UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Activation of molecular hydrogen, molecular oxygen, and olefins by solutions containing some univalent… Chan, Cheuk-Yin 1974

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

Download

Media
831-UBC_1974_A1 C49_3.pdf [ 8.72MB ]
Metadata
JSON: 831-1.0061133.json
JSON-LD: 831-1.0061133-ld.json
RDF/XML (Pretty): 831-1.0061133-rdf.xml
RDF/JSON: 831-1.0061133-rdf.json
Turtle: 831-1.0061133-turtle.txt
N-Triples: 831-1.0061133-rdf-ntriples.txt
Original Record: 831-1.0061133-source.json
Full Text
831-1.0061133-fulltext.txt
Citation
831-1.0061133.ris

Full Text

ACTIVATION OF MOLECULAR HYDROGEN, MOLECULAR OXYGEN, AND OLEFINS BY SOLUTIONS CONTAINING SOME UNIVALENT IRIDIUM COMPLEXES BY CHEUK-YIN CHAN B.Sc. (Magna cum laude), Chinese University of Hong Kong, 1967 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 January 1974 In presenting th i s thesis in pa r t i a l fu l f i lment of the requirements for an advanced degree at the Univers i ty of B r i t i s h Columbia, I agree that the L ibrary shal l make it f ree ly ava i lab le for reference and study. I further agree that permission for extensive copying of th is thesis for scholar ly purposes may be granted by the Head of my Department or by his representat ives. It is understood that copying or pub l i cat ion of th is thesis for f inanc ia l gain sha l l not be allowed without my written permission. Department of The Univers i ty of B r i t i s h Columbia Vancouver 8, Canada / - i i -ABSTRACT Kinetic and spectroscopic studies on solutions of two iridium(I) complexes—trans-chlorocarbonylbis(triphenylphosphine)iridium(I), Ir(CO)Cl(PPb-2)2» a n d y-dichlorotetrakis (cyclooctene)-di-iridium(I) , [IrCl(CQH,.) 0].—are described, especially for reactions involving activa-tion of molecular E^, molecular O^ , and olefins. The studies also i l l u s t r a t e the importance of solvent effects. The catalytic activity of Ir(CO)Cl(PPh 3) 2 for hydrogenation of maleic acid has been surveyed using a range of solvents—pyridine, dimethylsulfoxide (DMSO), dimethylacetamide (DMA), dimethylformamide (DMF), acetone, sulfolane, acetonitrile, nitromethane and formamide. Where activity i s observed, the mechanism appears to involve activation of hydrogen by a square-planar four-coordinate Ir(I) ol e f i n complex. The DMA, DMF and DMSO solvent systems, which are very similar in terms of coordinating a b i l i t y and di e l e c t r i c constant, do show catalytic activity and this results from the dissociation of a phosphine molecule from the iridium at some stage to form the required four-coordinate catalyst 1_: Ir(CO)Cl(PPh.) + olefin -PPh Ir(CO)Cl(PPh 3) 2(olefin) Ir(CO)ClPPh.