (Table IV). - 41 -Table VII. Spectrophotometry study of equilibrium reactions (3.1) and (3.7). Thermodynamic parameters. o t Reaction Solvent AH° AS° AS kcal/mole eu eu (3.1) benzene 4.2 + 0.2 2.2 + 0.6 -2.6 + 0.6 (3.1) DMA 3.1+0.1 4.1+0.3 -0.6+0.3 (3.7) DMA 2.2 + 0.1 -11.7 + 0.3 -16.4 + 0.3 * Obtained from Figures 9-11. t Obtained using the values of or which were calculated by including the solvent concentration (see Equation 3.10). - 42 -K l S + RuCl 2(PPh 3) 3 -—»- RuCl2(PPh3)2(S) + PPh3 (3.10) and the equilibrium constants calculated by incorporating the solvent concentration (11.3 M for benzene and 10.8 M for DMA), somewhat lower AS° values are obtained (Table VII). Since reaction (3.1) involves replacement of a triphenylphosphine II molecule in the coordination sphere of Ru by a solvent molecule, the A H 0 value for the reaction can be used to give an indication of the relative coordination strength of the phosphine and solvent molecules. From the slightly endothermic AH° values i t is evident that PPh3 coordinates more strongly than either solvent molecule, and that DMA coordinates more strongly than does benzene. DMA is known to be a good coordinating solvent, and'for example has a donor strength comparable to that of water. The order-of-magnitude difference between the values in DMA and in benzene results from both enthalpy and entropy factors being more favorable for the forward reaction in DMA. The negative AS° value observed for reaction (3.7) is characteristic of a dissociation reaction producing ions in a polar medium,"'1 and is due to the high degree of solvation of the ionic species with a resulting loss in freedom of motion. It is not easy to predict enthalpy changes for ionization reactions of this type;."'1 the change is small, but endothermic, showing that the heat of solution of the ions does not quite compensate for the energy required to separate the ions from each other. 52 A recent spectrophotometric study by Arai and Halpern on the dissociation of the rhodium complex RhCl(PPh3)3 - 43 -K RhCl(PPh3)3 —»» RhCl(PPh3)2 + PPh3 (3.11) gave a K value of 1.4 x 10 ^ M at 25° in benzene solution. The quantitative aspects of the spectrophotometric data have been questioned, 49 however, in terms of a possible contribution from the dimerization equilibrium, 2RhCl(PPh_)„ —»• [RhCl(PPh_)_]0 + 2PPh_ (3.12) o j — i l l i which was thought not significant by Aral and Halpern. Molecular weight and nmr studies on the nature of the complex RhH(CO)(PPh3)3 in solution indicated that extensive phosphine 53 dissociation occurred in organic solvents, although no value was determined quantitatively for the equilibrium constant. However, the reaction was endothermic, and in benzene at 25° the bisphosphine -3 -2 complex was said to predominate at 10 M, while at 10 M there was considerable reassociation to the tris complex. Thus, the equilibrium constant is probably similar to that determined here for the complex RuCl 2(PPh 3) 3 in benzene solution. Although the complex RuCl2(PPh3)^ has apparently been isolated in 39 the solid state, i t dissociates extensively in solution. In the present study, the data analysis gave no evidence for the existence of the tetrakis(phosphine) species in solution. - 44 -CHAPTER IV THE EQUILIBRIUM REACTION OF DICHLOROTRIS(TRIPHENYLPHOSPHINE)-RUTHENIUM(II) WITH MOLECULAR HYDROGEN IN DIMETHYLACETAMIDE SOLUTION 4.1 Introduction 8—10 Wilkinson's group first reported that the complex RuCl 2(PPh 3) 3 reacts with molecular hydrogen in benzene solution to produce RuClH(PPh3)3, which can be isolated from solution as the benzene solvate RuClH(PPh3)3*CgHg. The presence of a base such as triethylamine was required in order for the reaction to proceed. One mole of H^  was absorbed, and one mole of amine hydrochloride was produced per mole of RuCl 2(PPh 3) 3, the reaction being: RuCl2(PPh)3 + H2 + base >- RuClH(PPh3)3 + base(HCl) (4.1) Sodium borohydride could be used instead of molecular hydrogen, and alternative bases (sodium phenoxide, potassium hydroxide, alcohols) could also be employed. Violet-black crystals of the hydrido-complex slowly darkened in air ('v- 12 hr),and the hydride solutions were exceedingly air-sensitive, rapidly turning green on exposure to air. 54 A three-dimensional X-ray diffraction study on RuClH(PPh„)_'C..H, - 45 -showed the structure to be a higly distorted trigonal bipyramid with the phosphine groups approximately equatorial, and the benzene molecule merely occupying a site in the lattice. A high-field hydride resonance was observed in CDC1„ solution (T = 27.4, J „ = j r — n 26 cps), the symmetrical quartet structure of the resonance indicating that the hydride is mutually cis to three equivalent phosphorus atoms. A single Ru-H stretching frequency was observed in the infrared spectrum at 2020 cm * (Nujol mull).*^ Studies in this laboratory showed that the hydride complex could also be prepared in DMA by reaction of commercial ruthenium trichloride (RuCl^OR^O) with in the presence of a four-fold excess of triphenylphosphine.''"''' The reaction proceeded rapidly at 80°, with uptake of one mole of per mole of ruthenium, the final solution being violet. The proposed mechanism of this reaction involved production of the brown complex RuC^CPPh^^ as an intermediate. Addition of base was not required in DMA for the reaction (4.1) to proceed, because of the basic nature of the solvent itself. In this present work, i t was decided to prepare the complex 39 RuC^^Phg)^ according to the literature method, and to react this complex with hydrogen. This method would eliminate the excess chloride ion and water introduced through the use of RuCl^'SE^O as starting material. In addition, since a four-fold excess of PPh^ (a 2-equivalent reductant) had been used in the previous preparations in DMA, a slight excess of phosphine remained after the reduction to Ru**, since the commercial trichloride is a variable mixture of Ru*** , _ IV . 6,37,38 and Ru species. - 46 -4.2 Reaction of RuCl 2(PPh 3) 3 with Hydrogen in DMA The reaction of RuCl^CPPh^)^ with in DMA solution proceeded rapidly at 25°, with uptake of one mole of YL^ P e r mole of complex. Because of the extreme air-sensitivity of the solutions, the solvent was carefully degassed before addition of the RuC^CPPh^^ from a suspended glass bucket, and the reactions were performed under strictly -4 anaerobic conditions. If a dilute (7 x 10 M) solution of the violet hydride were prepared using purified solvent which had been stored under nitrogen for several months, the color of the solution was found to fade to pink rather quickly. With very "old" DMA the solutions became completely colorless, but no fading at a l l was observed using freshly distilled DMA. Additions of ^0, LiCl or _p_-toluenesulfonic acid to the solvent had no effect on the rate of fading, although the presence of excess PPh^ significantly decreased the rate. The addition of lithium acetate at low concentration (^[Ru]) resulted in rapid fading of the otherwise stable violet solutions, and the visible spectrum of the final yellow solution was the same as that of a DMA solution of RuH(OAc) (PPh^)^ (continuum above 420 nm, ~ 130) y faded solutions formed using "old" DMA had a similar spectrum (continuum above 400 nm, = 20-150). The fading possibly results, therefore, from the presence of dimethylammonium acetate in the solvent, this salt being the hydrolysis product of the hygroscopic DMA."'"' The fading problem was avoided by using the solvent within several weeks of its purification by distillation. -2 Because of the solubility limit of RuClH(PPh3)3 in DMA (^  10 M when prepared in situ), the weak infrared peak of the Ru-H stretch*^ - 47 -could not easily be observed in DMA solution. However, due to greatly increased solubility in chlorinated hydrocarbons, the hydride peak was easily distinguished at 2030 cm 1 in C^C^ solution after the reaction of hydrogen with RuCl^CPPh^)^ was carried out in the presence of triethylamine in this solvent. The hydride peak disappeared completely within several minutes, as the solution became oxidized from exposure to air. A carbonyl peak was observed at 1925 cm 1 in violet DMA solutions prepared at 80°, indicating that solvent decarbonyla-tion was occurring at high temperature, with production of a ruthenium-carbonyl complex. No high-field hydride resonance appeared in the nmr spectrum of the violet DMA solutions, apparently due to low solubility. This resonance was previously observed at x 27.4 using a CDCl^ solution of % 10 _ 1 M RuClH(PPh„).-C,H£.10 J J D O 4.2.1 Preparation of the DMA Solvate RuClHCPPh^yDMA The DMA solvate RuClHCPPh^^'DMA, analogous to the benzene solvate prepared by Wilkinson's group,could readily be isolated from concentrated solutions of the hydride complex in DMA. A suspension of RuC^CPPh.^ (0.1 g/ml) in DMA was warmed under hydrogen (1 atm) until reaction was complete. Dark violet crystals formed in the solution after several hours standing, and a reasonable yield (50%) was obtained by allowing crystallization to proceed overnight. The product was collected, washed quickly with DMA, and stored in vacuo. Microanalysis indicated the presence of one mole of DMA per mole of complex (Calculated for RuClP0CcoHcl.NO: C, 68.9; H, 5.5; N, 1.4; J J O J J - 48 -Cl, 3.5%. Found: C, 68.8; H, 5.5; N, 1.3; Cl, 3.7%). The carbonyl stretching mode in the infrared spectrum appeared at 1650 cm * (Nujol mull), very close to the position for the free solvent (1660 cm *). The DMA is thought to merely occupy a site in the crystal lattice, as was established for the benzene molecule in RuClH(PPh3)3'C^H^ since coordination of DMA results in much greater frequency shifts. A weak Ru-H stretching frequency appeared at 2030 cm *. The violet crystals were somewhat air-sensitive, darkening in air over several days. RuClH(PPh3)3"DMA could be redissolved in DMA, but dissolution took place very slowly at 25°, and solubility was limited. The visible spectrum of this solution was the same as that recorded on a violet solution prepared in situ from RuCl 2(PPh 3) 3 and hydrogen (Figure 12), indicating that the same species was present in both cases. Some variation in extinction coefficient (1800 + 200) was observed at the absorption maximum (520 nm), depending on the DMA batch used. The analogous hydride species RuBrH(PPh3)3 was likewise prepared in situ from RuBr2(PPh3)3 and H2, and the spectrum of this violet solution was quite similar to that for the corresponding chloride system (Figure 12). Conductivity measurements on RuClH(PPh3)3 in DMA solution at -4 3 x 10 M Ru(II) indicated that no loss of chloride by the complex occurred (A = 0), in contrast to the observations on the dichloro complex (Section 3.3). However, the equivalent conductance measured on hydride solutions of similar concentration prepared in situ was - 1 2 - 1 'v 40 ohm cm mol , indicating that the HC1 produced in the hydrogen reaction is substantially ionized at low concentration. e x 10 -3 A00 500 600 700 Figure 12. Visible absorption spectra recorded in DMA at 25°: (A) 8.1 x 10~ M RuClH(PPh„)» -4 (B) 5.1 x 10 M RuBrH(PPh3)3-- 50 -By analogy with the dichloro complex (Equation 3.1), dissociation of the hydride complex with loss of one PPh^ ligand seemed a likely possibility, K3 RuClH(PPh„)„ — ^ RuClH(PPh0)„ + PPh„ (4.2) J J -« j 2 j and in fact this dissociation has been proposed1^ to occur in benzene solution, based on the severe inhibitory effect observed for PPh^ on hydrogenation reactions catalyzed by the hydride. However, the nmr spectrum of the hydride in CDCl^ solution was found to be a sharp, symmetrical quartet, indicating l i t t l e dissociation in this solvent. 1^ In DMA, addition of even a very large excess of phosphine (PPh^Ru > 200) to the dilute violet solutions resulted in no change in the visible spectrum. As well, the diffuse reflectance spectrum on the violet solid RuClH(PPh3)2*DMA was very similar to the solution spectrum, but with a slight shift in the peak position to 534 nm. The data thus indicated that, either there was no RuClH(PPh,j)2 complex present in the DMA solution at the concentrations used,or else that the spectra of the bis and trisphosphine hydride species must be very similar in DMA. The rapid reaction between hydrogen and RuC^tPPh^^ in 1:1 benzene-ethanol solution resulted in a violet solution having the same visible spectrum as was recorded for the violet DMA solutions. 4.2.2 Kinetics of the Reaction of Hydrogen with RuCl_(PPh0)0 in . * L j j DMA As determined previously (Section 3.3), RuC^(PPh^)^ (species I) -3 II exists in DMA solution (at 10 M Ru ) as a mixture of the two species - 51 -RuCl 2(PPh 3) 2 (species II) and RuCl(PPh3)2 (species III); however, in the presence of a slight excess of LiCl, only species II is present. The reaction of H 2 with II was therefore studied in DMA solutions containing a too-fold excess of LiCl. The production of the violet hydride species (followed continuously and spectrophotometrically at 530 nm) was a pseudcrfirst-order reaction, a plot of log(O.D.^-O.D.) vs. time being linear over at least three half-lives. A representative log plot is shown in Figure 13. The pseudo-first-order rate constant k' (sec *) was determined from the slope of the straight line plots using the relationship k' = -2.3 (slope). Reproducibility was usually within 5%, except when using different solvent batches. Values of k' were determined at different temperatures and H2 concentrations, and are summarized in Table VIII and Figure 14; the dependence of H2 is f i r s t -order at low H 2 pressure, but approaches zero-order at pressures of about 1 atm. The production of the violet hydride at 25° and 1 atm was also _2 followed by measurement of gas uptake on a 10 M DMA solution of RuCl2(PPh3) containing a two-fold excess of LiCl. Because the ruthenium concentration in this experiment was an order of magnitude higher than the concentration used in the spectrophotometric investigations, phosphine dissociation by the complex was not expected to be complete in the solution used for the uptake experiment, so that the reaction rate is not strictly comparable to the spectrophotometric data. Nevertheless, the reasonably linear pseudo-first-order uptake plot gave -3 -1 a k' value of 2.8 x 10 s , compared to the spectrophotometric value -3 -1 of 4.87 x 10 s at the same temperature and hydrogen pressure. - 53 -Table V I I I . Reaction of RuCl 2 ( P P h 3 ) 3 i n DMA w i t h H . Dependence of k' on c o n c e n t r a t i o n and temperature. [ R u 1 1 ] = 0.81 x 10" 3 M, [ L i C l ] = 1.65 x 10" 3 M. Temp. H 2 [H 2J k' [IV" 1 (k')" 1 °C mm x 10 3,M x 1 0 3 , s - 1 x 10~3,M" 1 x 10" 2, s 10 100 0.21 0.45 a 4.77 22.3 10 175 0.37 0.73 2.72 13.8 10 278 0.59 0.99 1.71 10.1 10 475 1.00 1.17 1.00 8.55 10 760 1.60 1.42 0.63 7.05 15 143 0.30 1.08 b 3.32 9.26 15 278 0.59 1.50 1.71 6.67 15 339 0.84 1.77 1.19 5.65 15 760 1.60 2.28 0.63 4.39 20 143 0.30 1.49 C 3.32 6.71 20 278 0.59 2.16 1.71 4.63 20 429 0.90 2.72 1.11 3.68 20 760 1.60 3.36 0.63 2.98 25 141 0.30 2.36 d 3.37 4.24 25 278 0.59 3.20 1.71 3.13 25 446 0.94 4.46 1.06 2.24 25 760 1.60 4.87 0.63 2.05 k 2 = k'/[H 2] = 3 2.14 M " 1 s " 1 ; b 3.59 M ^ s " 1 ; C 4.95 M V " 1 ; d 7.95 M " 1 S ~ 1 . \ - 55 -In the presence of excess PPh^ and a slight excess of chloride, species I is present in DMA solution as well as species II (Section 3.3). The reaction of the solution with H2 w a s s t i H pseudo-first-order (see the representative logarithmic plot in Figure 13), and the rate constant (k") decreased with increasing phosphine concentration (Table IX, Figure 15). In contrast to the results obtained without added PPh^* the reaction at 0.146 M PPh^ showed a strictly first-order dependence on H2 concentration up to 1 atm pressure (Table X, Figure 16); at this high PPh^ concentration the presence of excess LiCl or HCl had l i t t l e effect on the rate of reaction (Table X). Without the presence of added LiCl or PPh3, RuCl 2(PPh 3) 3 exists in DMA solution as a mixture of species II and III (Section 3.3). Under these conditions, the formation of the violet hydride by reaction of H 2 with the solution was a pseudo-first-order reaction for only approximately 60% of the reaction (Figure 17), and a rate constant k"1 was obtained from this i n i t i a l linear portion of the log plot in order to reflect the reactivity of the i n i t i a l mixture of II and III. The rate showed a first-order dependence on hydrogen concentration up to 373 mm H2 (Table XI, Figure 18), no results being obtained at higher hydrogen pressure because of ftie extremely rapid reaction rate and the possibility of diffusion control. Addition of a slight excess of LiCl resulted in a marked decrease in the reaction rate, but addition of further chloride had no significant effect (Table XI, Figure 19). Hydrogen uptake experi-ments on DMA. solutions of RuCl 2(PPh 3) 3 at 35° confirmed that the rate of gas uptake was decreased in solutions containing added LiCl. - 56 -Table IX. Reaction of RuCl 2(PPh 3) 3 in DMA with H2. Dependence of k" on PPh_ concentration at 25°. [Ru11] = 0.81 x 10 3 M, [LiCl] = = 1.95 x 10"3 M, [H2] = 0.29 x 10"3 M. J a k" [II]" [PPh 3] T + [PPh 3] T _ 1 x 102, M x 10 4,s - 1 x 102,M x l O " 1 ^ " 1 0.00 16.8 0.55 13.1 3.60 0.61 16.4 1.09 9.87 1.44 1.14 8.77 2.27 7.45 0.80 2.31 4.33 5.24 5.80 0.53 5.27 1.90 10.60 3.99 0.31 10.62 0.94 19.80 2.04 0.14 19.81 0.51 100.10 0.35 0.02 100.10 0.10 Calculated from Equation ( 4 . 8 ) . Calculated from Equation (4.9). - 57 -Table X. Reaction of RuC^(PPh^)^ with in DMA solutions containing added PPh^. Dependence of k" on concentration of H 2 and temperature. [Ru11] = 0.81 x 10~3 M, [PPh„] = 0.146 M 3 a Temp. H 2 [H2] °C mm x 10 3, M 25 760 1.60 1.77 25 760 1.60 1.81a 25 760 1.60 1.67b 25 502 1.06 1.07b 25 258 0.54 0.55b 10 760 1.60 0.42b 15 760 1.60 0.70b 20 760 1.60 1.06b 3 [LiCl] = 3.16 x 10"3 M. b [HCl] = 2.76 x 10~2 M. k" i n 3 _ 1 x 10 , s - 58 -1.0 0.5 0.0 0.0 1.0 2.0 [PPh„] x 10, M - 1 II Figure 15. Dependence of k" on [PPh3] in DMA at 25° (0.81 x 10 M Ru . 1.95 x 10~3 M LiCl, 0.29 x 10~3 M H2> (Table IX). 