(olefin) 1 Ir(CO)(PPh 3) 2(olefin) + 2 - i i i -The sulfolane system is more active than the DMA, DMF and DMSO systems, but shows much more complicated kinetics. The hydrogenation appears to proceed in part via the phosphine dissociation path outlined in the above scheme, but the major pathway involves a cationic i n t e r s mediate Ir (CO) (PPh^)^ ( o l e f i n ) + , 2_, formed via chloride dissociation from the five-coordinate o l e f i n complex. Diethylmaleate is hydrogenated in sulfolane, however, primarily via the phosphine dissociation path. Solvents that are too strongly coordinating (pyridine) or too weakly coordinating (nitromethane) lead to catalytically inactive systems. The catalytic homogeneous hydrogenation of hexene-1, cyclooctene using DMA solutions of [IrCl(CgH.^) 2] 2 involves a monomeric species. The strongly coordinating solvent or the added olefin are thought to cleave the chloride bridge in [ I r C l ^ g H ^ ^ J ^ * The hydrogenation mechanism can be outlined as k 2 Ir + olefin ^ Ir (olefin) » [H.Ir(olefin) ] — ^ I _ r + product H (1) where _Ir_ is a complex already containing coordinated olefin. Selective hydrogenation of cyclooctene in a mixture of cyclooctene and hexene-1, the catalytic isomerization of hexene-2 and the catalytic hydrogenolysis of molecular 0_ to water, a l l using [IrCl(C 0H .) ] complex in DMA are described and discussed. Molecular 0 is activated by DMA solutions of [IrCl(C 0H. ) ] £• o IH 2. 2 containing excess chloride; the major species believed to be present in solution is [IrCl 2(CgH^^) 2] . The solution i n i t i a l l y absorbs one 0 2 per - i v -ir i d i u m . Product c h a r a c t e r i z a t i o n proved to be d i f f i c u l t but the solutions c a t a l y t i c a l l y o x i d i z e cumene l i k e l y v i a a hydroperoxide free r a d i c a l intermediate and the data are discussed i n terms of the follow-ing e q u i l i b r i a : — • ° 2 ^ l r ( 0 2 ) ^ l r ( l l ) 0 2 - (2) During some preliminary studies to i n v e s t i g a t e p o s s i b l e a c t i v a t i o n of CO under mild conditions using I r complexes i n aqueous s o l u t i o n s , the iri d i u m ( I I I ) dicarbonyl [Ph^As ] +[Ir(C0) 2C1^]~'2H 20, and a new c l u s t e r carbonyl t e n t a t i v e l y formulated [Ir(CO) 2] n» were synthesized. - v -TABLE OF CONTENTS Page ABSTRACT i i TABLE OF CONTENTS v LIST OF TABLES ix LIST OF FIGURES x i i ABBREVIATIONS xv ACKNOWLEDGEMENTS xv i i CHAPTER I. INTRODUCTION 1 1.1 Aim of Work 1 1.2 Homogeneous Hydrogenation 2 1.2.1. Heterolytic Splitting 3 1.2.2 Homolytic Splitting 4 1.2.3 Insertion 4 1.3 Catalytic Hydrogenation Using Iridium Complexes 5 1.4 Oxygen Ligand in Group VIII Transition Metal Complexes 7 References 10 CHAPTER II. APPARATUS AND EXPERIMENTAL PROCEDURES 13 11.1 Materials 13 11.1.1 Iridium Compounds 13 11.1.2 Organic Ligands and Substrates 14 1 II.1.3 Gases 14 11.1.4 Solvents 14 11.1.5 Other Materials 15 11.2 Gas Uptake Measurements 15 11.2.1 Apparatus for Gas Uptake Measurements 15 11.2.2 Procedure for a Typical Gas Uptake Experiment 17 11.2.3 Gas Solubility Measurements 19 - v i -11.3 Identification of Liquids and Solid Products 19 11.4 Instrumentation 20 References 21 CHAPTER III. SOLVENT EFFECTS ON HOMOGENEOUS HYDROGENATION CATALYZED BY VASKA'S COMPOUNDS, Ir(CO)Cl(PPh 3) 2 22 111.1 Introduction 22 111.2 Properties of Ir(C0)ClP 2 24 111.2.1 Dissociation of P in Various Solvents 24 a Pyridine 24 b DMSO, DMF and CH CN 28 c DMA 34 d Acetone and Sulfolane 36 e CH3N02 36 111.2.2 Reaction with Molecular H 2—Formation of the Dihydride Complex 36 111.2.3 Formation of an Olefinic Complex with MA 39 a Pyridine 41 b DMA 41 c DMSO and DMF 43 d Acetone 43 e Sulfolane 45 f CH3CN and CH3N02 45 111.3 Catalytic Hydrogenation of MA using Ir(CO)ClP 2 47 111.4 Discussion 53 References 65 CHAPTER IV. CATALYTIC HYDROGENATION OF OLEFINS USING Ir(CO)ClP 2 IN SULFOLANE SOLUTION 67 IV.1 Introduction 67 IV.2 Kinetic Measurements on Catalytic Hydrogenation of MA and DEMA Using Ir(C0)ClP 2 in Sulfolane 67 IV.2.1 Rate Dependence on Concentration of Catalyst, Substrate and Hydrogen 68 IV.2.2 Rate Dependence on Temperature 77 - v i i -IV.2.3 Effect of Addition of P on Hydrogenation Rate 77 IV.2.4 Effect of Added Acid on Hydrogenation Rate 83 IV.2.5 Effect of Acid Salts on Hydrogenation Rate 83 IV.3 Catalytic Hydrogenation of Other Olefins and Acetylenes 90 IV.4 Ir(C0)ClP 2 and i t s Properties in Sulfolane Solution 90 IV.4.1 Physical Properties in Sulfolane 90 IV.4.2 Reaction with Hydrogen to Form a Dihydride 93 IV.4.3 Reaction with Substrates: MA and DEMA 95 a Formation of Olefin Complex 95 b Dissociation of C l ~ from the Olefin Complex 100 c Conductance Measurements on the Hydride and the Olefin Complex in Sulfolane 102 d Possibility of Proton Dissociation from the Hydride and Olefin Complexes 102 IV. 5 Discussion 107 IV.5.1 Dihydride Formation 107 IV.5.2 Olefin Complexation 109 IV.5.3 Kinetic Data 113 a MA Substrate 114 b DEMA Substrate 120 IV. 5.4 General Conclusions 123 Appendix IV-1 127 Appendix IV-2 129 Appendix IV-3 132 Appendix TV-4 134 References 136 CHAPTER V. CATALYTIC HYDROGENATION REACTIONS USING THE Ir(I) CYCLOOCTENE COMPLEX IN DMA 138 V. l Introduction 138 V.2 Catalyzed Hydrogenation Using [IrCl(C0H-.)„]„ 1 in DMA 8 14 2 2 1 3 g V. 2.1 Reaction of 1 with H 2 138 V.2.2 Catalyzed Hydrogenation of Hexene-1 141 V.2.3 Selective Hydrogenation of a Mixture of Hexene-1 and Cyclooctene 149 V.2.4 Catalytic Hydrogenation of Other Monoolefins 153 V.2.5 Catalytic Isomerization 153 - v i i i -V.3 Discussion 157 V. 4 Catalytic Reduction of Molecular 0^ 169 Appendix V-l 174 References 175 CHAPTER VI. ACTIVATION OF MOLECULAR OXYGEN BY SOLUTIONS OF u-DICHLOROTETRAKIS(CYCLOOCTENE)DI-IRID IUM(I) 177 VI. 1 Introduction 177 VI.2 Experimental Results for the DMA System 178 VI.2.1 Kinetic and Stoichiometry 178 VI.2.2 Spectroscopic Measurements 187 a Infrared 187 b Visible 188 c ESR 191 VI.2.3 Oxidation of Substrates by 0.2 M LiCl/DMA Solution of 1 191 VI.3 Reactivity of _1 Toward 0„ in Benzene, Chloroform and Ethylene Chloride Solutions 194 VI. 4 Discussion 198 References 210 CHAPTER VII. ATTEMPTS FOR PREPARATION OF SOME IRIDIUM CARBONYL COMPLEXES 212 VII. 1 Introduction 212 VII.2 Tetraphenylarsonium tetrachlorodicarbonyliridate(III) 213 VII.3 An Iridium Carbonyl [Ir(C 0 ) 2 l n ? 