2.0 * CO o r-l X A! 1.0 [H2] x 10" Figure 16. Dependence of k" on [H2] in DMA at 25 0.146 M PPh3, 2.76 x 10 _ 2 M HC1) (Table X) -1 II (0.81 x 10 M Ru , - 59 -- 60 -Table XI. Reaction of RuCl 2(PPh 3) 3 in DMA with K . Dependence of k"1 on concentrations of H2 and LiCl. [«""] • 0.81 x 10~3 M. Temp. H2 [H2] [LiCl] k"' °C mm x 104, M x 103, M x 10 3,s _ 1 10 100 2.10 0.00 1.08 10 193 4.06 0.00 2.62 10 273 5.75 0.00 2.90a 10 278 5.85 0.00 3.02 10 373 7.85 0.00 4.46 25 100 2.10 0.00 3.80b 25 278 5.85 0.63 3.86 25 278 5.85 1.65 3.20 25 278 5.85 3.16 2.68 25 278 5.85 5.06 2.92 1.67 x 10 M sodium tetraphenylborate added. k1" = 10.6 x 10~ s~ at 5.85 x 10~ M.H2> since the reaction is first-order in IH 0]. - 61 -[LiCl] x i n 3 , M Figure 19. Dependence of k'" on [LiCl] in DMA at 25° (.0.81 x 10~ 3 M Ru 5.85 x 10~ 4 M H 2) (Table XI). - 62 -4.3 The Reaction of HC1 with RuClH(PPh3)3 in DMA The reaction between and RuC^CPPh^.^ in DMA solution is essentially complete for a l l of the conditions reported above in Section 4.2.2, since the visible spectrum (including extinction coefficients) of the violet solutions generated is the same as that of a DMA solution of RuClH(PPh3)3"DMA. However, removal of B.^ after an "in situ" preparation resulted in the fading of the violet color, with the visible spectrum of the solution gradually approaching that of the i n i t i a l solution of RuC^^Ph.^ before the hydrogen reaction. Addition of HC1 to the violet solutions formed in situ increased the rate of this back reaction, although neither H + (added as p_-toluenesulfonic acid) nor LiCl added alone had this effect. In the presence of excess HC1, the loss of the violet color at 530 nm was a pseudo-first-order reaction, since plots of logCO.D.-O.D.^ ) against time were linear (Figure 20); the visible spectrum of the final solution was that of the dichloro complex. The pseudo-first-order rate constant k^V showed a first-order dependence on HC1 concentration (Table XII, Figure 21) and inverse dependence on PPh3 concentration (Table XII, Figure 22). The total concentration of HC1 in the solution is the amount of HC1 added plus the amount produced by the preparation of the hydride complex in situ: [HC1]T = [HCl] a + [Ru I T] T (4.3) where [HC1]„, and [Ru**]_ refer to total concentrations and [HC1] to added i l a concentration. The results of temperature dependence experiments are summarized in Table XIII. 0.0 0.5 1.0 _ 3 time x 10 ,s 1.5 Figure 20. First-order rate plots for the reaction of RuClH(PPh 3) 3 in DMA with HC1 at 25c ( O ) 7»65 x 10~3 M HC1; ( A ) , 1.31 x 10~2 M HC1, 2.40 x 10~3 M PPt^. 2.0 (0.81 x 10~3 M Ru 1 1). - 64 -Table XII. Reaction of RuClH(PPh3)3 in DMA with HCl. Dependence of , iv , k on concentrations of HCl and PPh3 at 25°. [Ru11] = 0.81 x 10 3 M. [HCl] a [HC1]T* [PPh 3] a IV k IV k /[HC1]T x 102,M x 102,M x 103,M x 10 3,s - 1 x 10, M - 1s _ 1 0.26 0.34 0.00 1.423 4.23 0.51 0.59 0.00 2.11 3.57 0.55 0.63 0.00 2.52 3.98 0.77 0.85 0.00 3.49 4.12 1.10 1.19 0.00 4.41 3.72 0.37 0.45 0.00 1.38b 3.09 0.78 0.86 0.00 2.72b 3.18 1.04 1.12 0.00 3.58b 3.21 0.46 0.54 0.27 1.66 3.08 0.46 0.54 0.73 1.07 1.99 0.66 0.74 1.32 1.21 1.65 1.31 1.39 2.40 1.34 0.96 1.31 1.39 4.77 0.78 0.56 3.28 3.36 9.78 0.96 0.29 6.55 6.63 26.90 0.74 0.11 7.63 7.71 59.70 0.36 0.05 3.28 3.36 59.70 0.15 0.04 IHC1]T = [HCl] + [Ru 1 1]. a Initial rate. b 760 mm C.H. added. 2 4 0.0 0.4 „ 0.8 1.2 [HC1]_ x 10% M Figure 21. Dependence of k i V on [HCl] in DMA at 25° (0.81 x 10 M Ru ) (Table XII). - 66 -Table XIII. Reaction of RuClH(PPh3)3 in DMA with HC1. Temperature dependence of k*V. [Ru11] = 0.81 x 10~3 M. Temp. k i V * k3 K3 °C x l O ^ s - 1 x 10 3,s _ 1 25 0.87a 0.45 30 1.63a 0.84 35 2.48a 1.28 40 4.13a 2.12 10 4.97b 15 8.93b 20 15.4b 25 25.2b 30 47.2b Calculated according to Equation (4.32). 3 [HC1J = 2.76 x 10 2 M, [PPh ] = 0.146 M. 3. j Si b [HCl] = 5.53 x 10~3 M. - 67 -On addition of one atmosphere of ethylene to the violet hydride solution prepared in situ in the absence of added HCl, an immediate decrease in 0 - D « 5 2 o o c c u r r e d » d u e t o establishment of the equilibrium: RuI]C-H + C„H. ^ (C0H,)Rui:E-H RuII-C0H1. ( 4 . 4 ) 2 4 - « — 2 4 — I i> which will be discussed in Section (6.3). Further slow fading of the violet color then occurred at a rate similar to that observed for the reaction of "in situ" HCl with the hydride under vacuum, and the final solution spectrum again indicated that the dichloro species was the product of the reaction. Gas chromatographic analysis at this stage revealed the presence of ethane as well as ethylene above the solution; there appeared to be a net evolution of gas during the slower fading reaction. As in the case of the HCl plus hydride reaction + -1 under vacuum, neither H nor CI added in excess increased the rate of fading. However, a first^order dependence on HCl concentration was observed, and the calculated values of (pseudo-first-order rate constant)/[HCl] were similar to the k*V/[HCl] values for the reaction under vacuum (Table XII). 4 .4 The Overall Equilibrium The overall equilibrium constant K for the reaction RuCl„(PPh0) + H- RuClH(PPh„) + HCl ( 4 . 5 ) A o n z - i — j n where n = 2 or 3, was determined by optical density measurements at - 68 -520 nm, both with and without added PPh^ in the solutions. Good isosbestic points were observed in the recorded spectra, at 406 nm and 434 nm for solutions containing no added phosphine (Figure 23), and at 638 nm for solutions containing a high PPh^ concentration (Figure 24). Molar extinction coefficients were calculated from absorbance measurements on the limiting solutions containing (i) HC1 and no R^ , and (ii) 1 atm and no HC1, when the hydride species was completely formed. For the hydride species, = 1630 (independent of PPh^ concentration) and for the dichloro species £520 = ^® ^ a t '•*>*>b3-'a = ^ and 730 (at [PPh0] = 0.146 M). No RuCl(PPh„)„+ was expected to be j a 5 1 present in the solutions containing HC1, since HC1 is partially ionized + - - 1 2 - 1 -2 to H and Cl in DMA solution (A = 8.0 ohm cm mol at 2.0 x 10 M HC1 and 25°), and equilibrium (3.7) is completely to the left in the presence of excess chloride ion (Section 3.3). The total concentration of HC1 was determined as and at high added HC1 concentration was approximated by [HC1] . The c l concentrations of dichloro and hydride species were calculated using an equation such as (3.5). The overall measured equilibrium constant K is given by [HC1]T = [HCl] a + [Ru-H] (4.6) K = [Ru-H][HC1] [Ru-C1][H2] (4.7) where Ru-H and Ru-Cl represent the hydrido-chloro and dichloro species, O.D. r. i l I | 400 500 , 600 700 X , nm Figure 23. Visible absorption spectra of DMA solutions of RuC^CPPh^)^ containing at 25°. (3.30 x 10"2 M HCl, 0.81 x 10~3 M Ru 1 1). [H 2]: (A) 0.0 M; (B) 0.27 x 10~3 M; (C) 0.81 x 10 _ 3 M; (D) 1.63 x 10~3 M. - 71 -respectively. The value of K at 25° in the absence of added phosphine was determined from a plot of [Ru-H]/[Ru-Cl] vs [H2] (Table XIV, Figure 25: K = slope[HCl] = 20.6) and from a plot of [Ru-H]/[Ru-Cl] vs. [HC1] T _ 1 (Table XIV, Figure 26: K = slope/[H2] = 15.0). A value of K at high added phosphine concentration was estimated to be 309 at 25°, using equation (4.7) for one experiment at 128 mm H 2 and an added HC1 -2 concentration of 2.76 x 10 M (Table XV). The temperature dependence of K was determined both in the presence and absence of added PPh^ (Table XV). 4.5 Discussion 4.5.1 The Effect of Excess Triphenylphosphine on the Forward Reaction In contrast to results obtained on DMA solutions of RuCl^PPh^)^ containing no added phosphine (discussed in Section 4.5.3), the hydrogen reaction rate at 0.146 M PPh^ shows a first-order dependence on H 2 concentration and also no dependence on chloride ion concentration (Table X), indicating that ionic species such as RuCl (PPh.j)3+ are not important kinetically. In the presence of a 2-fold excess of LiCl, DMA solutions of RuCl 2(PPh 3) 3 (species I) contain RuCl2(PPh^)2 (species II), as discussed in Section 3.3. The decrease in the rate of the hydrogen reaction with increasing concentration of added phosphine (Table IX, Figure 15) is concomitant with an increase in the species ratio [I]/[II] resulting from a shift in equilibrium (3.1). The solution reactivity with H2 becomes very low at high concentration of PPh^, indicating that the reactivity of species I is negligible compared to that of II. The rate of reaction may therefore be used as an indirect i - 72 -Table XIV. Spectrophotometry investigation of equilibrium reaction (4.5). Dependence of 0.D.^ 2Q o n concentrations of HCl and H2 at 25° • [Ru11] = 0.81 x io" -3 M [HC11 a x 102,M [HC1]T* x 102,M H2 mm [H2] x 103,M °-D-520 [Ru-H] + [Ru-Cl] 3.30 3.30 0 0.00 0.43 0.00 3.30 3.31 128 0.27 0.55 0.16 3.30 3.32 245 0.52 0.65 0.32 3.30 3.33 384 0.81 0.73 0.51 3.30 3.33 509 1.07 0.80 0.71 3.30 3.34 638 1.34 0.85 0.87 3.30 3.34 775 1.63 0.89 1.07 0.26 0.30 128 0.27 0.94 1.35 0.51 0.55 128 0.27 0.82 0.78 0.77 0.79 128 0.27 0.74 0.54 1.10 1.13 128 0.27 0.67 0.37 [HC1]T = [HCl] + a [Ru-H]. Calculated as explained in text (Section 4.4). - 73 -1.0 [H2] x 103, M Figure 25. Plot of [Ru-H]/[Ru-Cl] vs. [H ] for equilibrium (4.5) in DMA at 25° (0.81 x 10~3 M Ru1 , 3.30 x 10~2 M HCl) (Table XIV). 0.0 0.0 1.0 [HCl], 2.0 L x 10" 3.0 M Figure 26. Plot of [Ru-H] / [Ru-Cl] vs. [HCl] " for eciuilibrium (4.5) -3 1 in DMA at 25° (0.81 x 10 M Ru 1 1, 0.27 x 10 M H^. (Table XIV). - 74 -Table XV. Spectrophotometric investigation of equilibrium reaction (4.5). Temperature dependence of K. [Ru11] = 0.81 x 10~3 M. Temp. O.D.520 * K °C x 10 - 1 25 0.89a 2.2 30 0.8l a 1.6 35 0.753 1.2 40 0.70a 0.9 25 1.14b 31 30 1.08b 22 35 1.04b 17 40 0.99b 13 Calculated according to Equation (4.7). a {HCl] = 3.30 x 10"2 M, [H_] = 1.60 x 10~3 M. b [HCl] = 2.76 x 10"2 M, [H_] = 0.27 x 10~3 M, [PPh„] = 0.146 M. 3. /\ J 3. 9. - 75 -measure of the relative concentrations of species I and II in the solution. The maximum reaction rate (recorded for a solution containing a 2-fold excess of LiCl and no added phosphine) is a measure of the reactivity of species II, since the phosphine dissociation from I is complete in the absence of added PPh^ - The relative concentrations of I and II may therefore be calculated using the measured values of k": [II] _ k" ns Fxi ~ k"» _ k» v^.o; max Values of the ratio [II]/[I] calculated at the various phosphine concentra-tions are summarized in Table IX, along with the total concentration of free PPh^ calculated from [PPh 3] T = [PPh 3] a + [II] ' (4.9) A plot of [II]/[I] against [PPh.j]T 1 gives a line with slope = = -2 2.2 x 10 M (Figure 27). This value of is in reasonable agreement _2 with the spectrophotometric value of 4.6 x 10 M (Section 3.3). Using these values, the concentration of species II calculated to be present in the solution at 1.0 M PPti3 is sufficient to completely account for the observed rate of reaction (Table IX), verifying that species I is in fact essentially unreactive towards Hj. As in DMA solution, addition of PPh^ to solutions of RuC^^Pl^^ in 1:1 benzene-ethanol decreased the rate of formation of the violet hydride in the hydrogen reaction; this result suggests that phosphine dissociation from species I occurs in benzene-ethanol as well as in DMA solution. - 76 -[H ]/[ I ] 2.0 h 0.0 [PPh 3] T - 1 x 10"2,M-1 Figure 27. Plot of [II]/[I] vs. [ P P h ^ - 1 for DMA solutions of RuCl 2(PPh 3) 3 at 25° (0.81 x 10~3 M Ru 1 1, 1.95 x 10~3 M LiCl, 0.29 x 10~3 M H2) (Table IX, Figure 15). - 77 -4.5.2. Reaction of RuCl(PPh.p 2 + with Hydrogen In the spectrophotometry determination of for the reaction K2 + RuCl„(PPh0)0 — ^ RuCl(PPh_)_ + CI (4.10) z 5 2. «« 5 Z II III in DMA, a 2-fold excess of LiCl was sufficient to completely suppress the loss of chloride ion by species II (Section 3.3). This result can be correlated with the effect of LiCl on the reaction rate of with these solutions (Figure 19), with the conclusion that species III is significantly more reactive towards H^  than is species II. The non-linearity of the logarithmic plots (Figure 17) towards the end of the reaction (with no added LiCl) is indicative of an increasing species ratio I l l l/Il l l] , probably as a result of the production of ionic chloride by the hydrogen reaction (HCl being significantly ionized at low concentration). The reaction of species III, which accounts for ^ 80% of the solution reactivity in the absence of excess chloride, shows first-order dependence on H2 concentration (Figure 18). A rate expression for the reaction of III with hydrogen is therefore given by Rate = k 1[H 2][HI] (4.11) At low hydrogen pressure, the reaction of II is also first-order in hydrogen concentration, so that a rate expression for species II is Rate = k 2[H 2][II] (4.12) - 78 -The reaction rate, at low hydrogen concentration, of a solution contain-ing a mixture of species II and III is given by Rate = k 1[H 2][III] + k 2[H 2][II] (4.13) The experimentally determined psuedo-first-order rate constant k"' is defined by Rate = k"'[Rui:i:]T (4.14) From Equations (4.13) and (4,14) is obtained the relationship k-'fRu11] - k [H ][II] k l " [ H 2 ] [ I I I ] ( 4 - 1 5 ) Values for k 2 were determined at low hydrogen pressure for the temperature range 10-25°C (Table VIII), and yielded the activation t + parameters AH = 12.8 + 0.8 kcal/mole, AS = -11.5 + 2.9 eu. The equilibrium constant K2 was previously measured over this temperature range (Table IV, Figure 11), allowing calculation of the i n i t i a l concentrations of species II and III in the reacting solutions. Values of k^  were calculated from Equation (4.15) (Table XVI), and an + Arrhenius plot (Figure 28) yielded the activation parameters AH = 12.1 + 0.7 kcal/mole, AS+ = -10.4 + 2.4 eu. Studies on the dehydrochlorination of IrH2Cl(CO)(PPh^)2 in the presence of bases"*7 revealed that heterolytic splitting of hydrogen could occur via such a dihydro intermediate, and led to the suggestion - 79 -Table XVI, Reaction of RuCl 2(PPh 3) 3 in DMA with H2. Temperature dependence of k^. [Ru11] . 0.81 x 10"3 M, [H2] = 0.21 x 10~3 M. • Temp. * [n] t + t t [III] k"' °C x 105,M x 104,M x 104,M x 10 3,s - 1 x 10-1,M 1s 1 10 5.80 6.19 1.90 1.08 1.49 15 6.20 6.14 1.95 1.66 2.15 20 6.63 6.08 2.01 2.39 3.09 25 7.04 6.03 2.06 3.80 4.77 * Values interpolated from plot of log K2 vs. T * (Figure 11). Calculated from given K2 values. Calculated according to Equation (4.15). - 80 -3.40 3.50 -1 3 -1 T x 10 , °K Figure 28. Arrhenius plot for the reaction of RuCl(PPh3)2 with H in DMA (Table XVI). - 81 -that oxidative addition of hydrogen to give a Ru(IV) dihydride occurred as an intermediate step in the base-promoted reaction of with RuCl 2(PPh 3) 3 in benzene solution. The existence of the Ru(IV) complex RuH^(PPh^)'"^ provides some further evidence for such intermediates. Activation parameters similar to those calculated for the present system (i.e. a moderate activation enthalpy and moderate negative activation entropy) have been determined for the oxidative addition of to complexes of the type IrX(CO)(PPh^),^ where X is . ,. , 60,61 . . . . an anionxc ligand. A three-centre transition state, H H \ i Ph P \ / „ CO Ir PPh3 in which both hydrogen atoms interact simultaneously with the metal 61 ion, has been proposed for these reactions. The reaction of RuCl(PPh3)2 with hydrogen thus possibly involves oxidative addition of hydrogen k RuCl(PPh3)2 + H2 RuClH 2(PPh 3) 2 + (4.16) followed by reductive elimination of a proton in a fast step, RuClH 2(PPh 3) 2 + > RuClH(PPh3)2 + H+ (4.17) - 82 -The negative activation entropy for oxidative addition reactions such as (4.16) reflects mainly the loss of translational entropy of 61 t the hydrogen molecule upon coordination. Much larger negative AS values have been obtained for reactions involving oxidative addition of methyl iodide^ and silicon hydrides^2 to iridium(I) complexes. This result has been attributed to increased solvation of a highly polar transition state. 4.5.3 Reaction of RuCl^ (PPh.^ )^  with Hydrogen The rate of reaction of RuC^CPPh^^ with hydrogen approaches a limiting value at high concentration of hydrogen (Figure 14). Inverse plots of vs. rj| j (Table VIII, Figure 29) were linear, and allow calculation of a k' from the intercept: k' is the first-order max max rate constant for the reaction at high hydrogen concentration. An Arrhenius plot (Figure 30) using the temperature dependence data on + k' (Table XVII) yielded the activation parameters AH = 13.2 + max 3 r — 0.5 kcal/mole, AS+ = -24.2 + 1.7 eu. The non-linear hydrogen dependence of this reaction could result from an i n i t i a l rapid equilibrium involving formation of a dihydride intermediate species in appreciable concentration (cf. Equations 4.16 and 4.17 above): RuCl.(PPh0). + H„ RuCl0H_(PPhQ)„ RuClH(PPh.)_ + HCl 2 5 2 2 -« I 2. 