215 References 217 - ix -LIST OF TABLES Table Page I I I - l Spectrophotometric study of equilibrium reaction (3.3) in pyridine solution 29 III-2 Spectrophotometric study of equilibrium reaction (3.3) in DMSO, DMF and CH CN solvents 33 III-3 Spectrophotometric data in DMA (freshly d i s t i l l e d ) solvent in vacuum at 35° 35 III-4 Spectrophotometric study of equilibrium reaction (3.3) in acetone, sulfolane 37 III-5 Spectrophotometric study of dihydride formation 40 III-6 Spectrophotometric study of ol e f i n complex formation in different solvents 42 III-7 Catalytic hydrogenation of MA using Ir(C0)ClP 2 in different solvents 48 III-8 Catalytic hydrogenation of MA using Ir(CO)C1P in DMSO at 80° 51 III-9 Catalytic hydrogenation of MA using Ir(C0)ClP 2 and related properties in different solvents 57 +2 111-10 Dq values of six-coordinate Ni complexes in different solvents 61 IV-1 Kinetic data for the Ir(C0)ClP. catalyzed hydrogenation of MA at 80° 70 IV-2 Kinetic data for the Ir(C0)ClP ? catalyzed hydrogenation of DEMA at 80° 71 IV-3 Temperature dependence of hydrogenation rate (MA ' system) 78 IV-4 Temperature dependence of hydrogenation rate (DEMA system) 78 IV-5 Effect of added P on rate of hydrogenation (MA system) 80 Effect of added P on rate of hydrogenation (DEMA system) 80 Effect of added P on rate of hydrogenation (MA system) 82 Effect of added acid on hydrogenation rate (MA system) 84 Effect of added acid on hydrogenation rate (DEMA system) 84 Variation of hydrogenation rate on added L i C l (MA system) 87 Variation of hydrogenation rate on added LiNO^ (MA system) 87 Variation of hydrogenation rate on added L i C l (DEMA system) 89 Effect of added salts and other reagents on hydrogenation rate (DEMA system) 91 Hydrogenation of olefins and acetylenes using Ir(C0)ClP 2 in sulfolane 92 Equilibrium constant for dihydride formation 94 Equilibrium constants for equilibrium (4.2) (MA system) 97 Equilibrium constant for equilibrium (4.2) (DEMA system) 101 Determination of Cl dissociation in sulfolane (DEMA system) 103 Conductance measurements i n sulfolane 106 Kinetic data for the Ir(C0)ClP~ catalyzed hydrogenation in DMA 119 [IrCl(C_H.., ).]„ catalyzed hydrogenation of hexene-1: kinetic data at 25° i n DMA 144 Ir* catalyzed hydrogenation of hexene-1 in DMA— Temperature dependence 150 Conductance measurements in DMA at 25° 151 Catalyzed hydrogenation of CgH^ by 1_ in DMA 154 Catalyzed hydrogenation of monoolefins using [ I r C l ( C 8 H 1 4 ) 2 ] 2 in DMA 156 - x i -V- 6 [ I r C l ( C 0 r L .)„]„ catalyzed reduction of 0„ i n DMA 170 VI- 1 Reaction of [IrClCCgH-j^^lo with oxygen; k i n e t i c data at 25° i n 0.2 M L i C l (DMA) 181 VI-2 Oxygenation of [ I r C l ( C 8 H 1 4 ) 2 ] 2 i n DMA at 25°; k i n e t i c data for v a r i a t i o n of L i C l concentrations 184 VI-3 V a r i a t i o n of k 2 with temperature 186 VI-4 K i n e t i c data f o r the oxygenation of [IrCl(C 0H-.)