5 2 5 2 (4.18) Such a mechanism gives rise to the rate law - 84 -Table XVII. Reaction of RuCl 2(PPh 3) 3 in DMA with H2. Temperature dependence of k' max [Ru11] = 0.81 x 10"3 M, [LiCl] = 1.65 x 10"3 M. Temperature k* max °C i n 3 "! x 10 , s 10 2.0 15 3.1 20 4.8 25 6.3 k' values determined from Figure 29. max - 85 -- 86 -kK[H JtRu 1 1] R a t e = 1 + K[H2] ( 4* 1 9> which will account for the observed first-order hydrogen dependence at low [H 2], and zero-order dependence at high [H2] when the dihydride is fully formed. However, gas uptake experiments at 25° and 1 atm H2 indicated that there was no rapid i n i t i a l hydrogen uptake by solutions containing RuCl2(PPh^)2, although such i n i t i a l uptake would be required by this mechanism; spectroscopic results also gave no evidence for the presence of a rapidly formed intermediate species in the reaction. A mechanism involving pre-dissociation of chloride, followed by reaction of the dissociated species with hydrogen RuCl~(PPh,)_ RuCl(PPh_)0+ + Cl~ RuClH(PPh0)0 + HCl 2 O Z -r j Z. H 2 j Z (4.20) gives rise, via a steady-state treatment on the reactive intermediate RuCl(PPh 3) 2 +, to the rate law kk [RuCl (PPh ) ][H ] Rate = — — — (4.20a) k_[Cl~] + k[H2] It should be noted that this steady-state treatment can apply only to systems containing added chloride, where the concentration of RuCl(PPh 3) 2 + is low. The inverse chloride dependence required by this mechanism is observed only £. low chloride ion concentration (Figure 19), and the lower limiting rate attained at high chloride concentration must result - 87 -from the reaction of RuCl 2(PPh 3) 2 itself with hydrogen. Thus, in the systems containing added chloride, the rate of reaction might be given by an expression such as kk [RuCl (PPh ) ] [ H ] Rate — ± — — + k [RuCl (PPh ) ] [H ] (4.20b) k_[Cl ] + k[H2] l L J which will qualitatively explain the chloride dependence as well as the less than first-order hydrogen dependence.observed at high concentration of hydrogen. However, quantitative analysis suggests that the second term in the rate law is dominant, and the curvature in the hydrogen-dependence plots (Figure 14) is much greater than would be expected on the basis of this mechanism. Salt effects, which could influence the relative importance of the two terms in the rate law, were found to be not significant (Table XI). A mechanism involving pre-dissociation of phosphine RuCl 2(PPh 3) 2 RuCl2(PPh3) + PPh3 (4.21) followed by reaction of RuCl2(PPh3) with hydrogen gives a rate expression similar to (4.20a), using a steady-state treatment on RuCl2(PPh3). Although an inverse phosphine dependence for the hydrogen reaction is in fact observed, in accordance with this mechanism, this dependence was previously explained in Section 4.5.1 in terms of the equilibrium reaction (3.1), and good agreement was observed between the kinetic and spectroscopically determined values of K^ ; inclusion of a second phosphine dissociation reaction (4.21) in the -' 88 -mechanism gives an unsatisfactory analysis of the kinetic data for the phosphine dependence, which includes an inverse square term. Since the starting solution in these reactions (Table VIII, no added phosphine) contains only RuC^(PPh^^ according to the visible spectra recorded, the equilibrium itself (i.e. essentially using a steady-state treatment for RuCl2(PPh3)2) cannot account for the non-linear hydrogen dependence. At this stage, we cannot give a totally satisfactory explanation for the hydrogen dependence results obtained in solutions containing added chloride. 4.5.4 Reaction of RuClH(PPh3)3 with HCl The rate of reaction of HCl with the violet hydride solutions, expressed by either or both of the equations k 3 RuClH(PPh3)2 + HCl —^> RuCl 2(PPh 3) 2 + H 2 (4.22) k4 RuClH(PPh3)3 + HCl — R u C l 2 ( P P h 3 ) 3 + H2 (4.23) is much more sensitive to the presence of low phosphine concentration than is the forward reaction, as can be seen by a comparison of Figures 22 and 15. This result would seem to indicate that there is a large difference in the magnitude of the equilibrium constants for phosphine dissociation in reactions (4.2) and (3.1), and that K3 << K^ . As before, the trisphosphine species appears to be essentially unreactive (i.e. k^ = 0), since the reaction rate with HCl becomes negligible at high concentration of PPh, (Table XII). Although the phosphine dissociation - 89 -of RuCl 2(PPh 3) 3 was complete at 10 3 M Ru 1 1, RuClH(PPh3)3 (species IV) must be only partially dissociated to RuClH(PPh3)2 (species V) under the same conditions because of the apparently lower equilibrium constant, The first-order dependence of k*V on HCl concentration (Table XII, Figure 21) gives rise to the rate expression Rate = k*[HCl][Ru I 3 :] T (4.24) * k 1 V where k = [ H C 1 j (4.25) * The magnitude of k is dependent on the concentration of V in the solution, and k [V] k = -^ -== (4.26) [Ru ] T The ratio of species IV and V in the solution can be expressed in terms * of k and k 3: [VI _ k* TivT ~ k 3 - k* ( 4 ' 2 7 ) Even with no added phosphine in the solution, [V] < [Ru**]^, so that * k 3 > k in this case. An estimated value of k 3 must therefore be used in the calculation, the accuracy of the estimation being indicated by a lack of curvature in a plot of [V]/[IV] vs. [PPh 3] T \ where [PPh 3] T = [PPh 3] a + [V] (4.28) - 90 -The best straight line for this plot is obtained using a value (determined by successive approximation) of 0.85 + 0.05 M *s * (Table XVIII, Figure 31), with the slope of the line giving a value of 3.1 x 10~4 M for K3 at 25°. Using this K3 value, RuClH(PPh3)3 is calculated -4 II to be 46% dissociated at 8 x 10 M Ru , in the absence of added PPh-j' The lack of change in the visible spectrum of the violet solutions upon addition of excess phosphine (Section 4.2.1) must therefore be the result of very similar spectra for species IV and V. The hydrido-chloro complex having three bulky phosphine ligands might be expected to encounter less intramolecular steric strain than the corresponding dichloro complex, possibly accounting for the difference in the magnitudes of and K3« A positive AS0 value was in fact the reason for the relatively large value (Section 3.4) of 4.6 x 10~2 M . The non-reactivity of the trlsphosphine species RuClH(PPh3)3> compared to RuClH(PPh3)2» with HCl is analogous to the lack of reaction between HNi[P(0Et)3]^ and H , although the ligand-deficient species 63 HNi[P(OEt) 3l 3 does react with evolution of hydrogen. From Equations (4.24) and (4.26) is derived the rate expression Rate = k3[HCl][V] (4.29) In terms of total ruthenium, this becomes k K [HClHRu11] R a t e = K3 + [PPh 3] T <*-30> - 91 -Table XVIII. Reaction of RuClH(PPh3)3 in DMA with HCl Dependence of [V]/[IV] on concentration of PPh3 at 25°. [Ru11] = 0.81 x 10"3 M. 3 a [PPh 3] T* [V] + [V]/[IV] + + [PPh 3] T- 1 x 103, M x 103, M x 104, M x 10 -3 -1 x 10 ,M 0.00 0.37 3.71 8.48 2.70 0.27 0.56 2.93 5.68 1.79 0.73 0.92 1.89 3.05 1.10 1.32 ' 1.48 1.57 2.40 0.68 2.40 2.49 0.92 1.28 0.40 4.77 4.82 0.53 0.70 0.21 9.78 9.81 0.27 0.35 0.10 !PPh3JT = [PPh 3] a + [V] • ^ Calculated according to Equation (4. 26") using k 3 = 0. 85 M~1s~1. tt Calculated according to Equation (4. 27) using k 3 = 0. 85 M_ 1s - 1. - 92 -- 93 -Equation (4.30) can be rewritten, in terms of the pseudo-first-order iv rate constant k , as [HCl] 1 ^ [ P P h3 ]T „ - n . I s" " S + "W (4-30a) Thus a plot of [HClJ/k"^ vs. [PPh^]^ should give a straight line with slope l/k^K^ and intercept l/k^* t^S^a m a y D e u s e d as an approximation for the total phosphine concentration, i f [PPh„] >> [V] (see Equation 4.28). A limitation on the use of this inverse plot (Figure 31, inset) is the difficulty in determining the intercept on the ordinate axis, since this intercept is so close to the origin. Approximate values of -1 -1 -4 k^ and obtained from the plot were 0.6 M s and 4.7 x 10 M, respectively, in reasonable agreement with the values determined above. In solutions containing a high concentration of added phosphine, [PPhg] >> K3 and [PPh.^ = [PPhg],^ so that the rate expression (4.30) becomes k K [HCl][Ru11] [PPh_l (4.31) j a The experimentally determined pseudo-first-order rate constant k 1 V can be related to the combined constant k^ K^  using Equation (4.31): k i v[PPh ] k 3 K 3 " - [ H C l l ( 4 ' 3 2 ) A good Arrhenius type plot (Figure 32) was constructed using the calculated values of k-jK-j (Table XIII), with the activation parameters - 94 -for the overall reaction (4.2 and 4.22) being A H = 18.9 + 1.4 kcal/mole, A S ^ = -10.5 +4.9 eu. If the thermodynamic parameters for the i n i t i a l rapid equilibrium are both small and positive (cf. reaction 3.1 in Section 3.4), the activation parameters for the reaction of HCl with RuClH(PPh.j)£ will be slightly less positive than those given above. Examples of reversible oxidative addition of HCl exist for complexes 64 of Rh(I), Ir(I) and Pt(II). As well, oxidative addition of a proton to square-planar Pt(II) phosphine complexes has been proposed as one step 65 66 in isotopic exchange and ligand substitution mechanisms. Oxidative addition of a proton to the complex Ni[P(0Et).j].j yields the hydride + 63 complex HNi[P(0Et 3)] 3 in a reversible reaction. It should be noted that the present reaction is first-order in HCl, but independent of H + concentration. An alternative to an oxidative addition reaction of HCl with RuClH(PPh.j)2 followed by reductive elimination of is concomitant electrophilic attack at the hydride and nucleophilic attack at the metal ion by the polar HCl molecule: - PPh 3 PPh 3 C l Ru H Cl Ru Cl + H 2 (4.33) This type of reaction is known for post-transition metal alkyl compounds. A further possibility is a reversible intramolecular hvdrogen elimination, via a process involving the ortho hydrogen of a phenyl - 95 -ring (see Equation 6.12 in Section 6.4), followed by a rate-determining reaction with HCl: (Ph3P)Ru + HCl RuCl 2(PPh 3) 2 (4.34) The rate of reaction of HCl with the violet hydride appears to be almost unaffected by the presence of ethylene in the solution. Hydrogen which is liberated in reaction (4.22) apparently reacts with the hydrido-ethylene complex RuClH(C2H^) (PPh.j)2 or ethyl complex to evolve ethane (see Equations 5.3-5.5). The rapid hydrogen reduction of ethylene in the presence of the violet hydride is discussed in Section 5.2.5. 4.5.5 The Overall Equilibrium The overall measured equilibrium constant K for reaction (4.5) increases with addition of PPh^ to the ruthenium solutions. Using the data in Table XV, Van't Hoff plots (Figures 33 and 34) gave the thermodynamic parameters: AH° = -11.2 + 0.8 kcal/mole, AS0 = -31.6 + 2.7 eu (with no added phosphine); AH° = -11.1 + 0.9 kcal/mole, AS° = -25.8 + 2.9 eu (at 0.146 M PPh3). High solvation of the polar (and partially ionized) HCl undoubtedly accounts for the large entropy decrease in the reaction.^ 7 - 96 --2.7 log k 3K 3 -2.9 -3.1 -3.3 3.20 3.30 -1 Figure 32. x 103, °K -1 -1 Plot of log k 3K 3 vs. T for the reaction of RuClH(PPh3)3 in DMA with HCl (Table XIII). 1.3 -log K 1.1 -0.9 " Figure 33. 3.40 Van't Hoff plot for equilibrium (4.5) in DMA solution (Table XV). - 98 -In equilibrium (4.5), the measured equilibrium constant K is equal to the ratio k./k , where k_ and k are the second-order rate constants f r f r for the forward and reverse reactions, respectively. Using rate data obtained for the forward and reverse reactions, value of k.. and k can f r be calculated. At 25° and 0.146 M PPhg, k = k"/[H2] =1.1 M _ 1s _ 1 (Table X) and k r = k ± V/[HCl] = 3.2 x 10~3 M~1s~1 (Table XIII), so that K = 340; this K value compares favorably with the value of 309 obtained from spectrophotometric measurements under the same conditions (Section 4.4). At 25° and with no added phosphine, k f - k'/[H2] = 3.1-7.9 M_1s~ (Table VIII) and k r = k i v/[HCl] = 0.38 M~1s~1 (Table XII), so that K = 8.4-21.4; K values obtained spectrophotometrically under the same conditions were 15.0 and 20,6 (Section 4.4). - 99 -CHAPTER V HOMOGENEOUS HYDROGENATION OF OLEFINS USING HYDRIDOCHLOROTRIS(TRIPHENYL-PHOSPHINE)RUTHENIUM(II) AS CATALYST 5.1 Introduction The hydrogenation of olefins using RuClHCPPh^)^ as catalyst was first investigated in benzene and toluene solution by Wilkinson's 8—10 group. Detailed kinetics were not obtained because of several experimental difficulties, including low catalyst solubility, the extreme sensitivity of the catalyst to oxygen, and possible diffusion control in the very rapid reactions. In contrast, hydrogenation studies performed using this catalyst in DMA solution7'"'"*'*2 gave good reproducibility, the catalyst in this case being prepared in situ by reaction of RuCl^'SH^O with H^  at 80° in the presence of a four-fold excess of triphenylphosphine. The hydrogenation of maleic acid in DMA was studied extensively, and a reaction mechanism was postulated to account for the experimental results. This chapter describes the results obtained by extending this work to a study of the hydrogenation of various other olefins. A simple "in situ" preparation of the hydride from RuC^(PPh.j).j and w a s used, instead of the more complicated preparative method from R U C I 3 . 3 H 2 O . The several advantages of this new method of catalyst preparation have been previously discussed in - 100 -Section 4.1. The olefins studied included maleic and fumaric acids and similar compounds, terminal alkenes, internal and cyclic alkenes, and terminal olefins having functional groups. Determination of the importance of steric and electronic effects on the olefin hydrogenation rate was a primary aim of this study, together with the further elucidation of the reaction mechanism. The results of previous studies of this nature using the RhCl(PPh3)3 catalyst have been ambiguous and only partially comprehensible, and reproducibility has 22 often been poor. In addition, i t is difficult to draw any firm 68 conclusions from the incomplete data available in many cases, since the relative i n i t i a l rates reported at some "fixed" conditions for a series of olefins incorporate both olefin complexity constants and rate constants. A few individual olefin systems have been studied in detail, for example using RhCKPPh^, 6 9 RhH(CO) (PP h ^ , 7 0 RuH(OAc) (PPh 3) 3, 4 0 IrCl(CO) (PPh 3) 2, 7 1 and IrH(CO) (PPh^. 7 2 5.2 Catalytic Hydrogenation of Olefins Since the "in situ" reaction of RuCl 2(PPh 3) 3 with E^ ^ -s effectively instantaneous in DMA at 35° (see Chapter IV), i t was not necessary to prepare the hydride complex prior to the addition of olefin to the reaction flask. In trials with maleic acid, i t was determined that this method gave the same hydrogenation rate as was obtained by first preparing RuClH(PPh3)3 and then adding olefin to the frozen violet solution, followed by the actual hydrogen uptake experiment (Table XIX). The rate of olefin reduction was studied under different conditions of temperature, hydrogen pressure, and catalyst and olefin concentrations, - 101 -Table XIX. Hydrogenation of maleic acid and similar olefins. Dependence of i n i t i a l rate on Ru** and olefin concentration at 35°. [H ] = 1.79 x 10"3 M. Olefin [Olefin] [Ru ] Initial rate x 102, M x 103, M x 105, M s~* maleic acid 1.2 0.4 0.18 3.0 1.0 0.56 6.0 2.0 1.39 9.0 3.0. 2.36 12.0 4.0 3.23 15.0 5.0 4.17 1.0 2.0 1.09 2.0 2.0 1.28 4.0 2.0 1.39 10.0 2.0 1.46 10.0 2.0 1.42a 10.0 2.0 1.42b 30.0 2.0 1.37 30.0 2.0 1.37C diethylmaleate 10.0 2.0 1.32 maleic acid monoamide 10.0 2.0 0.0 10.0 2.0 1.39d 0.1 M 1-octene added. b 4.0 x 10~3 M PPh3 added, c H9 added before M.A. d 0.1 M £-toluenesulfonic acid added. - 102 -The i n i t i a l region of the uptake plots was linear, with a decrease in rate being observed as the olefin concentration decreased (especially in cases where the i n i t i a l olefin concentration was low) or sometimes as the catalyst became deactivated. All determinations of hydrogenation rate were made by measuring the slope of the i n i t i a l linear region of the uptake plot. Examples of hydrogen uptake plots recorded during the hydrogenation of several different olefins are shown in Figure 35. Good reproducibility was obtained for solid olefins (+ 2%), with lower reproducibility for liquid olefins (+ 5%). In addition, rates up to 20% lower were recorded using solvent which had been stored under nitrogen for several months. This accounts for differences in some rates reported for olefin-dependence experiments compared to the rate reported for the temperature-dependence experiments. 5.2.1. Hydrogenation of maleic acid, fumaric acid, and similar olefins During the hydrogenation of maleic acid, the color of the catalyst solution was yellow rather than violet. However, the violet color returned towards the end of the hydrogenation reaction, as the substrate concentration decreased. Catalyst solutions containing a very low in i t i a l concentration of maleic acid were orange or red. In general, the catalytic activity of a solution was found to be somewhat decreased after the complete reduction of the substrate. For example, following the reduction ('v 2 hr.) of a solution of 0.1 M maleic acid at 35°, the hydrogen uptake experiment was repeated using the same catalyst solution with more maleic acid added; the rate of hydrogenation at 0.1 M olefin concentration was found to be reduced by 0-0 1.0 _ 2 2.0 3.0 time x 10 , s Figure 35. Rate plots for RuClH(PPh3)3~catalyzed hydrogenation reactions in DMA at 35° (2.0 x 10~3 M Ru 1 1, 1.79 x 10 M H2). Substrate: (O) 1.09 M 1-nonene; (•) 0.60 M 4 -methoxystyrene; ( A ) 0.69 M methyl vinyl ketone. - 104 -12%, compared to the first experiment. Catalyst deactivation evidently occurs even more rapidly at higher maleic acid concentration, since a solution of 0.3 M maleic acid could not be completely hydrogenated, even after many hours, when using a catalyst concentration of 2.0 x -3 -3 10 M. The deactivation of catalyst (at 0.1 M substrate, 2.0 x 10 M Ru) was most marked in the diethylmaleate hydrogenation reaction, where the rate of hydrogen uptake decreased by 50% of its i n i t i a l value within the first 20 minutes of reaction. The yellow solutions containing deactivated catalyst appeared to be reasonably air-stable,in contrast to the extreme air-sensitivity of the original violet catalyst solutions. Diethylmaleate, maleic acid and maleic acid monoamide were found to be hydrogenated at almost identical i n i t i a l rates; a l l gave yellow reactant solutions. However, the monoamide could be hydrogenated only in the presence of added p_-toluenesulfonic acid. Table XIX summarizes the results of hydrogenation experiments using maleic acid, diethylmaleate and maleic acid monoamide as substrates. Using a constant molar ratio ofmaleic acid to catalyst (30:1), the hydrogenation rate was found to be linearly dependent on catalyst concentration, except when using very dilute catalyst solutions (Figure 36). Dependence of i n i t i a l rate on olefin concentration was between zero-and first-order, with the order decreasing as the catalyst concentration increased (Figure 37). Previous, work in this lab showed that the hydrogen dependence of the reaction was first-order. 1 1 Addition of 2 moles of PPh^ per mole of catalyst had no effect on the hydrogenation rate of a 0.1 M solution of maleic acid. As in the case of maleic acid, the color of a hydride solution - 105 -containing fumaric acid was yellow or red, with more red color being evident at lower substrate concentration; the red hydride color was regenerated towards the completion of olefin hydrogenation. The same observation holds true for monomethylfumarate, monoethylfumarate and diethylfumarate; these latter three substrates are hydrogenated at almost identical rates to fumaric acid (Table XX). A lower rate of catalyst deactivation was observed when using fumaric acid and its analogs compared to using maleic acid as substrate. At equivalent substrate concentrations,the hydrogenation rate of fumaric acid was lower than that of maleic acid, although the limiting rates at high olefin concentration appear to be similar (Figure 37). Addition of PPh^ resulted in a decrease in the fumaric acid hydrogenation rate (Table XX). Four other olefins structurally related to fumaric acid were also used as hydrogenation substrates. Crotonic acid and trans-cinnamic acid are related in that one of the carboxyl groups of fumaric acid is replaced by a methyl group and a phenyl group, respectively. Methyl-fumaric acid has a methyl group replacing one of the olefinic hydrogen atoms of fumaric acid, and trans-stilbene has two phenyl groups instead of the carboxyl groups of fumaric acid. At a substrate concentration of 0.5 M, only crotonic acid of these four olefins was reduced at a measurable rate, and this rate was only about 10% of the maximum rate obtained with fumaric acid. In addition, no change in the violet . color of the catalyst solution was observed in the presence of any of these four substrates. Subsequent addition of maleic acid to these systems resulted in the formation of a yellow solution and a subsequent - 106 -Table XX. Hydrogenation of fumaric acid and similar olefins. Dependence of i n i t i a l rate on olefin concentration at 35°. [Ru11] = 2.0 x 10"3 M, [H2] = 1.79 xl0~ 3 M. Olefin [Olefin] x 102, M Initial rate x 105, M s" 1 fumaric acid 3.0 0.61 4.0 0.78 6.0 0.97 10.0 1.10 30.0 1.12 30.0 0.96a monomethylfumarate 10.0 1.05 monoethylfumarate 10.0 1.02 diethylfumarate 10.0 1.05 methyl-fumaric acid 10.0 0.0 50.0 0.0 crotonic acid 10.0 0.0 50.0 0.12 10.0 1.25b 10.0 o . o c trans-cinnamic acid 10.0 0.0 50.0 0.0 trans-stilbene 10.0 0.0 50.0 0.0 10.0 1.30d 3 4.0 x 10 M PPh., added. Subsequent addition after 1 hr of 0.10 M M.A. c 0.10 M _p_-toluenesulfonic acid added. d Subsequent addition after 4 hr of 0.10 M M.A. - 107 -[Ru11] x 103, M 36. Dependence of i n i t i a l hydrogenation rate of M.A. on [Ru ] in DMA at 35° (1.79 x 10~3 M H 2 > [M.A.]/[Ru11] = 30) (Table XIX). 1.5 0.5 -~—o cr-- / -I / i 1 [olefin] x 10, M 37. Dependence of i n i t i a l hydrogenation rates of M.A. and F.A. on [olefin] in DMA at 35° (1.79 x 10 _ 3 M K t 2.0 x 10~3 M Ru11) (Tables XIX and XX). - 108 -hydrogen uptake rate very similar to that measured for maleic acid itself (Table XX). The temperature dependence data for the i n i t i a l hydrogenation rates of maleic acid and fumaric acid at 0.1 M olefin concentration are summarized in Table XXI. 5.2.2. Hydrogenation of terminal olefins The olefinic substrates 1-hexene, 1-octene, 1-nonene and 1-decene were a l l hydrogenated at a rapid rate at 35°, using a catalyst concentra--3 tion of 2.0 x 10 M (Table XXII). The product of reduction was in each case the normal alkane, and no isomeric alkenes were observed in gas chromatographic analysis of the solution during the reaction. The rate of reduction was found to be between zero- and first-order in olefin concentration (Figures 38 and 39), with a limiting rate being approached above 1.0 M olefin concentration. The hydrogenation rate of 1-octene was unaffected by the addition of a four-fold excess of trans-2-octene (Figure 38). At the lower olefin concentrations used, the color of the catalytic solution was violet or red, at least in the i n i t i a l period of the reaction. However, i t was found that these violet solutions gradually faded to orange during the reaction and on standing under hydrogen after completion of the uptake experiment. These color changes, not evident for the systems discussed in Section 5.2.1} indicated that some change was occurring in the nature of the metal complex. The catalytic hydrogenation of a solution containing 0.3 M 1-octene for 10 minutes resulted in a loss of ^30% of the i n i t i a l catalytic activity - 109 -Table XXI. Hydrogenation of maleic acid and fumaric acid. Dependence of i n i t i a l rate on temperature. [Ru11] = 2.0 x 10~ 3 M, [olefin] = 0.10 M. Olefin Temperature [H2] Initial rate k* °C x 1 0 3 , M x 1 0 5 , M s" 1 M - l -1 M s maleic acid 30 1.79 0.71 2.0 35 1.79 1.46 4.1 40 1.79 2.27 6.3 45 1.78 4.24 11.9 fumaric acid 30 1.79 0.57 35 1.79 1.10 40 1.79 1.56 45 1.78 2.56 Calculated using Equation (5.7a). - 110 -Table XXII. Hydrogenation of terminal olefins. Dependence of i n i t i a l rate on olefin and hydrogen, concentrations at 35°. [Ru11] = 2.0 x 10~3 M. Olefin [Olefin] M H2 mm x [H2] 103, M Initial rate 3 -1 x 10 , M s 1-hexene 1-octene 1-nonene 1-decene 0.40 748 1.77 0.66 0.52 746 1.77 0.84 0.75 741 1.75 0.98 1.24 731 1.73 1.15 1.24 731 1.73 0.43a 0.29 755 1.79 0.84 0.53 755 1.79 1.25 1.02 754 1.79 1.71 1.02 354 0.84 0.68 0.04b 756 1.79 0.18 0.06b 756 1.79 0.28 0.08b 756 1.79 0.32 0.12b 756 1.79 0.44 0.23b 756 1.79 0.69 0.23b 756 1.79 0.17a 0.13 756 1.79 0.39 0.20 756 1.79 0.53 0.36 756 1.79 1.00 0.43 756 1.79 1.09 0.63 755 1.79 1.29 1.09 755 1.79 1.64 1.09 755 1.79 0.61a 0.35 756 1.79 1.14 0.45 756 1.79 1.36 0.66 756 1.79 1.55 1.16 756 1.79 1.82 1.16 756 1.79 0.74a 1.16 756 1.79 0.10c 3 4.0 x 10~3 M PPh3 added. k Containing a 4-fold excess of trans-2-octene. C 25.6 x 10~3 M PPh3 added. - I l l -It : 1 1_ 0.0 0.5 1.0 [olefin], M 38. Dependence of i n i t i a l hydrogenation rates of 1-decene and 1- octene on [olefin] in DMA at 35° (1.79 x 10~3 M H_, 2.0 -3 II 10 M Ru ) . ( A ) 1-octene plus a 4-fold excess of 2- octene (Table XXII). .0 0.5 1.0 [olefin], M re 39. Dependence of i n i t i a l hydrogenation rates of 1-nonene and 1-hexene on [olefin] in DMA at 35° (1.73 x 10~3-1.79 x 10' H2, 2.0 x 10"3 M Ru11) (Table XXII); - 112 -of the solution as determined by using the same solution in a second experiment with maleic acid added as substrate. The presence of excess triphenylphosphine (PPh^rRu = 2:1) decreased the rate of terminal olefin reduction by a factor of 2.5-2.7 for solutions above 1.0 M concentration in olefin (Table XXII), and also appeared to decrease the rate of fading of the violet catalyst solutions. Table XXIII summarizes the temperature dependence data for the hydrogenation rates of the four terminal olefins. These experiments were performed at high concentration of alk-l-ene (> 1.0 M) except for 1-octene, which was studied at 0.23 M concentration in the presence of a four-fold excess of _trans-2-octene. 5.2.3 Hydrogenation of cyclic olefins, and trans_-2-octene Cyclohexene, cyclooctene and trans-2-octene were hydrogenated more slowly than were terminal olefins (Table XXIV). In a l l three cases, the color of the catalytic solution containing the olefin was violet, although some fading of the color was sometimes evident after the completion of an uptake experiment, especially at high olefin concentration. Trans-2-octene was available only as a mixture containing 20% of 1-octene; an uptake plot using this olefin mixture as substrate is shown in Figure 40. The i n i t i a l hydrogen uptake by the solution is rapid, but a second distinct region of much slower uptake is also observed. The amount of hydrogen rapidly absorbed corresponds to the quantity of 1-octene present in the solution, indicating that the subsequent slow uptake is due to reduction of the trans-2-octene. The hydrogenation rate of this substrate showed a - 113 -Table XXIII. Hydrogenation of terminal olefins. Dependence of in i t i a l rate on temperature. [Ru11] = 2.0 x 10"3 M. Olefin [Olefin] M [ H 2 ] x 10- M Initial rate 3 -1 x 10 , M s k* x 10~ 2, M~1s"1 1-hexene 1.24 30 35 40 45 1.75 1.73 1.71 1.69 1.03 1.34 1.84 2.23 2.94 3.88 5.38 6.60 1-nonene 1.09 30 35 40 45 1.79 1.79 1.78 1.78 1.47 2.00 2.54 3.23 4.10 5.59 7.14 9.07 1-decene 1.16 30 35 40 45 1.79 1.79 1.79 1.78 1.68 2.15 2.72 3.15 4.70 6.00 7.60 8.85 1-octene a 0.23 30 35 40 45 1.79 1.79 1.78 1.78 0.53 0.69 0.81 0.96 Calculated using Equation (5.7a), In the presence of 0.93 M trans-2-octene. - 114 -Table XXIV. Hydrogenation of cyclic olefins and trans-2-octene. Dependence of i n i t i a l rate on olefin and hydrogen concentrations at 35°. [Ru11] = 2.0 x 10~3 M. Olefin [Olefin] M H2 mm [H2] x 103, M Initial rate 5 -1 x 10 , Ms cyclooctene 0.25 756 1.79 13.6 0.42 755 1.79 24.0 0.46 755 1.79 24.0 0.62 755 1.79 27.7 0.73 755 1.79 31.4 1.08 754 1.79 41.7 1.08 754 1.79 6.4a cyclohexene 0.49 751 1.78 0.61 1.03 746 1.77 1.18 1.48 741 1.75 1.53 1.03 326 0.77 0.41 trans-2-octene 0.44 755 1.79 0.74 0.80 754 1.79 1.41 a 4.0 x 10 3 M PPh0 added. - 116 -first-order dependence on olefin concentration up to the maximum concentration used (0.80 M) , while the two cyclic olefins showed less than first-order olefin dependence at high concentration of substrate (Figures 41 and 42). However, the curvature observed in the cyclohexene plot was very slight. The hydrogenation of the three substrates was also followed using gas chromatographic analysis of the reaction solution. No isomeric octenes were observed as products in the trans-2-octene system; the only product was that of reduction, n-octane. Temperature dependence data for the cyclooctene reduction are summarized in Table XXV. 5.2.4 Hydrogenation of terminal olefins having functional groups Of the olefins a l l y l alcohol (CH2=CH-CH2OH), 4-methoxystyrene (CH2=CH-C6H^-OCH3), methyl vinyl ketone (CH2=CH-C0CH3) and acrylamide (CH2=CH-CONH2), a l l but al l y l alcohol were found to be hydrogenated at rates somewhat less than those observed for straight-chain terminal olefins at corresponding conditions (Table XXVI). The rate of hydrogenation of 4-methoxystyrene approached a limiting value at high olefin concentration (Figure 42) whereas a limiting rate was attained even at very low concentration of substrate in the case of methyl vinyl ketone and acrylamide. The reduction of these latter two substrates was not inhibited by addition of a two-fold excess (PPh3:Ru) of phosphine; the methoxystyrene reduction showed a marked inhibition (Table XXVI). The color of the hydride catalyst was yellow or orange in the presence of the alcohol, ketone and amide substrates, but remained violet in solutions containing 4-methoxystyrene. - 117 -co X 4-1 Pi ca •H 4-1 •H c 6.0 3.0 0.0 4-methoxystyr &p& ^ — cyclooctene 1 0.0 1.0 0.5 [olefin], M Figure 42. Dependence of i n i t i a l hydrogenation rates of 4-methoxystyrene _3 and cyclooctene on [olefin] in DMA at 35° (1.79 x 10 M H -3 II 2.0 x 10 M Ru ) (Tables XXIV and XXVI). 2' - 118 -Table XXV. Hydrogenation of cyclooctene. Temperature dependence of k. [Ru11] = 2.0 x 10~3 M, [cyclooctene] = 1.08 M. Temp. °C [H2] x 103, M Initial rate x 105, M s" 1 k* x 10"2, M"1s"1 30 1.79 31.4 0.88 35 1.79 41.7 1.17 40 1.78 49.7 1.40 45 1.78 62.0 1.74 * Calculated using Equation (5.7a). - 119 -Table XXVI. Hydrogenation of terminal olefins having functional groups. Dependence of i n i t i a l rate on olefin concentration at 35°. [Ru11] = 2.0 x 10~3 M. Olefin [Olefin] H2 [H2] Initial rate M mm x 103, M 4 -1 x 10 , M s all y l alcohol 0.40 754 1.79 8.06 * 0.50 754 1.79 7.39 0.75 753 1.78 6.63 1.00 752 1.78 4.81 4-methoxystyrene 0.30 756 1.79 2.48 0.60 756 1.79 4.07 1.10 756 1.79 6.36 1.10 756 1.79 1.42a methyl 'vinyl ketone 0.25 754 1.79 0.98 0.40 752 1.78 0.98 0.69 750 1.78 1.00 0.25 754 1.79 0.94a acrylamide 0.03 756 1.79 0.98 0.04 756 1.79 1.01 0.06 756 1.79 1.08 0.10 756 1.79 1.15 0.20 756 1.79 1.24 0.30 756 1.79 1.17 0.50 756 1.79 0.96 0.20 756 1.79 1.373 a 4.0 x 10 M PPh„ added. i - 120 -The violet color returned in the acrylamide solutions as this substrate approached complete reduction. However, at the higher acrylamide concentrations (> 0.2 M) the i n i t i a l hydrogenation rate became less than the limiting rate (Table XXVI), suggesting increased catalyst deactivation under these conditions. After complete reduction _3 35 min) of a 0.1 M acrylamide solution using 2.0 x 10 M catalyst at 35° the catalyst activity was decreased by 30%, as determined by repeating the hydrogen uptake experiment after addition of more substrate to the solution. Under the same conditions, using methyl vinyl ketone as substrate,the catalyst activity was reduced by 90%, and the final solution color after complete reduction of the substrate was bright yellow rather than violet. Allyl alcohol was found to react with and deactivate the dichloro complex RuCl2(PPh3)3; for example, allowing the ruthenium complex to stand in a DMA solution of a l l y l alcohol (0.1 M) for 30 minutes at room temperature before the addition of hydrogen resulted in a 50% decrease in the i n i t i a l hydrogenation rate. For this reason, the hydrogen uptake rates measured using this substrate were not very reproducible, and the rapid catalyst deactivation also resulted in decreased hydrogenation rates with higher olefin concentration (Table XXVI). The yellow catalyst solution resulting after nearly complete reduction ( ^ 30 min) of a 0.2 M a l l y l alcohol solution at 35° and -3 2.0 x 10 M Ru showed only 1% of the catalytic activity of the i n i t i a l violet solutions. The variation of hydrogenation rate with temperature was determined for the olefins other than a l l y l alcohol, and these results are summarized In Table XXVII. - 121 -Table XXVII. Hydrogenation of terminal olefins having functional groups. Temperature dependence of k. [Ru 1 1] = 2.0 x 10 3 M. Olefin [Olefin] T [HJ Initial rate k* M °C z xl03,M 4 -1 xlO ,M s xl0~2,M - 1s" 1 4-methoxystyrene 1.10 30 1.79 4.60 1.28 35 1.79 6.36 1.78 40 1.79 8.62 2.41 45 1.78 11.60 3.26 methyl vinyl ketone 0.25 30 1.79 0.66 0.18 35 1.79 0.98 0.27 40 1.78 1.38 0.39 45 1.78 1.79 0.50 acrylamide 0.20 30 1.79 0.89 0.25 35 1.79 1.24 0.35 40 1.79 1.69 0.47 45 1.78 2.29 0.64 * Calculated using Equation (5.7a). - 122 -5.2.5 Hydrogenation of ethylene and butadiene A reproducibility problem arises in uptake experiments performed using a mixture of two gases because of possible incomplete mixing of the gases. For this reason, no detailed kinetics were attempted in the investigation of ethylene and butadiene hydrogenations. However, ethylene was found to be reduced very rapidly at 35°, and butadiene at a slower rate. In an experiment using a catalyst -4 concentration of 5.2 x 10 M and 0.5 atm. pressure of each of hydrogen and ethylene (JH^ = 0.90 x 10~3 M, T C ^ J = 0.02 M) , the observed - 4 - 1 rate of hydrogen uptake was 1.4 x 10 Ms . Assuming a first-order -3 dependence on catalyst concentration, the rate at 2.0 x 10 M Ru -4 -1 would be 5.4 x 10 Ms , which is comparable to the rate of hydrogenation observed for terminal olefins at <v 0.2-0.3 M substrate concentration. The color of the catalytic solution during the hydrogenation of ethylene was yellow ( E ^ Q = 640) but the violet color of the hydride catalyst returned within a few seconds i f the solution were not vigorously shaken to maintain the equilibrium concentrations of the two gases in the solution. The catalyst was slowly deactivated and, for example, repressurizing the system with the two gases after several hours resulted in an uptake rate less than 20% of the i n i t i a l rate -4 (at 5.2 x 10 M Ru). The deactivated solution was yellow, and exhibited no color change either on treatment with hydrogen or on exposure to air. The catalyzed hydrogen reduction of ethylene to ethane was also followed in a qualitative manner using gas chromatographic and mass spectrographic analysis of the gas mixture above the catalytic solution. The reaction of ethylene alone with RuClH(PPh_)„ is discussed - 123 -in Sections 4.3 and 6.3. Ethylene was also catalytically reduced with deuterium, using _3 0.5 atm. of each gas over a 7 x 10 M DMA solution of RuCl^CPPh^)^; after 24 hr at 35°, mass spectrographic analysis of the gas above the solution revealed that the r eaction product contained a mixture of polydeuterated ethanes, including even a small amount of C„D-. 1,3-Butadiene was hydrogenated more slowly than ethylene under comparable conditions, even though the concentration of butadiene in the solution was much higher than that of ethylene due to its greater solubility in DMA. During hydrogenation, the color of the catalytic solution was yellow; the violet color of the catalyst could be regenerated by removal of the gas mixture and treatment with 1 atm. of hydrogen. Gas chromatographic analysis of the gas phase reaction products revealed that this catalytic reduction of butadiene produced cis-2-butene, and that this product was then further reduced to butane at a rate of less than 10% that of the butadiene reduction. 5.3 Reaction of acetylene with RuClH(PPh3)3 Gas chromatographic analysis indicated that a very small amount of ethylene was ini t i a l l y produced by reaction of an acetylene-hydrogen mixture with a DMA solution of RuClH(PPh3>3 at 25°. Acetylene reacted rapidly and irreversibly with the catalyst to produce a brown solution; the violet color of the original hydride complex could not be regenerated by treatment of this solution with hydrogen. The brown solution gradually decomposed under C 2 H 2 or in air, with deposition of metal on the sides of the reaction flask. An ir spectrum on some brown - 124 -solid recovered from a solution after 2 days at 25° under a H2~^2^2 mixture showed the presence of an intense, broad carbonyl band at 1930 cm \ indicating that a Ru-carbonyl complex was produced. Ir spectra recorded immediately after the reaction of C2^2 w ^ t n a solution of 73 the hydride showed bands due to coordinated acetylene (3160, 1947, 790 cm"1). 