„] 0 l n benzene 196 - x i i -LIST OF FIGURES Figure Page II- l Apparatus for constant gas-uptake measurements 16 III- l Visible absorption spectra of Ir(CO)ClP_ in pyridine at 70° under N 2 25 26 III-2 Variation of optical density of Ir(C0)ClP 2 at 387 nm in pyridine at 70° under N 2 III-3 Visible absorption spectra of Ir(C0)ClP 2 in various solvents under N 2 30 III-4 Optical density plot at 387 nm for P dissociation of Ir(C0)ClP 2 in DMF at 35° under N 2 32 III-5 Optical density plots for Ir(C0)ClP in CH,N0~ under N 2 at 25° 38 III-6 Visible absorption spectrum of Ir(C0)ClP in CH CN at 35° under N 2 J 46 III-7 H 2 uptake plot of Ir(C0)ClP 2 and MA in different solvents under 1 atm H 2 49 III- 8 Dependences of hydrogenation rate on [Ir], [MA] and [H 2] i n DMSO at 80° 52 IV- 1 H 2 uptake plot for hydrogenation of MA using Ir(C0)ClP 2 at 80° and p[H 2] = 760 mm Hg in sulfulane 69 IV-2 Variation of hydrogenation rate vs. concentration of Ir(C0)ClP 2 73 IV-3 Variation of hydrogenation rate vs. concentration of MA 74 IV-4 Variation of hydrogenation rate vs. concentration of DEMA 74 i IV-5 Variation of hydrogenation rate vs. pressure of H 2 (MA system) 75 IV-6 Variation of hydrogenation rate vs. pressure of H 2 (DEMA system) 76 IV-7 Arrhenius-type plot for Ir(C0)C1P2-MA-H2 system 79 IV-8 Arrhenius-type plot for Ir(C0)C1P2-DEMA-H2 system 79 - x i i i -IV-9 Variation of hydrogenation rate vs. added [P] for Ir(CO)ClP2-MA-H2 system . 81 IV-10 Variation of hydrogenation rate vs. added [P] for Ir(CO)ClP -DEMA-H2 system 81 IV-11 Variation of hydrogenation rate on added acid (MA system) 85 IV-12 Variation of hydrogenation rate on added acid (DEMA system) 86 IV-13 Variation of hydrogenation rate on added L i C l or LiNO„ (MA system) 88 F/-14 Variation of hydrogenation rate on added L i C l (DEMA system) 89 IV-15 Change of absorption for Ir(CO)ClP -MA in sulfolane under N 2 98 IV-16 First-order log plot of absorption change for the Ir(CO)ClP2-MA system 99 IV-17 Apparatus set up for determination of [Cl ] and acid-base t i t r a t i o n in sulfolane 104 IV-18 [Cl ] calibration curve in the potentiometric method 105 IV- 19 Potentiometric acid-base t i t r a t i o n in sulfolane at 80° 108 V- l H 9 uptake plot of [IrCl(C 0H. . ) J 0 in DMA at 25° 139 o 1H 2. L V-2 Visible absorption spectra of [ IrCl (C„H.. , ) A „ recorded in DMA at 25° 142 V-3 H 2 uptake plot of [ I r C l ( C g H 1 4 ) 2 ] 2 i n DMA at 25° 143 V-4 Dependence of linear hydrogenation rates of hydrogen pressure in DMA at 25° 145 V-5 Dependence of linear hydrogenation rates of [Ir] concentration in DMA at 25° 146 V-6 Dependence of linear hydrogenation rates of [hexene-1] in DMA at 25° 147 V-7 H 2 uptake plot of [I r C l ( C 8 H 1 4 ) 2 ] 2 in DMA at 25° with high concentration of ole f i n 148 V-8 H 2 uptake plot of [IrCl(C H ) ] in DMA at 25° with added L i C l ^ 1 152 - xiv -V-9 H2 uptake plot of [IrCl (CgH - i ^ ^ ^ and mixture of cyclooctene and hexene-1 in DMA at 25° 155 V-10 Plot of rate" 1 vs. [hexene-1]"1 for [IrCl(C^IL , ) ~ ] 9 in DMA at 25° ^ 159 V - l l Arrhenius plot of [IrCl (CgH-^^^ in DMA 160 V- 12 H 2 uptake plot of [ I r C l ( C 8 H 1 4 ) 2 ] 2 in DMA at 65° with mixture of IL, and O2 170 VI- 1 0 0 uptake plot of [IrCl(C QH, .)„]„ in 0.2 M LiCl/DMA solution at 25 179 VI-2 Dependence of i n i t i a l rate on [Ir] in [IrCl(CgH,,) 2] ?