5.4 Catalytic hydrogenation using other ruthenium(II) complexes 1-Nonene and maleic acid were used as substrates for hydrogenation 39 40 reactions catalyzed by RuBr 2(PPh^, RuH(OAc) (PPh^, Ik 7S 3Q RuCl2(C0)(PPh3)2, J and cis-RuCl 2(CO) 2(PPh 3) 2, and the results of these experiments are summarized in Table XXVIII. The violet solution formed by reaction of RuBr2(PPh.j)3 with hydrogen in DMA was of comparable efficiency to RuClH(PPh3)3 for catalyzing olefin hydrogenation. The complex RuH(OAc)(PPhg)^ was of somewhat lower efficiency, and the two carbonyl complexes were found to be inactive as hydrogenation catalysts under the conditions used. 5.5 Discussion Studies conducted by Wilkinson's group1^ on the hydride catalyst RuClH(PPh3)3 in benzene solution revealed a very high selectivity for the hydrogenation of terminal olefins, with internal and cyclic 3 olefins being reduced more slowly by a factor of at least 10 . This hydride complex is apparently the most active yet discovered for the catalytic hydrogenation of alk-l-enes. Relative i n i t i a l rates were determined for several terminal and internal olefins as well as cyclohexene, - 125 -Table XXVIII. Catalytic hydrogenation of olefins using Ru*1 complexes. Initial rates at 35°. [Ru11] = 2.0 x 10~3 M, [H2] = 1.79 x 10~3 M. Catalyst Olefin* Initial rate x 105, M s" 1 RuClH(PPb3)3 1-nonene M.A. 200 1.46 RuBrH(PPh3)3 1-nonene M.A. 229 0.74 RuH(OAc)(PPh3)3 1-nonene M.A. 68.3 0.60 RuCl2(CO)(PPh3)2 1-nonene 0.0 RuCl 2(C0) 2(PPh 3) 2 1-nonene 0.0 * [1-nonene] = 1.09 M, [M.A.] = 0.10 M. From Table XXIII. From Table XIX. - 126 -but no olefins having functional groups were studied. Data obtained -4 at 25° (using 8.3 x 10 M catalyst, 1.2 M olefin concentration and 500 mm H^ ) are given below: Alkene Rate x 10^, M s * 1-pentene 300 1-hexene 340 1-heptene 210 1- decene 280 cyclohexene < 0.2 2- hexene < 0.06 2- octene < 0.06 3- heptene < 0.06 2-methyl-l-pentene < 0.06 Investigations*^ with 1-octene revealed that a limiting rate was approached at high olefin concentration and in addition, excess triphenylphosphine severely inhibited the hydrogenation of 1-heptene. However, no other kinetic data were obtained due to the experimental difficulties mentioned earlier (Section 5.1). The nature of the observed dependence of rate on olefin concentration and added triphenyl-phosphine, and the selectivity towards terminal alkenes were similar to the results obtained7^ using the catalyst RhH(CO)(PPh^)^, which had previously been studied in detail, and an analogous mechanism was therefore postulated. The proposed mechanism involved dissociation of the catalyst to give a square monomer RuClH(PPh.j)2 having trans-PPh^ groups (Equation 5.1), coordination (Equation 5.2) and insertion of the II olefin into the Ru-hydride bond giving a Ru -alkyl complex (Equation 5.3), the rate-determining oxidative addition of hydrogen to produce a Ru*^ complex (Equation 5.4), and reductive elimination of alkane to - 127 -regenerate the hydride complex (Equation 5.5). K l RuClH(PPh3)3 - RuClH(PPh3)2 + PPh3 (5.1) K2 RuClH(PPh3)2 + alkene >- RuClH(PPh3)2 (alkene) (5.2) K3 RuClH(PPh„)0(alkene) — ^ RuCl(alkyl)(PPh„)_ (5.3) RuCl (alkyl) (PPh 3) 2 + H2 H2RuCl (alkyl) (PPh ) 2 (5.4) H2RuCl(alkyl) (PPh 3) 2 »- RuClH(PPh3)2 + alkane (5.5) The high selectivity towards alk-l-enes was proposed to arise from steric interactions between the olefin and the bulky phosphine groups during the alkyl formation step. As illustrated in Scheme I, hydrogen transfer through the four-centre transition state cannot readily occur unless R ' = H . R R R H Ph-P H ' / . H P H P Y/\ J J p h , p v / ° \ Ru . » ^•Ru""' *- Ru / \ ^ ~ / \ ^ ~ / \ Cl PPh3 Cl PPh3 Cl PPh3 Scheme I A similar argument has been used to explain the selectivity of the catalyst RhH(CO) (PPh 3) 3 in hydrogenation reactions,7^* and the selectivity observed in olefin isomerization systems using RuCl 2(PPh 3) 3 in the presence of air or hydroperoxides, IrCl(CO)(PPh3)2, and PtCl 2(PPh 3) 2-SnCl 2-H 2. 7 6 - 128 -10 Evidence for alkyl complex formation was obtained using a CDCl^ solution of RuClH(PPh3)3,where the species (PPh^ClRu-C^ was detected under high ethylene pressure. However, there was no evidence of chemical reaction when benzene solutions of RuClH(PPh3)3 were treated with alkenes (including C2**4 a t ^ a t m*)» indicating that the alkyl equilibrium (Equation 5.3) was well to the hydride side. Hydrogen atom exchange and isomerization, which have been observed between the hydride complex and olefins,-'-'-',77,78 occur via an equilibrium reaction such as Equation (5.3), and add further support to a mechanism involving an alkyl intermediate. Much stronger evidence for interaction (i.e. coordination, not necessarily insertion) between olefin and hydride complex was obtained in DMA during the present studies, where an obvious color change resulted from the addition of some olefins (e.g. maleic acid) to the violet catalyst solution. The reversible olefin coordination and insertion steps (5.2) and (5.3) are verified by the observation that highly deuterated ethanes (> 2 deuteriums) are produced during the deuterium reduction of ethylene. These equilibria must be more rapid than the quite rapid reduction of ethylene, which is governed by steps (5.4) and (5.5). The recently discovered reaction of the complexes Ru(N2)K2(PPh3)3 and RuH2(PPh3)3 with H2 to yield RuH^  ( P P h ^ 5 8 ' 5 9 involves oxidative addition of hydrogen to produce a Ru1^ complex from Ru 1 1, and provides some evidence for the similar step in Equation (5.4). - 129 -5.5.1 The reaction kinetics in DMA As discussed in Chapter IV, molecular weight measurements were the only evidence obtained in benzene solution to suggest the existence of equilibrium (5.1), although results in DMA (Section 4.5.4) did suggest -3 partial catalyst dissociation at 10 M Ru. The first-order dependence on catalyst concentration for the maleic acid hydrogenation rate at a high substratercatalyst ratio (Figure 36) suggests that, at least in the presence of a strongly-coordinating olefin such as maleic acid (see later discussion), no reassociation of phosphine to produce RuClH(PPh.j).j occurs. (Previously obtained data 1 1 show that less than a first-order dependence in Ru(II) can result at high catalyst concentration i f a high substrate:catalyst ratio is not maintained.) The i n i t i a l non-linearity in Figure 36 could be due to further dissociation in very dilute catalyst solutions to give a catalytically less active species such as RuClH(PPh.j) or RuE^PPh,^ , although some loss of activity due to traces of oxygen impurity might also result in a similar plot for the catalyst dependence. Some results presented in Section 6.2.1 indicate that no such traces of oxygen would exist in the maleic acid system (see Figure 55). Confirming evidence that there is no significant phosphine reassociation at 0.1 M maleic acid is found in the observation that addition of triphenylphosphine in a 2:1 mole ratio with the catalyst has a negligible effect on the hydrogenation rate at this substrate concentration. The effect of added phosphine on the hydrogenation rates of other olefins was sometimes substantial, however, as summarized in Table XXIX. The "rate reduction factor" is calculated as the ratio of the rate - 130 -Table XXIX. RuClH(PPh3)3 catalyzed hydrogenation of olefins. Effect of added triphenylphosphine on i n i t i a l rates at 35°. [Ru11] = 2.0 x 10"3 M, [PPh3] = 4.0 x 10"3 M. Olefin [Olefin] 3 M Rate reduction factor^3 maleic acid 0.10 1.0 acrylamide 0.20 1.0 methyl vinyl ketone 0.25 1.0 fumaric acid 0.30 1.2 1-decene 1.16 2.5 1-hexene 1.24 2.7 1-nonene 1.09 2.7 4—methoxystyrene 1.10 4.5 cyclooctene 1.08 6.6 Concentration at which the maximum hydrogenation rate was measured. Calculated as defined in Section 5.5. - 131 -observed without added phosphine to the rate observed with PPh^ added. The rate reduction factor, which represents a competition of the back reaction of equilibrium (5.1) and the forward reaction of (5.2), is likely to be inversely related to the olefin coordination strength. Electron-withdrawing groups are known to increase olefin complexity constants, 7 9'^* in agreement with the observation that maleic acid, acrylamide and methyl vinyl ketone have negligible rate reduction factors and also give yellow solutions on reaction with the hydride. Steric effects are also known to be important in olefin coordination, even in 79 82 83 the absence of bulky phosphine ligands; ' ' for example, the equilibrium constants (at 40°) for the formation of a silver nitrate-olefin complex in ethylene glycol are 2.2 M * for cis-2-octene and 0.4 M * for trans-83 2-octene. Bulky substituents on an olefin decrease complexity constants, since only certain directions of approach of the olefin to the. metal will permit complexing to take place.^>84 ^ t ^ e s ubstrates in Table XXIX, the trans-arrangement of fumaric acid, the bulky phenyl group of 4-methoxy-styrene and the cyclic nature of the olefin cyclooctene would appear to be steric factors which serve to increase the rate reduction factor. Strongly coordinating olefins are able to compete successfully with the added phosphine for a coordination site, while poorly coordinating olefins such as the alk-l-enes are less successful. However, in the presence of excess substrate and the absence of added PPh^, the back reaction in Equation (5.1) becomes less important, as evidenced by an approach to a limiting rate with increasing substrate concentration. Since the oxidative addition of is expected to be rate-determining,*^ the proposed reaction mechanism leads to the rate expression - 132 -d[H ] R = - = klH 2]{alkyl complex] (5.6) In terms of the total catalyst concentration [Ru**]^, this becomes kK*[H 2][Ru I I] T[olefin] R = 1 + K»[olefin] ( 5 ' 7 ) where the combined equilibrium constant K' = K^^. In agreement with this rate expression, the hydrogenation reaction is first-order in H 2 and catalyst concentrations. At high olefin concentration, Equation (5.7) reduces to R = k[H 2]lRu I T] T (5.7a) Equation (5.7) can be rewritten in the form I = 1 + 1 ( 5 < 8 ) R kK'[H 2][Ru I i :] T[olefin] k[H2] [Ru11]^, According to Equation (5.8), a plot of 1/Rate against 1/[olefin] should be a straight line, and linear plots are in fact obtained for maleic and fumaric acids (Figure 43), 1-octene (Fig. 44), 1-nonene (Fig. 45), 1-hexene and 1-decene (Fig. 46), cyclooctene and 4-methoxy-styrene (Fig. 47), cyclohexene (Fig. 48) and acrylamide (Fig. 49). Inverse plots were not constructed for methyl vinyl ketone, since hydrogenation rates were not measured at sufficiently low concentration of this olefin, or for trans-2-octene, which showed first-order dependence on olefin up to the maximum concentration used. Values of - 133 -i c rH Ctf Pi rt •rt 4J •H c r H 0.0 0.5 _ x _ 2 [olefin] x 10 , M -1 Figure 43. Dependence of i n i t i a l hydrogenation rates of M.A. and F.A. on [olefin] at 35°, as plotted in accordance with Equation (5.8) (Tables XIX and XX, Figure 37). CO i c rH X pi rH (fl •H 4J •H c 0.0 Figure 44. 1 , 0 -1 -1 -1 [1-octene] x 10 ,M 0 Dependence of i n i t i a l hydrogenation rate of 1-octene on [olefin] at 35°, as plotted In accordance with Equation (5.8) (Table XXII, Figure 38). - 134 -3.0 -ro I o cu •u •H C3 2.0 1.0 -0.0 4-° -1 -1 [1-nonene] ,M Figure 45. Dependence of i n i t i a l hydrogenation rate of 1-nonene on [olefin] at 35°, as plotted in accordance with Equation (5.8) (Table XXII, Figure 39). I c cu 4J Pi cd •H 4-1 •H c 1.5 1.0 0.5 _ 1-hexene - _____—-i i 1-decene i 0.0 1.0 1 12.0 3.0 [olefin] ,M~ Figure 46. Dependence of i n i t i a l hydrogenation rates of 1-hexene and 1-deeene on [olefin] at 35°, as plotted in accordance with Eauation (5.8) (Table XXII, Figures 38 and 3 9 ) . - 135 -0.0 2.0 n 4.0 [olefin] ,M Figure 47. Dependence of i n i t i a l hydrogenation rates of cyclooctene and 4-methoxystyrene on [olefin] at 35°, as plotted in accordance with Equation (5.8) (Tables XXIV and XXVI, Figure 42). 0.0 1.0 _ x _ ± 2.0 [cyclohexene] ,M Figure 48. Dependence of i n i t i a l hydrogenation rate of cyclohexene on [olefin] at 35°, as plotted in accordance with Equation (5.8) (Table XXIV, Figure 41). - 136 -0.0 2.0 _ ± _ ± 4.0 [acrylamide] x 10 ,M Figure 49. Dependence of i n i t i a l hydrogenation rate of acrvlamide on [olefin] at 35°, as plotted in accordance with Equation (5.8) (Table XXVI). - 137 -k and K' were calculated from the gradient and intercept of the inverse II plots, using the relationships k = 1/intercept[H^][Ru ]^ and K' = intercept/gradient. The results of these calculations are summarized in Table XXX. In addition, K' << 1.2 for trans-2-octene and K' » 4 for methyl vinyl ketone, since the olefin dependence of hydrogenation rate was first-order for the former substrate at 0.8 M concentration and zero-order for the ketone at 0.25 M olefin. Three apparent olefin complexity constants (K1) were measured at 25° by recording the decrease in intensity of the visible absorption peak of the violet solution at 520 nm upon addition of olefin. Assuming negligible absorption by the olefin or alkyl complex at this wavelength, the equili-brium constants 1.0 M 1 and 23 M 1 were calculated for 1-decene and fumaric acid, respectively, in reasonable agreement with the K' values of Table XXX. The measured equilibrium constant for ethylene was 11.6 M 1 (see Section 6.3). Ethylene has also been observed to complex more 83 readily than alk-l-enes with AgNO^  in ethylene glycol solution: the complexity constants 22.3 M 1 and 2.6 M 1 were measured at 40° for ethylene and 1-octene, respectively. The observation that a DMA solution of 0.02 M ethylene is reduced at a rate equivalent to that of a solution containing 0.2-0.3 M alk-l-ene is probably due to the difference of a factor of ^ 10 in complexity constants, and the limiting rate of ethylene hydrogenation is likely very similar to that observed for terminal olefins. A large range of k and K' values is apparent for the olefins in Table XXX. K' is largest (> 20 M_1) for the olefins having electron-withdrawing groups: maleic acid, fumaric acid, methyl vinyl ketone and acrylamlde. K' values for the terminal alkenes are a l l of comparable magnitude (1.1-2.6 M "*"); 4-methoxystyrene and cyclooctene have - 138 -Table XXX. Catalytic hydrogenation of olefins using RuClH(PPh3) Values of K' and k determined at 35°. [Ru11] = 2.0 x 10~3 M. Olefin [H 2]* K' k x 103, M M"1 -2 -1 -1 x 10 , M s 1-hexene 1.75 1.5 5.2 1-octene 1.79 2.4 5.8 1-nonene 1.79 1.1 8.7 1-decene 1.79 2.6 6.8 4-methoxystyrene 1.79 0.7 4.1 cyclooctene 1.79 0.6 2.8 acrylamide 1.79 120 0.35 methyl vinyl ketone 1.78 » 4 a 0.28b cyclohexene 1.77 0.2 0.18 trans-2-octene 1.79 «1.2 a maleic acid 1.79 260 0.04 fumaric acid 1.79 25 0.04 * Average H2 concentration recorded in olefin variation experiments. Calculated using Equation (5.7). Calculated using Equation (5.7a). - 139 somewhat lower K' values (^  0.6 M *), and cyclohexene and trans-2-octene have much lower values (<? 0*2 M *). Equilibrium constants in the range 0.3-2.3 M * have been similarly calculated from kinetic 85 data at 25° for the formation of olefin complexes with RhCl(PPh_)_ j J , RhH(CO)(PPh3)3 7 0 and RuH(0C0CF3)(PPh3>3,40 using terminal, internal or cyclic alkenes without functional groups. Since K' = K^^, the magnitude of K' can give an indication of the effect of steric and electronic factors on olefin coordination (K^), and on the insertion of the olefin into the metal-hydride bond (K 3). It is evident from the relative K' values that a trans-isomer is much less suited for these reactions than a cis-isomer (cf. M.A. and F.A.) and that the presence of a bulky phenyl group or a cyclic arrangement is also unfavorable. Electron-withdrawing groups increase the value 79—81 86 of K', in agreement with the findings ' that olefin coordination strength is greatly increased by this type of functional group. Whether this inductive effect also influences the insertion step ( 3^) to any significant extent is not obvious. A comparison of the K' values in Table XXX with the rate reduction factors in Table XXIX reveals that these two sets of data are exactly parallel, an increase in rate reduction factor corresponding to a decrease in K'. Electron-withdrawing groups appear to significantly decrease the value of k, as evidenced by the results obtained for acrylamide and methyl vinyl ketone (Table XXX), and this seems consistent for an oxidative addition reaction. Steric affects may also be important in reducing the value of k, as shown by the result for cyclo-hexene. The very low k values calculated for maleic and fumaric acids might reflect unfavorable steric as well as electronic effects, perhaps - 140 -because these two substrates are in fact internal olefins. Alfrey-Price 87 "e" values, which are considered to reflect the electron density of olefinic linkages, are 1.30, 1.25 and 1.49 for acrylamide, diethylfumarate and diethylmaleate, respectively. In other words, the inductive effect of two -GT^Et groups is approximately equivalent to that of one amide group. Since diethylfumarate and diethylmaleate are hydrogenated at the same rate as their acid analogues, and acrylamide has a k value which is an order of magnitude higher, steric effects due to the internal olefinic structure of maleic acid and fumaric acid might apparently reduce the value of k. However, the apparent k values calculated for these latter two substrates may not refer to the oxidative addition reaction (5.4) (see following). Arrhenius plots were constructed using the temperature dependence data of Tables XXI, XXIII, XXV, and XXVII and the plots were found to be linear in a l l cases (Figures 50, 51 and 52). Values of t t enthalpy and entropy of activation (AH and AS ) calculated from these plots are summarized in Table XXXI. It should be noted that the olefin concentrations used for cyclooctene and 4-methoxystyrene were not sufficient to attain a maximum rate (Figure 42) so that the activation parameters calculated for these two olefins are approximate. Except for maleic acid, the activation parameters are compatible with the prediction that oxidative addition of H^  is the rate-determining step. Activation parameters similar to these have been determined for other oxidative addition reactions (see Section 4.5). The anomalous activation parameters calculated for the maleic acid system may result from its unique position as a strongly-coordinating internal olefin. The large K' value calculated for this - 141 -- 142 -3.0-2.5 log k 2.0 1.5 cyclooctene methyl vinyl ketone 3.10 T 1 x 10 3.20 3 3.30 o„- l 'K Figure 51. Arrhenius plots for the RuClHCPPh^^-catalyzed hydrogenation of 1-nonene, 1-hexene, cyclooctene and methyl vinyl ketone in DMA (Tables XXIII, XXV and XXVI). - 143 -3.0 2.5 log k 2.0 1.5 1-decene 4-methcxy-styrene acrylamide _L 3.10 3.20 -1 3 T x 10 , 3.30 'K Figure 52. Arrhenius plots for the RuClH(PPh3)^catalyzed hydrogenation of 1-decene, 4-methoxystyrene and acrylamide in DMA. (Tables XXIII and XXVI). - 144 -Table XXXI. Catalytic hydrogenation of olefins using RuClH(PPh3) Activation parameters. Olefin AH+ AS+ kcal/mole eu 1-hexene 9.8±1.0 -14.9+3.5 1-nonene 9.5±0.4 -15.3+1.4 1-decene 8.3±0.6 -18.9±2.1 4-me thoxys tyrene 11.3+0.2 -11.6+0.7 cyclooctene 8.0±1.0 -23.2+3.5 acrylamide 11.6+0.2 -13.8+0.7 methyl vinyl ketone 13.2±0.7 -9.3±2.4 maleic acid 22.0±2.0 +15.5+6.9 - 145 -olefin could reflect a large value of K^ , since for this substrate may in fact be very small because of steric hindrance to the insertion of internal olefins. In other words, the hydrido-olefin complex RuClHCPPh^^(M.A.) may be completely formed at the limiting rate, so that the mechanism can be represented under these conditions by the two steps K 3 RuClH(PPh„)_(M.A.) — ^ RuCl(PPh.).(alkyl) (5.9) J z j z RuCl(PPh3)2(alkyl) + H2 -^-»- RuClH(PPh3)2 + succinic acid (5.10) Assuming the i n i t i a l (K3) equilibrium to be fast, this mechanism gives rise to the rate law R= kK3[H2][RuClH(PPh3)2(M.A.)] • = kK 3[H 2][Ru T I] (5.11) since [RuClH(PPh3)2(M.A.) ] = [Ru 1 1^ when K3 is small. Thus, using this analysis, the K' value quoted for maleic (or fumaric) acid in Table XXX may actually be the value of K2, and the quoted k value would then be kK3« Although the value of K' is much smaller for fumaric acid than for maleic acid, presumably reflecting mainly differences in K2» the values of k (which may be kK-j) are very similar for these two substrates. Such a rationalization seems reasonable, since the same alkyl intermediate results, and shows that in the case of a trans-isomer - 146 -additional steric control is exercised in the coordination step, but not in the insertion step. The stable hydrido-olefin complex RuClH(C^HQ)(PPh_)_,10 formed / o 5 2. by reaction of norbornadiene with RuClHCPPhg)^, is analogous to the reacting species proposed to exist in solution here. Other metal complexes containing both hydride and olefin ligands are known, 88 89—91 including Ir(III) and Pt(II) complexes. If reaction (5.4) were not greatly influenced by steric factors, the value of k would be expected to be essentially the same for the cyclohexene, cyclooctene and terminal olefin systems; the much lower apparent k value determined for cyclohexene (Table XXX) may result from a mechanistic variation such as that proposed for the maleic acid reduction, but with much smaller for the cyclic olefin. The limiting hydrogenation rate may therefore occur in this case when a l l of the catalyst is present as the hydrido-olefin complex; even though and k for cyclohexene may be similar to the values for the cyclooctene and terminal olefin systems, the much lower limiting hydrogenation rate of cyclohexene could result from a smaller value. Apparently a conformation of the more flexible cyclooctene ring is available which experiences only limited steric interactions during insertion, since this olefin is reduced at a rate comparable with that of the terminal alkenes. Anomalous behaviour of cyclohexene, compared to larger cyclic alkenes, has also been observed in RhCl(PPh^^-catalyzed deuteration 92 experiments. Although clean addition of 2 D atoms per molecule of cyclohexene occurred in this case, the larger rings underwent H-D "scrambling" during the reaction; the anomaly was proposed to arise in the insertion (i.e., K^ ) step. - 147 -As with cyclohexene, the low hydrogenation rate of the internal olefins trans-2-octene and cis-2-butene (Section 5.2.5) may result primarily from a low value. In addition, the value for trans-2-octene is expected to be lower than that of the cis-isomer (cf. maleic and fumaric acids). The claim that cyclic olefins are reduced more slowly than terminal 3 10 olefins by a factor of at least 10 in benzene solution should be treated with caution, since cyclohexene was the only cyclic olefin used, and cyclooctene is in fact reduced much more rapidly than the six-membered ring in DMA solution. Nevertheless, the selectivity of reduction between terminal alkenes and cyclohexene in benzene solution 3 (a factor of ^ 10 ) is much higher than in DMA solution (*\» 40). Decreased selectivity in the more polar and more strongly coordinating amide solvent could result from a slightly different steric arrangement of the bulky phosphine groups in the catalytic species. The temperature dependence data obtained for maleic acid and fumaric acid (Table XXI) can be used to obtain an estimate of the thermodynamic parameters AH° and AS0 for the possible equilibrium reaction involving coordination of fumaric acid: K2 H-Ru + F.A. —=•»- H-Ru(F.A.) (5.12) Because is much greater for maleic than for fumaric acid (see Figure II 37),at 0.1 M olefin concentration a l l of the Ru is present as the olefin complex in the case of maleic acid, whereas H-Ru(F.A.) is only partially formed. Since the limiting rate is essentially the same for - 148 -both olefins, [H-Ru(F.A.)] = [Ru11]™ (5.13) where R^, ^  and R^. refer to the respective hydrogenation rates of the two olefins at 0.1 M substrate concentration. The value for fumaric acid is given by [H-Ru(F.A,)] 2 [H-Ru][F.A.] U * ± 4 ; Substitution using Equation (5.13) in Equation (5.14) yields the expression K . ^ . A ^ V ^ A . RF.A. (5.15) where [F.A.] = [F.A.]T since [F.A.]T » [Ru11] . The values calculated for at four temperatures are summarized in Table XXXII. A plot of log against T * yields a good straight line, from which are derived the parameters AH° = -12.5 + 0.8 kcal/mole,ASC -33.7+ 2.8 eu. Similar thermodynamic parameters have been calculated for the formation of ir-olefin complexes of strongly-coordinating olefins 81 with Ni(dipy); for example, AH" = -12.8 kcal/mole, AS0 = -22.0 eu for the formation of the maleic anhydride complex in tetrahydrofuran solution. A crude estimate of K' values at four temperatures can be obtained for the 1-octene system. Comparison of the hydrogenation rate of 0.23 M 1-octene solutions (Table XXIII) with the maximum rate obtained at high - 149 -Table XXXII. Hydrogenation of fumaric acid and 1-octene. Temperature dependence of for fumaric acid and K' for 1-octene. T K9 (F.A.)3 K* (l-octene) b 2 -1 -1 °C M M 30 40.6 2.42 35 30.6 2.26 40 22.0 2.02 45 15.2 1.83 Calculated according to Equation (5.15). Calculated according to Equation (5.16). - 150 -olefin concentration allows calculation of K' values (cf. Equation 5.15); Ki = [Ru-alkyl] _ R . [H-Ru] [1-octene] (R -R)(0.23) ^-^) max where R is the hydrogenation rate of the 0.23 M 1-octene solution. Since R values were not obtained at several temperatures for this olefin, max the comparable values measured for 1-nonene (Table XXIII) were used in this crude calculation. The values of K' calculated using this method are summarized in Table XXXII. A plot of log K' against T 1 gave a good straight line, from which the parameters AH° = -3.8 + 0.2 kcal/mole, AS° = -10.9 +0.9 eu were derived. The only substrates which could not be hydrogenated at a measurable rate using the RuClH(PPh3)3 catalyst were methy1-fumaric acid, trans-stilbene and trans-cinnamic acid; these olefins would a l l be expected to have very low complexity constants 7 9' 8 4 as well as low values because of steric factors. Crotonic acid (CH3CH=CH-C00H) at 0.5 M concentration is reduced much more slowly than fumaric acid, and this could be attributed to a lower complexity constant due to the presence of an electron-releasing methyl group instead of the second carboxyl group of fumaric acid. Olefin mixtures -are reduced at the same rate as a solution containing only the more strongly-coordinating olefin (e.g. maleic acid vs. trans-stilbene, and maleic acid vs. 1-octene), verifying the importance of the i n i t i a l olefin coordination step in the reaction mechanism. Catalyst poisoning, which occurred during most olefin hydrogenation reactions, was more severe for strongly coordinating olefins such as - 151 -maleic acid, diethylmaleate and acrylamide than for fumaric acid, which coordinates less strongly. Maleic acid monoamide, which might be expected to form the most stable metal-olefin complex of any of the 81 substrates used, was not hydrogenated except in acidic solution. Presumably protonation of the amide group modifies the reactivity of the otherwise inert olefin complex. The substrate methyl vinyl ketone deactivated the catalyst rapidly, possibly via decarbonylation of the ketone; the comparable decarbonylation reaction with aldehydes 93 is well known; catalyst deactivation by a l l y l alcohol may result 94 from the formation of an inactive all y l i c or carbonyl complex, although these points were not pursued experimentally. Poisoning of the catalyst was also observed under hydrogenation conditions in benzene solution. 1^ Catalyst recovered after the hydrogenation of 1-octene in benzene was studied thoroughly, and the inactive species was suggested to be an alkyl, alkyl-alkene or ally l i c complex.1^ 5.5.2 Other ruthenium(II) complexes The complex RuH(OAc)(PPh^)^ has been found previously to be only about 1% as active as RuClH(PPh3)3 for the catalytic hydrogenation of 40 1-hexene in benzene solution. The present data in DMA for the substrates 1-nonene and maleic acid show that the activity of the acetate is ^ 35% that of the chloride in this solvent; RuBrH(PPh3)3 shows catalytic activity comparable to that of the chloride (Table XXVIII). The inactivity of the mono- and dicarbonyl complexes in Table XXVIII follows the previously reported trend of decreasing hydrogenation activity - 152 -with carbonylation of ruthenium(II) solutions.*''" It is of interest to note that some recent s t u d i e s ^ h a v e shown that the rate of olefin isomerization catalyzed by RuC^^Ph^)^ is greatly accelerated after reaction of the catalyst with peroxide impurities in the olefin to give a carbonyl complex, RuC^ (CO) (PPh^^ (olefin). - 153 -CHAPTER VT HOMOGENEOUS POLYMERIZATIONS OF ETHYLENE AND BUTADIENE, CATALYZED BY HYDRIDOCHLOROTRIS(TRIPHENYLPHOSPHINE)RUTHENIUM(II) 6.1 Introduction The violet hydride complex RuClH(PPh3)3'DMA was isolated from concentrated DMA solutions of RuC^(PPh^)^ after reaction with hydrogen, as described in Section 4.2.1. Continuous uptake of both ethylene, and 1,3-butadiene was observed by DMA solutions of this complex in the absence of hydrogen, suggesting a polymerization reaction. Salts of the group VIII metals are known to catalyze the polymerization and dimerization of olefins in polar media,9^ 9^ and the mechanism of chain propagation is thought to be analogous to that of the Zlegler catalysts (see Section 1.3). Little kinetic data have been reported for this type of homogeneous reaction, except for Cramer's detailed kinetic and mechanistic study on the dimerization 99 of ethylene catalyzed by rhodium trichloride; this system involves initiation via a metal hydride. By careful exclusion of oxygen, reproducible kinetic data could be recorded, when using the same catalyst and solvent batch, on the present ethylene and butadiene uptake reactions. The results obtained contribute to the understanding of the mechanism of polymerization reactions involving coordination catalysis. - 154 -6.2 Polymerization Reactions in DMA The uptake of olefin gas by DMA solutions of RuClHCPPh^-DMA was recorded from the point at which the catalyst was added to the solvent, which had previously, been equilibrated with the gas at the required temperature and pressure; the catalyst dissolved immediately at the temperatures used (50-85°C). Ethylene and butadiene were -3 -2 absorbed at conveniently measurable rates using ca. 10 - 10 M catalyst and up to one atmosphere olefin pressure. No measurable uptake of vinyl fluoride occurred over a period of 3 hr at 80°, at _3 4.5 x 10 M catalyst and one atmosphere pressure. The solubilities of ethylene, butadiene and vinyl fluoride in DMA at 80° were found -2 -1 -1 -1 experimentally to be 4.2 x 10 M atm , 9.2 x 10 M atm and 1.7 x 10 1 M atm 1 respectively, and Henry's Law was obeyed at least up to 1 atm pressure. The dependence of polymerization rate on temperature was investigated using a different batch of catalyst and solvent, accounting for the slight discrepancy in the rate data (see Tables XXXIII-XXXVI). 6.2.1 Ethylene polymerization On reaction with ethylene, the color of the catalyst solution changed rather quickly from violet to amber (a continuum down to 350 nm, e = 660 at 400 nm). An "autocatalytic" type gas uptake region was observed during the color change (y 1000 sec) followed by a period of linear uptake for about one hour. Representative uptake plots recorded at three different catalyst concentrations are shown in Figure 53. The kinetics of the reaction were obtained from the linear 6.0 5.0 Ul 2.0 time x 10 ,s Figure 53. Rate plots for the RuClH(PPh3 ^ -catalyzed polymerization of ethylene in DMA at 80° (4.0 x 10 _ 2 C 2H 4). t R u I T ] : ( 0 ) 2.3 x 10'3 M; (•) 4.5 x 10"3 M; ( O ) 9.8 x 10~3 M. M - 156 -region of the uptake plots (Table XXXIII). The data showed a f i r s t -order dependence on ethylene (Figure 54); the reaction was also first-order in catalyst concentration up to the solubility limit of the -.3 catalyst (estimated to be 7 x 10 M). At higher catalyst concentrations, undissolved particles of catalyst were observed in the solution after the completion of uptake measurements. If CP. grade ethylene were used without purification, the plot of rate vs. catalyst concentration gave a positive intercept on the abscissa axis (Figure 55). This effect was presumably due to destruction of a portion of the catalyst by oxygen present as impurity in the ethylene; purification of the ethylene resulted in increased rates, with the linear portion of the ruthenium dependence plot now passing through the origin (Figure 55, broken line). When using unpurified ethylene, solutions with low catalyst concentration had a greenish color which is evidence for partial oxidation of the catalyst; no green color was observed when using purified ethylene. The rate of reaction can be expressed as: Rate = k 1[C 2H 4][Ru I I] T (6.1) where [Ru**]^ is the total ruthenium concentration. Neglecting loss of some catalyst by reaction with oxygen impurity, Figures 54 and 55 - 2 - 2 give the bimolecular rate constant at 80° as 6.5 x 10 and 8 x 10 M *s * respectively; allowing for catalyst loss, the first value will be somewhat higher. Table XXXIV lists k values determined at 70-85°, and the Arrhenius plot shown in Figure 56 gives the activation - 157 -Table XXXIII. RuClH(PPh3)3- catalyzed polymerization of ethylene in DMA solution. Dependence of linear rate on concentratii II of Ru and C „H. at 80°. 2 A [Ru11] C 2 H 4 [C2H4] Linear rate x 10 , M mm x 10, M x 10 , Ms 4.69 725 0.A0 2.8A 1.49 725 0.A0 2.33 0.98 725 0.A0 2.08 0.45 725 0.A0 1.31 0.34 725 0.A0 0.98 0.29 725 0.A0 0.76 0.23 725 0.AO 0.A7 0.17 725 0.A0 0.24 0.18 725 a 0.A0 0.63 0.23 725 a 0.A0 0.89 0.34 725 a 0.A0 1.12 0.34 b 725 a 0.A0 0.55 0.34 C 725 3 0.A0 0.34 0.44 725 0.A0 1.09 0.44 441 0.2A 0.71 0.44 242 0.13 0.42 0.44 375 0.21 0.28 d 0.44 735 0.A1 0.50 d 3.31 6 725 0.A0 0.0 0.45 7 2 5 a ' f 0.A0 0.0 a Substrate purified; b 0.61 x 10 2 M PPh added; ° 3.4 x 10~ 2 M LiCl added; At 70°; 6 RuCl2(PPh3>3 used instead of RuClHCPPh^'DMA; C„H0F used instead of C„H - 158 -Figure 54. Dependence of linear rate of ethylene polymerization in DMA at 80° on I C ^ ] (4.4 x 10 _ 3 M Ru 1 1). (Table XXXIII). 3.0 [Ru11] x 102, M Figure 55. Dependence of linear rate of ethylene polymerization in DMA at 80° on [Ru11] (4.0 x 10"2 M C H : CP. grade; purified) (Table XXXIII). - 160 -Table XXXIV. RuClHCPPh^)^ —catalyzed polymerization of ethylene in DMA solution. Temperature dependence of k^. [Ru11] = 4.4 x 10"3 M, IC 2H 4] = 4.0 x 10"2 M. Temperature °C * k l x 102, M~1s"1 70 2.5 75 4.0 80 6.1 85 8.7 Calculated using Equation (6.1). - 161 -2.80 2.85 2.90 -1 3 -1 T x 10 , °K ure 56. Arrhenius plot for the RuClH(PPh3)3-catalyzed polymerization of ethylene in DMA (Table XXXIV). - 162 -parameters AH = 20.0+0.7 kcal/mole and AS = -7.7 + 2.0 eu. A gradual decrease in the uptake rate, accompanied by a solution color change to yellow (a continuum down to 350 nm, with e = 735 at 400 nm), was observed over several hours (Figure 57). The plots levelled off after an ethylene consumption of about 20 moles per mole of catalyst. It was thought that this effect might be due to a build-up of gaseous product, with consequent reduction in the partial pressure of ethylene. However, evacuating the system after the reaction had proceeded for several hours and repressurizing with fresh ethylene did not increase the uptake rate. Addition of about 2 moles of triphenylphosphine per mole of catalyst reduced the linear rate by a factor of about 2; addition of a 10-fold excess of lithium chloride Inhibited activity by a factor of 3. No gaseous products were detected at the end of a two day reaction _2 time at 80° (4 x 10 M catalyst, 1 atm. pressure); solvent fractions collected also gave no trace of product by glc analysis. The remaining sticky orange residue showed an infrared peak at 1950-55 cm thought to be due to a carbonyl formed by decarbonylation of the DMA. Infrared bands at 2960, 2925, 2865, 1460 and 1380 cm"1 were attributed to an aliphatic hydrocarbon. By heating the residue to 200° in vacuo, small amounts of a clear condensible liquid were obtained; the i r spectrum identified i t as the aliphatic hydrocarbon, and the strength of the methyl peaks (1380 cm 1) indicated considerable chain branching. Detection of any olefinic structure in the i r was hampered by the presence of sublimed triphenylphosphine in the condensed o i l . - 164 -Although no characterizable complexes were isolable from the final, inactive yellow solutions, i t was found that removing the ethylene atmosphere and adding hydrogen at room temperature rapidly regenerated the violet color of the starting catalyst. However, an - 3 inactive solution containing 3 x 10 M catalyst was reactivated (after 9 hr under C^R^ at 80°) by the hydrogen treatment to only about 40% of its maximum activity. Solutions of RuC^^Ph^-j were found to be inactive for the ethylene polymerization; RuClHtPPh^)^ formed in situ by reaction of RuC^CPPh-p^ with H2 was also inactive, since the dichloro complex is obtained on removal of hydrogen (Section 4 . 3 ) . 6.2.2 Butadiene polymerization The i n i t i a l part of a butadiene uptake plot at 80° using the hydride catalyst is shown in Figure 58. The violet solution immediately became yellow on addition of the diene, and no "autocatalytic" type uptake region was observed. The rate decreased with time, but the in i t i a l rate was some 200 times greater than the linear region of the ethylene uptakes for corresponding conditions. Initial rate data for the butadiene system are given in Table XXXV. After about one hour, when ^400 moles of diene per mole of catalyst had been consumed (for the conditions of Figure 58), the reaction rate had decreased from the in i t i a l value of 36.1 x 10~5 M s" 1 to a value of 2.20 x 10~5 M s" 1, which then remained constant for vL.5 hours before slowly decreasing. Initial rate data indicated a first-order dependence on ruthenium (Figure 59) but, in contrast to the ethylene system, the rate was now - 166 -Table XXXV. RuClH(PPh^)^-catalyzed polymerization of butadiene in DMA solution. Dependence of i n i t i a l rate on concentrations of Ru 1 1 and C.H- at 4 6 65°. [Ru11] * C4 H6 Initial rate x 103, M mm M 4 -1 x 10 , M s 0.29 743 0.90 0.51 0.70 743 0.90 1.49 1.05 743 0.90 2.46 0.50 743 0.90 0.91 0.50 509 0.62 0.90 0.50 317 0.38 0.87 0.50 166 0.20 0.86 0.50 725 0.88 3.613 0.50b 725 0.88 0.23a'° * a b Substrate purified; At 80°; RuCl 2(PPh 3) 3 used instead of RuClH(PPh ) 'DMA; c linear rate. - 167 -co o 2.0 o r H X Initial Rate 1.0 Initial Rate 0.0 i 1 0.0 II O'5 [Ru ] x 10 ,M 1.0 Figure 59. Dependence of i n i t i a l rate of butadiene polymerization in DMA at 65° on [Ru 1 1] (0.90 M C.H,) (Table XXXV). Figure 60. Arrhenius plot for the RuClH(PPh3)3~catalyzed polymerization of butadiene in DMA (Table XXXVI). - 168 -independent of the substrate concentration. The unimolecular rate constants (k^) estimated for the butadiene system are given in Table XXXVI. An Arrhenius plot (Figure 60) yielded the activation parameters, AH+ = 27.8 + 1.9 kcal/mole and AS+ = 21.0 + 6.0 eu. During the rapid butadiene uptake, the DMA solutions quickly became cloudy, and the o i l droplets which were formed coalesced to a gel on cooling. The gel, while soft, was partially soluble in dichloromethane or carbon disulfide; the gel became semisolid on standing in air for several days. Infrared analysis of the polybutadiene microstructure was carried out in carbon disulfide solutions using extinction coefficients calculated by Hampton1^ for the cis-vinylene, vinyl and trans-vinylene linkages. Absorbances for the three linkages were measured at 724, 911 and 967 cm 1 respectively; the results indicated that the polymer contained approximately equal percentages of.the three types of olefinic structure. Add i t ion of triphenylphosphine (PPh^^Ru = 30) decreased the butadiene polymerization rate by a factor of 5; no change in the microstructure of the isolated product was observed as a result of the presence of excess phosphine (PPh^rRu = 5). Following a polymerization reaction, i f the butadiene were pumped off and replaced by 1 atm hydrogen, rapid gas uptake (initial rate -5 -1 -4 12 x 10 Ms at 5.0 x 10 M Ru) occurred at 80°. At the stage when this reaction had slowed to ^ 20% of its i n i t i a l rate ( VL hr), an infrared spectrum of the gel product indicated that ^50% of a l l three types of olefinic linkage had been hydrogenated. - 169 -Table XXXVI. RuClHCPPh^-catalyzed polymerization of butadiene in DMA solution. Temperature dependence of k^. iRu 1 1] = 5.0 "x 10~4 M, IC4H 63 = 1.0 M. * Temperature k2 °C x 10, s _ 1 50 0.41 55 0.87 60 1.75 65 2.64 Calculated using the equation, Rate = k_[Ru ]. - 170 -The dichloro complex RuC^CPPhg)^ was also found to catalyze the formation of polybutadiene (Figure 58). Addition of butadiene to the brown solution of dichloro complex resulted in immediate production of a yellow coloration. The uptake rate increased for 1200 sec and then remained linear for several hours before slowly decreasing. This linear rate (2.32 x 10 ^  M s 1 for the conditions of Figure 58) corresponded closely to that of the linear region observed with the hydride-catalyzed reaction at the same conditions and concentrations. Infrared analysis of the gel product obtained from the reaction catalyzed by RuC^^Ph^) 3 indicated the same relative percentages of the three types of polybutadiene linkages as found in the product of the hydride-catalyzed reaction. 6.3 Reaction of RuClH(PPh3)3 with Ethylene and Butadiene at 25° Color changes were observed on subjecting DMA solutions of the hydride complex to the olefinic substrates at the polymerization temperatures. Some quantitative measurements were also made at 25° by following changes in optical density at 520 nm, the absorption maximum of the hydride. The observed rapid loss of the violet color was analyzed in terms of reaction (6.2), the resulting ethyl or butenyl complex having no significant absorption at this wavelength: HRuCl(PPh0)_ + C_H. (or C.H.) C0HC (or C.H_)RuCl(PPh„) . 5 1 2 4 4 6 -«— 2 5 4 7 5 2 (6.2) -4 Using 7.4 x 10 M hydride solutions and an ethylene pressure of - 171 --2 -1 1 atm ([C2H4] = 7.3 x 10 M atm at 25°), K for ethylene was estimated to be 11.6 M *; the ethyl complex is thus approximately half formed at 1 atm a n d 25°. Butadiene complexes much more strongly, since even at 30 mm pressure ([C^H^J = 1.6 M atm * at 25°) the hydride color is completely removed; thus, K for butadiene must be > 200 M_1. During the spectrophotometric investigation of reaction (6.2) for ethylene, a very much slower fading of the hydride color occurred following the rapid decrease due to the K equilibrium (Figure 61). This slow fading gave linear log (optical density) vs. time plots (Figure 62), showing a first-order dependence on the hydride concentration remaining after reaction (6.2) had taken place; the pseudo-first-order rate constant was determined as 2.8 x 10 s *. 6.4 Discussion The maximum polymerization rates were first-order in the catalyst RuClH(PPh,j)3; the ethylene reaction was first-order in ethylene, but the butadiene reaction was independent of the diene concentration. These results can be explained in terms of a common mechanism for both systems, i f the stronger complexing ability of the diene (possibly as a chelate) is considered. Ethylene ini t i a l l y complexes weakly with the hydride, giving an alkyl species (Equation 6.2); the rate-determining step in the subsequent polymerization involves reaction of the ethyl complex with further ethylene. The i n i t i a l non-linear uptake region (Figure 53) is attributed to a build-up of alkyl concentration as the reaction proceeds: - 173 -K °2H4 Ru-H + C0H. Ru-C_HC — f — p r o d u c t s (6.3) 2 4 2 5 k. The slow fading of the hydride color after the i n i t i a l K equilibrium at 25° must give a measure of the subsequent polymerization rate at this temperature. Assuming a first-order dependence on ethylene, _2 which is present in large excess (7,3 x 10 M) in the spectrophotometric -4 -1 -1 study, gives a second-order rate constant of 3.8 x 10 M s ; extrapolation of the Arrhenius plot in Figure 56 yields a k^ value of -4 -1 -1 2.8 x 10 M s at 25°. Considering the lengthy extrapolation, the agreement between these values is good. 99 Following the reaction schemes of Cramer, the slow reaction in the polymerization can be written as follows: K1 k' Ru-CH2-R + C2H4 - — w (C2H4)Ru-CH2-R *- Ru-(CH2)3~R (6.4) Coordination of a second ethylene, like the first addition to the hydride complex, is likely to be fast, and the subsequent insertion reaction is likely to be slow (k'). Continuation of the processes outlined in Equation (6.4) gives rise to a growing alkyl chain. The extended linear uptake region observed indicates that the reactivities of the various intermediate alkyls are very similar; II this linear rate can thus be written as k'K' [C-J^] [Ru (alkyl)], and the estimated second-order rate constant (k ) values are in fact k'K' values. The rate law arising from the mechanism outlined in Equation (6.4) is given in the general case by: - 174 -. dlC 2H 4] ..k ,K'[C 2H 4][Ru I I] T dt = 1 + K'IC2H4] ( 6 , 5 ) Since the rate law in this case assumes the limiting form II k'K'[C2H4][Ru ] T > the K' equilibrium must therefore l i e well to the left at the reaction conditions. In other words, K ?[C 2H 4J << 1 -1 -? which means that K' << 25 M at 80°, since [ C ^ ] = 4 x 10 M. This result seems reasonable considering that the equilibrium constant for the reaction of ethylene with the hydride (Equation 6.2) is 11.6 M 1 at 25° and that such equilibrium constants usually decrease with increasing temperature.1^1 Butadiene completely removes the hydride color at the start of the reaction, i.e. equilibrium (6.2) lies well to the right, and presumably a butenyl complex exists in solution; no i n i t i a l curved region is seen in the uptake plot (Figure 58). The kinetic data can again be interpreted in terms of the mechanism outlined in Equation (6.4) and the rate law shown in (6.5). The essential independence of the rate on [C^Hg] implies that K'[C4Hg] >> 1. One factor affecting the predominance of this term in the denominator of Equation (6,5) is the butadiene solubility (0.92M atm ^ ) , which is about 25 times greater than that of ethylene. Since the butadiene polymerization rate is s t i l l independent of [C4H^] at 166 mm pressure, K' must be >> 5 for this system. Considering the measured equilibrium constant (> 200 M 1 at 25°) for reaction with the hydride, this result seems reasonable. Thus, compared with the ethylene system, both factors in the K'[olefin] term are much larger, and the rate law takes the limiting form k'[Ru11]^, for the butadiene system. The k 2 values listed in Table XXXVI - 175 -for the diene system are thus values of k', and are a direct measure of the insertion rate (Equation 6.4). It should be noted that the above analysis leads to the conclusion that reaction of the hydride with butadiene (investigated spectrophotometrically) produces a butenyl-butadiene species (C^ H-^ RuCC^ Hg) , and that the measured equilibrium constant in fact refers to a KK' value for a reaction with two butadiene molecules (Equations 6.2 and 6.4). Comparison of data from Tables XXXIV and XXXV indicates that, at 80° and about 1 atm of substrate, the polymerization rate for butadiene (k') is some 12 times greater than that for ethylene (k'K'). The rate of the actual.insertion step (k') is thus probably at least 50 times higher for the butadiene system. The activation parameters determined for the ethylene system t t (AH = 20,0 kcal/mole, AS = -7.7 eu) must apply to the composite constant k'K', although the rate constant is expected to be much more temperature dependent than K'. The parameters may be compared to those reported for ethylene dimerization catalyzed by rhodium chlorides in 99 HCl-ethanol media (16.6 kcal/mole, -20.1 eu), which were said to refer to the following insertion reaction: (C 2H 5)Rh(C 2H 4)Cl 3" (C2H5CH2CH2)RhCl3~ (6.6) The activation parameters for the rhodium system appear in fact to apply to a composite constant kK , where K refers to the pre-equilibrium: - 176 -K (C2H5)RhCl3" + C2H4 — ^ (C 2H 5)Rh(C 2H 4)Cl 3~ (6.7)' was measured and found to decrease with increasing temperature; from the data presented, the thermodynamic parameters AH° = -4 kcal/mole, AS0 = -6 eu may be estimated (K£ was 10 M * at -10°C). Allowing for the pre-equilibrium, the activation parameters for k would be closer to 21 kcal/mole and -14 eu. Presumably corrections of a similar magnitude could be made to the activation parameters in the present ruthenium system. Activation parameters of 12.2 kcal/mole and -14.9 eu have been reported for the propagation step in an ethylene polymerization 102 catalyzed by a soluble titanium Ziegler catalyst in toluene. The parameters obtained for the butadiene system (AH^  = 27.8 kcal/mole, AS^ = 21 eu) refer directly to the insertion reaction (cf. Equation 6.4). Comparison of the data shows that the activation energy for the butadiene system is likely greater than that for the ethylene system, and this is feasible considering that butadiene coordinates (possibly as a chelate) more strongly than the ethylene. The factor responsible for the greater rate of the insertion reaction with the butadiene system is the much more favorable entropy of activation. Such an effect could be related to the breaking of a butadiene chelate ring in the transition state required for the overall insertion, or perhaps a reduction in the degree of solvation of the transition state complex in this polar medium. The high degree of chain branching in the ethylene polymer could arise through isomerizatipn via an alkyl complex: - 177 -Ru-CH2CH2R — H - R u -CH„ CH. 2 | 3 — R u - C H R (6.8) CHR Butene isomerization has been found to be -very fast under the conditions 99 of ethylene dimerization at a rhodium centre, i.e. isomerization is much faster than insertion. Product release could occur from a hydrido-olefin intermediate formed as in Equation (6.8), or via an intramolecular hydrogen transfer in an alkyl-ethylene intermediate: (C2H4)Ru-CH2CH2R > (C2H5)Ru + CH2=CH-R (6.9) Similar reaction schemes can be written for the butadiene system. A reaction such as Equation (6.3) can give two possible butenyl intermediates, Ru-CH(CH3)CH=CH2 or Ru-CH2CH2CH=CH2, depending on whether the hydrometalation is Markownikoff or anti-Markownikoff, respectively. Similarly, subsequent insertion reactions involving further butadiene can give product as follows: Ru-R + C.H, > Ru-CH--CH-CH=CH- (6.10) H O I | 2. R Ru-CH-CH=CH, I 2 CH2R Vinylene linkages could arise via ir-allyl and o-fr-intermediates, which have sometimes been postulated for reactions involving conjugated 103,104 . dienes, for example: - 178 -CH -R CH_-R CH -R / \ CH Ru CH ^ Ru*-» CH Ru (6.11) •// I II The inhibition of polymerization by added chloride or PPh^ shows the importance of available coordination sites on the metal complex. Similar inhibition by PPh^ was observed in the hydrogenation of olefins using RuClH(PPh.j)3 as catalyst (Chapter V). It is significant that polybutadiene with the same microstructure was obtained from reactions catalyzed by both the hydride complex and the dichloro complex. The initiation period required for the dichloro complex is likely due to formation of the hydride by a reaction with butadiene; that the two systems eventually give extended linear regions of essentially equal rates (2.20 x 10~5 M s _ 1 for RuClH(PPh3)3, and 2.32 x 10~5 M s - 1 for RuCl2(PPh3)3) indicates that the same catalytic species are present. The dichloro complex is inactive for ethylene polymerization, presumably because the hydride is not formed by 97 reaction with C„H.. Canale and coworkers have used 6:1 PPh_-RuCl0 2 4 3 3 systems at 25° in aqueous emulsion to polymerize butadiene; the microstructure of their product was somewhat similar to that obtained in the present case. These workers suggested possible initiation via a metal-hydride complex and steric control effected through a metal-diene complex. A similar polybutadiene product has also.been obtained recently"^"' using RuCl 2(PPh 3) 3 in dichloromethane solution at 60°; the molecular weight of this polymer corresponded to ^ 50 units of butadiene. - 179 -The propagation reaction was considered to proceed through successive insertion of butadiene monomers into Ru-allylic bonds, with a-allylic bonds (as in species I) being more active than T r-allylic bonds (as in species II) for this reaction. The growing end of the polybutadiene chain was also considered to have some dynamic all y l character (cf. Equation 6.11), thus accounting for the presence of vinylene as well as vinyl linkages in the microstructure. Some deactivation of the catalyst undoubtedly arises from decarbonylation of the solvent which gives less active ruthenium carbonyl species. However, observations that (a) deactivation occurs much more readily in the butadiene reaction than in the ethylene reaction under the same conditions, and (b) that the catalyst activity can be partially regenerated by treatment with hydrogen, rules out decarbonylation as the only cause of deactivation. The complex RuClH(PPh3)3 in solution is thought to undergo a reversible intramolecular H2 elimination, via a process involving the ortho hydrogen of a . . . 106 phenyl ring: ? , H -H2 f 1 HRuCl(PPh^)„ —*• (Ph„P)Ru-Cl —-=*• (Ph0P)Ru (6.12) V h 2 V \ } P h 2 P - V } III A similar reaction has been postulated for the complex (C2H^)RuCl(PPh3)2, with irreverisble elimination of ethane, to account for a stoichiometric 77 hydrogenation of ethylene in the absence of molecular hydrogen. Deactivation during the polymerization reactions (and the regeneration - 180 -the RuClHCPPhg)^ catalyst by R^ ) could result from the formation of a complex such as I I I . The partial hydrogenation of the butadiene polymer on subjecting the system to a hydrogen atmosphere is presumably catalyzed by RuClHCPPhg)^, a species which is highly efficient in hydrogenation reactions (Chapter V). Such reductions catalyzed by homogeneous 22 Ziegler catalysts are well known. Although vinyl fluoride was not polymerized by the violet RuClH(PPh3)3 solutions at 80°, a yellow coloration was produced rapidly, suggesting that the fluoroalkyl complex is thermodynamically too stable to react with further substrate. - 181 -CHAPTER VII GENERAL CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK The present studies involve investigation of the nature of the complex RuCl 2(PPh 3) 3 in solution, the reversible reaction of DMA solutions of this complex with molecular hydrogen to produce the hydride species RuClHCPPh^)^ and HCl, and a study of homogeneous hydrogenation and polymerization reactions catalyzed by the hydride complex. Spectrophotometric studies on benzene and DMA solutions of RuCl„(PPh„)o reveal that extensive phosphine dissociation occurs in both solvent systems; dissociation is virtually complete for -3 a 10 M DMA solution of the complex at 25°, and ^ 80% complete in benzene under the same conditions. The forward reaction involves an increase in both enthalpy and entropy, but is slightly less endothermic in DMA than in benzene. In the highly polar amide solvent, a further ionic dissociation reaction occurs, RuCl 2(PPh 3) 3 RuCl 2(PPh 3) 2 + PPh 3 (7.1) RuCl 2(PPh 3) 2 RuCl(PPh3)2 + + C l (7.2) - 182 -_3 corresponding to ^ 25% dissociation at 10 M Ru and 25°. The forward reaction is again endothermic, but involves a decrease in entropy due to high solvation of the ionic species. An obvious extension of this work would be an investigation of the dependence of the equilibrium constants and thermodynamic parameters of these equilibria on the anionic ligand X and the phosphine ligand P in RuX2P3. A quantitative knowledge of such equilibria is important for a detailed understanding of the catalytic activity of these ruthenium complexes, which are known to activate a number of small covalent gas molecules such as H2, 02, CO and C0H.. 2 4 DMA solutions of RuCl2(PPh.j)3 absorb H2 gas under ambient conditions to produce a hydridochloro complex in a reversible reaction, for example, RuCl 2(PPh 3) 3 + H2 RuClH(PPh3)3 + HCl (7.3) DMA is a basic amide solvent, and thus promotes hydride formation by effectively stabilizing the released HCl;at 1 atm hydrogen pressure, the equilibrium position is completely to the right in the absence of added HCl. In benzene or toluene solution, the hydrogen reaction will occur only in the presence of added base. The DMA solvate RuClH(PPh3)3. DMA can be isolated from concentrated solutions of the hydride complex prepared in situ in DMA. Spectrophotometric kinetic studies on the forward reaction reveal that RuCl 2(PPh 3) 3 itself is not reactive with hydrogen; RuCl 2(PPh 3) 2 is highly reactive, however, and RuCl(PPh 3) 2 + even more so. The forward - 183 -reaction is first-order in ruthenium, and also first-order in R at low concentration of R^ . Addition of either LiCl or PPh^ to the solutions decreases the reaction rate; data from chloride-dependence and phosphine-dependence experiments support the equilibrium constants measured spectrophotometrically for reactions (7.1) and (7.2). The reaction of HCl with the violet hydride is first-order in both HCl and hydride complex, and the rate is independent of H concentration. Activation parameters for both the forward and reverse reactions lend support to a mechanism involving oxidative addition of R^ (or HCl) to give an TV intermediate Ru species; this result suggests that heterolytic splitting of molecular hydrogen may in fact occur via an i n i t i a l oxidative addition reaction. The presence of excess PPh-j in the DMA solutions increases the apparent equilibrium constant for hydride formation; this observation correlates with a much stronger inverse phosphine dependence shown by the reverse reaction than the forward reaction. Good correlation is observed between the measured spectrophotometric equilibrium constant K and the ratio of the kinetically determined second-order rate constants for the forward and reverse reactions. Although the equilibrium reaction involving loss of phosphine by RuClH(PPh3)3 in solution could not be investigated spectrophotometrically because of the similarity of the spectra of the bis and trisphosphine species, the kinetically determined equilibrium constant for this reaction indicated a much smaller degree of dissociation than that observed for the dichloro complex. If, as expected, oxidative addition of R^ t n e rate-controlling step in the forward reaction, an increase in basicity of the ligands on the ruthenium should accelerate the reaction. Further work studying - 184 -such ligand effects would seem worthwhile; as yet, relatively l i t t l e kinetic and thermodynamic data have been reported for oxidative addition reactions, which are of the utmost Importance in homogeneous catalytic systems. A wide variety of olefins is hydrogenated under ambient conditions in DMA solution using molecular hydrogen in the presence of RuClHCPPhg)^. The catalytic reaction is first-order in hydrogen and in catalyst; the olefin dependence is first-order at low olefin concentration, but approaches zero-order as the substrate concentration increases. Olefins having electron-withdrawing functional groups coordinate strongly to the violet hydride complex, resulting ln a color change to yellow. Olefin complexity constants are much lower if the olefin lacks this inductive effect, or if unfavorable steric interactions occur between the olefin and the ruthenium complex. In reactions involving hydrogenation of olefin mixtures, the olefin i n i t i a l l y hydrogenated is the one which coordinates more strongly, and not necessarily the olefin which is reduced more rapidly. Addition of excess PPh^ to the solutions can reduce the rate of olefin hydrogenation; the comparative magnitude of this rate reduction for a series of olefins is inversely proportional to the olefin coordination strength. The proposed mechanism for the catalytic hydrogenation reaction involves dissociation of the hydride complex with loss of a PPh^ molecule (K^) , olefin coordination (K^ ,) , insertion of olefin into the Ru-hydride bond (K^), oxidative addition of to the resulting a-alkyl complex (k), and reductive elimination of alkane to regenerate the hydride complex: - 185 -RuClH(PPh_)0 — ^ RuClH(PPh„)„ + PPh TV k R u C l H 2 (alkyl) (PPh 3) 2 RuCl (alkyl) (PPh3) 2Scheme I The kinetically determined values of k and K' (where K' = K2K3) indicate the importance of steric and electronic factors in the olefin on the olefin coordination (K2) and insertion (K3) reactions, and on the oxidative addition of hydrogen (k). The presence of electron-withdrawing functional groups on the olefin serves to increase K2, but also decreases k. K2 and K3 can both be decreased in magnitude by unfavorable steric factors present in the olefin, via interaction with the bulky phosphine ligands. Activation parameters obtained for the hydrogenation reactions support the assignment of oxidative addition of H2 as the rate-determining step in the mechanism. Higher activation enthalpies are observed for olefins having electron-withdrawing functional groups, accounting for the smaller k values calculated for this type of substrate. A possible extension of this work is the study of steric and electronic effects in the phosphine or anionic ligands on the activity or selectivity of the catalyst for hydrogenation reactions. Some - 186 -preliminary work included in this thesis indicates that replacement of the chloride ligand in RuClHCPPhg)^ by bromide has only a minor effect on measured hydrogenation rates, whereas replacement by acetate appears to decrease rates somewhat. Optically active phosphine ligands could be used instead of PPh^, with the possibility of generating optically active products. Such asymmetric syntheses using corresponding rhodium complexes are currently being studied intensively. Ethylene and butadiene are both polymerized in DMA solution at temperatures of 50-85° using RuClH(PPh.j)3 as catalyst. The reaction rates are first-order in catalyst, and are first-order in substrate for the ethylene system, but are independent of the substrate concentration for the diene system. The polymerization reactions are inhibited by addition of PPh^ to the solutions , suggesting the importance to the reactions of free coordination sites on the catalytic species. The data are interpreted in terms of a common mechanism analogous to that of Ziegler catalysts, where, following coordination of the monomer to the metal, the polymer chain grows by insertion of the monomer at a polarized metal-carbon bond, for example, Ru-H + C-H. 2 k (7.4) (C2H4)Ru-C2H5 (7.5) (C2H4)Ru-C2H5 Ru-C2H4-C2H5 (7.6) - 187 -The present data indicate that the insertion reaction (7.6) is rate-determining subsequent to relatively rapid coordination of the olefinic substrates. The differences in kinetics of the two substrate systems are accounted for by the stronger complexing ability of the diene. The activation parameters obtained show that the butadiene polymerization rate is higher because of a more favorable entropy of activation. The microstructure of the polybutadiene product contains approximately equal amounts of vinyl, cis-vinylene and trans-vinylene linkages; these olefinic linkages can be catalytically reduced by addition of molecular hydrogen to the reaction solution at 80°. During polymerization, the catalyst becomes deactivated due to decarbonylation of the solvent, and possibly due to an intramolecular hydrogen transfer process involving a phenyl ring of the phosphine ligand. The complex RuC^^Ph^^ also catalyzes at 80° the polymerization of butadiene, but not the ethylene reaction. The diene polymerization rate in this case is much lower than the i n i t i a l rate obtained using the hydride catalyst; nevertheless, a polybutadiene product of the same microstructure is obtained. The dichloro complex is apparently converted slowly to the hydride by reaction with butadiene. Further kinetic studies on related systems investigating ligand, substrate and solvent effects, and particularly molecular weight distribution in the products could help elucidate further the factors controlling the rate-determining step. Such investigations should contribute significantly to the understanding of stereoselective polymerization. - 188 -REFERENCES 1. M. Calvin, Trans. Faraday Soc, 3_4, 1181 (1938). 2. See for example, G. Szonyi, Adv. in Chem. Series, Vol. 70, Amer. Chem. Soc., Washington, D.C., 1968, p. 53. 3. See for example, R.F. Gould, ed., "Bioinorganic Chemistry", Adv. in Chem. Series, Vol. 100, Am. Chem. Soc, Washington, D.C, 1971. 4. CG. Bond, Adv. in Chem. Series, Vol. 70, Am. Chem. Soc, Washington, D.C, 1968, p. 25* 5. J. Halpern, J. Harrod and B.R. James, J. Am. Chem.Soc, 83, 753 (1961). 6. J.F. Harrod, S. Ciccone and J. Halpern, Can. J. Chem., 39_, 1372 (1961). 7. B.R. James, Inorg. Chim. Acta Rev., 4^, 73 (1970), and references therein. 8. D. Evans, J.A. Osborn, F.A. Jardine and G. Wilkinson, Nature, 208, 1203 (1965). 9. P.S. Hallman, D. Evans, J.A. Osborn and G. Wilkinson, Chem. Comm., 305 (1967). 10. P.S. Hallman, B.R. McGarvey and G. Wilkinson, J. Chem. Soc. (A), 3143 (1968). 11. B. Hui, Ph.D. Dissertation, Univ. of British Columbia, Vancouver, 1969. 12. B. Hui and B.R. James, Proc. 4th Intern. Conf. on Organometallic Chem., Bristol, 1969, Abstr. L6. 13. J. Halpern, Adv. in Chem. Series, Vol. 70, Am. Chem. Soc, Washington, D.C, 1968, p. 1. 14. J. Halpern, Ann. Rev. Phys. Chem., 16, 103 (1965). - 189 -15. J. Halpern, Proc. 3rd Intern. Congr. Catalysis, Amsterdam, 1964, North Holland, Amsterdam, 1_, 146 (1965). 16. J. Halpern, Chem. Eng. News, 44^  68 (1966). 17. J. Halpern and B.R. James, Can. J. Chem., 4_4, 671 (1966). 18. J. Halpern, J. Harrod and B.R. James, J. Am. Chem. Soc, 88_, 5150 (1966). 19. R.G.S. Banks and J.M. Pratt, J. Chem. Soc. (A), 854 (1968) and references therein. 20. J. Kwiatek and J.K. Seyler, Adv. in Chem. Series, Vol. 70, Am. Chem. Soc, Washington, D.C., 1968, p. 207, and references therein. 21. L. Vaska, Accounts Chem. Res., 1_, 335 (1968), and references therein. 22. B.R. James, "Homogeneous Hydrogenation", Wiley, New York, in press. 23. R.S. Coffey, in "Aspects of Homogeneous Catalysis", Vol. 1, Carlo Manfredi, Milan, 1970, p. 5. 24. A. Andreetta, F. Conti and G.F. Ferrari, ibid., p. 204. 25. W.S. Knowles, M.J. Sabacky and B.D. Vineyard, Chem. Comm., 10 (1972). 26. J.D. Morrison, R.E. Burnett, A.M. Aguiar, C.J. Morrow and C. Phillips, J. Am. Chem. Soc, 93, 1301 (1971). 27. H. Hirai and T. Furuta, J. Polym. Sci. (B), 9_, 459 (1971); ibid., 2, 729 (1971). 28. P. Abley and F.J. McQuillin, J. Chem. Soc. (C), 844 (1971). 29. J. Manassen, Platinum Metals Rev., 15, 142 (1971) and references therein. 30. G. Natta and U. Giannini, "Encyclopedia of Polymer Science and Technology", Vol. 4, Interscience, New York, 1966, pp. 137-150. 31. G. Natta, Science, 147, 261 (1965). - 190 -32. P. Cossee, Tetrahedron Letters, 17, 12, 17 (1960); J. Catalysis, 3, 80 (1964); E.J. Arlman and P. Cossee, ibid., 3, 99 (1964). 33. C.W. Bird, "Transition Metal Intermediates in Organic Synthesis", Logos Press, London, 1967. 34. CP. Chiusoli in "Aspects of Homogeneous Catalysis", Vol. 1, Carlo Manfredi, Milan, 1970, p. 77. 35. G. Lefebvre and Y. Chauvin, ibid., p. 107. 36. R. Cramer, J. Am. Chem/ Soc, 87_» 4717 (1965). 37. B.R. James and J. Louie, Inorg. Chim. Acta, 3_, 568 (1969). 38. J. Louie, M.Sc. Thesis, Univ. of British Columbia, Vancouver, 1968. 39. T.A. Stephenson and G. Wilkinson, J. Inorg. Nucl. Chem., J28_, 945 (1966). 40. D. Rose, J.D. Gilbert, R.P. Richardson and G. Wilkinson, J. Chem. Soc. (A), 2610 (1969). 41. For a summary of preparative methods for these carbonyl complexes, see Ref. 74 and references therein, and Ref. 75. 42. D. Cook, Can. J. Chem., 42, 2721 (1964). 43. See for example, R.S. Nyholm, Proc. Chem. Soc, 273 (1961). 44. S.J. La Placa and J.A. Ibers, Inorg. Chem., 4^, 778 (1965). 45. D.D. Lehman, D.F. Shriver and I. Wharf, Chem. Comm., 1486 (1970). 46. J.A. Osborn, F.H. Jardine, J.F. Young and G. Wilkinson, J. Chem. Soc. (A), 1711 (1966). 47. D.R. Eaton and S.R. Suart, J. Am. Chem. Soc, 9_0, 4170 (1968). 48. T.H. Brown and P.J. Green, J. Am. Chem. Soc, 92_, 2359 (1970). 49. P. Meakin, J.P. Jesson and CA. Tolman, J. Am. Chem. Soc, 94, 3240 (1972). - 191 -50. V. Gutmann, "Coordination Chemistry in Non-Aqueous Solvents", Springer-Verlag, New York, 1968, pp. 31-34. 51. A.A. Frost and R.G. Pearson, "Kinetics and Mechanism", 2nd. ed., Wiley, New York, 1961, p. 135. 52. H. Aral and J. Halpern, Chem. Comm., 1571 (1971). 53. D. Evans, G. Yagupsky and G. Wilkinson, J. Chem. Soc. (A), 2660 (1968). 54. A.C. Skapski and P.R. Troughton, Chem. Comm., 1230 (1968). 55. DMA general information bulletin, Industrial and Biochemicals Department, E.I. DuPont de Nemours & Co., Wilmington, Delaware. 56. A.J. Carty, Can. J. Chem., 44_, 1881 (1966). 57. J.F. Harrod, D.F.R. Gilson and R. Charles, Can. J. Chem., 4_7, 1431 (1969). 58. A. Yamamoto, S. Kitazume and S. Ikeda, J. Am. Chem. Soc, 90, 1089 (1968). 59. W.H. Knoth, J. Am. Chem. Soc, 90_, 7172 (1968); ibid, 94, 104 (1972). 60. P.B. Chock and J. Halpern, J. Am. Chem. Soc, 88_, 3511 (1966). 61. L. Vaska and M.F. Werneke, Trans. M.Y. Acad. Sci., 31, 70 (1971). 62. J.F. Harrod and CA. Smith, J. Am. Chem. Soc, £2, 2699 (1970). 63. CA. Tolman, J. Am. Chem. Soc, j_2, 4217 (1970). 64. See Ref. 66 and references therein. 65. CD. Falk and J. Halpern, J. Am. Chem. Soc, 87_, 3523 (1965). 66. U. Belluco, M. Giustiniani and M. Graziani, J. Am. Chem. Soc, 89, 6494 (1967). 67. Ref. 51, pp. 132-137. 68. W. Strohmeier and S. Hohmann, Z. Naturforsch., 25b, 1309 (1970). 69. S. Siegel and D. Ohrt, Inorg. Nucl. Chem. Letters, _5, 15 (1972), and references therein. - 192 -70. C. O'Connor and G. Wilkinson, J. Chem. Soc. (A), 2665 (1968). 71. B.R. James and N.A. Memon, Can. J. Chem., k6, 217 (1968). 72. M.G. Burnett and R.J. Morrison, J. Chem. Soc. (A), 2325 (1971). 73. K. Nakamoto, "Infrared Spectra of Inorganic and Coordination Compounds", Wiley, New York, 1963, pp. 230-231. 74. B.R. James and L.D. Markham, Inorg. Nucl. Chem. Letters, _7, 373 (1971). 75. B.R. James, L.D. Markham, B. Hui and G.L. Rempel, to be published. 76. J.E. Lyons, J. Org. Chem., 36, 2497 (1971). 77. R.A. Schunn, Inorg. Chem., 9_, 2567 (1970), and references therein. 78. B. Hudson, D.E. Webster and P.B. Wells, J.C.S. Dalton, 1204 (1972); ibid., 1287 (1972). 79. R. Cramer, J. Am. Chem. Soc, 89_, 4621 (1967). 80. R.G. Denning, F.R. Hartley and L.M. Venanzi, J. Chem. Soc. (A), 324 (1967). 81. T. Yamamoto, A. Yamamoto and S. Ikeda, J. Am. Chem. Soc, 93, 3360 (1971). 82. J.R. Joy and M. Orchin, J. Am. Chem. Soc, 81, 310 (1959). 83. M.A. Muhs and F.T. Weiss, J. Am. Chem. Soc, 84, 4697 (1962). 84. R.J. Cvetanovic, F.J. Duncan, W.E. Falconer and R.S. Irwin, J. Am. Chem. Soc, 87, 1827 (1965). 85. F.H. Jardine, J.A. Osborn and G. Wilkinson, J. Chem. Soc. (A), 1574 (1967). 86. W. Strohmeier and R. Fleischmann, Z. Naturforsch., 24b, 1217 (1969). 87. J. Brandup and E.H. Immergut, "Polymer Handbook", Interscience, New York, 1966, p. 341. 88. S.D. Robinson and B.L. Shaw, Tetrahedron Letters, 1301 (1964); J. Chem. Soc. , 4997 (1965). - 193 -89. P. Uguagliati and W.H. Baddley, J. Am. Chem. Soc., 90, 5446 (1968). 90. A.J. Deeming, B.F.G. Johnson and J. Lewis, Chem. Comm., 598 (1970). 91. H.C. Clark and H. Kurosawa, Chem. Comm., 957 (1971); Inorg. Chem., 11, 1275 (1972). 92. J.G. Atkinson and M.O. Luke, Can. J. Chem., 48, 3580 (1970). 93. R.H. Prince and K.A. Raspin, J. Chem. Soc. (A), 612 (1969). 94. J.K. Nicholson and B.L. Shaw, Proc. Chem. Soc, 282 (1963). 95. J.E. Lyons, Chem. Comm., 562 (1971). 96. R.E. Rinehart, H.P. Smith, H.S. Witt and H. Romeyn Jr., J. Am. Chem. Soc, 83, 4864 (1961); ibid., 84, 4145 (1962). 97. A.J. Canale, W.A. Hewett, T.M. Shryne and E.A. Youngman, Chem. and Ind., 1054 (1962). 98. A.J. Canale and W.A. Hewett, J. Polym. Sci. (B), 2_, 1041 (1964). 99. R. Cramer, J. Am. Chem. Soc, 87, 4717 (1965). 100. R.R. Hampton, Anal. Chem., 21, 923 (1949). 101. For example, see Ref. 21. 102. J.C.W. Chien, J. Am. Chem. Soc, 81, 86 (1959). 103. H.A. Tayim and J.C. Bailar, J. Am. Chem. Soc, 89, 4330 (1967). 104. K.C. Ramey and G.L. Statton, J. Am. Chem. Soc, 88_, 4387 (1966). 105. K. Hiraki and H. Hirai, Macromolecules, 3, 382 (1970). 106. G.W. Parshall, Accounts Chem. Res., _3» 139 (1970). 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items