-02-LiCl/DMA system at 25° 182 VI-3 Dependence of i n i t i a l rate on O2 pressure in [IrCl(C 8H 1 4) 2]2-0 2-LiCl/DMA system at 25° 183 VI-4 Dependence of i n i t i a l rate on [LiCl] in the [IrCl(C 8H 1 4) 2] 2-0 2-DMA solution at 25° 184 VI-5 Arrhenius plot of [IrCl(C 8H 1 4) 2] 2-0 2-LiCl/DMA system 186 VI-6 Visible absorption spectra of [ I r C l ( C 8 H 1 4 ) 2 ] 2 in 0.2 M LiCl/DMA at room temperature under 0 2 189 VI-7 Variation of O.D. at 620 nm for [IrCl(CgH 1 4) 2] 2-02~ LiCl/DMA solution at R.T. and the corresponding log plot 190 VI-8 Esr spectrum of [ I r C l ( C 8 H 1 4 ) 2 ] 2 - 0 2 - L i C 1 / D M A after exposed to air for 6 min. and cooled to li q u i d N 2 temperature 192 VI-9 Esr measurement at 3245 G for [ I r C l ( C 8 H 1 4 ) 2 ] 2 - L i C l / DMA solution under a i r at liquid ^ temperature 193 VI-10 0 2 uptake plot of [IrCl(CgH 1 4) 2]2 i n benzene at 25°, and the corresponding log plot 195 - XV -ABBREVIATIONS The following l i s t of abbreviations, most of which are commonly adopted in chemical research literature, w i l l be employed in this thesis. A l l temperatures are in °C unless specifically denoted °K. atm atmosphere calc calculated DEMA diethyl maleate, Et02CCH=CHC02Et DMA dimethylacetamide, CH3CON(CH3)2 DMF dimethylformamide, HCON(CH3)2 DMSO dimethylsulfoxide, (CH3)2SO eq. or eqn. equation esr electron spin resonance Et ethyl, -C 2H 5 FA fumaric acid, trans-butenedioic acid, trans-H02CCH=CHC02H Fig figure g gyromagnetic ratio hr hour i r infrared L ligand log common logarithm M molar MA - ' male% acid, cis-butenedioic acid, c i s -H 0CCH=CHC0oH 2 I - xvi -Me methyl, -CH3 min minute m.pt. melting point m.wt. molecular weight _9 nm nanometer (10 meter) = 1 millimicron (my) O.D. optical density, absorbance ol olefin P triphenylphosphine unless sp e c i f i c a l l y denoted Ph phenyl, "CgH^ py pyridine R alkyl R.T. room temperature S solvent SA succinic acid sec second T or temp temperature UV ultra violet VIS v i s i b l e vpc vapor phase chromatography e molar extinction coefficient v frequency, cm 1 - x v i i -ACKNOWLEDGEMENT I wish to thank Dr. B. R. James for his expert guidance and inexhaustive encouragement throughout the course of this work. CHAPTER I INTRODUCTION 1.1 Aim of Work The aim of work undertaken for this thesis was to investigate the use of certain iridium complexes in solution for the activation of molecular hydrogen, molecular oxygen and to a lesser extent carbon monoxide. Detailed kinetics of some of the reactions were then investigated and reaction mechanisms have been postulated. Some results obtained by earlier workers (14) i n this laboratory showed that the Ir(I) complex, Ir(C0)Cl(PPh.j)2> i s a homogeneous hydrogenation catalyst i n DMA, although the complex i s inactive i n benzene. More detailed solvent effects on this catalytic system were thus investigated and these studies are described i n Chapter III. Such studies led to an interesting finding in sulfolane solvent; a quite different mechanism i s involved for the corresponding hydrogenation system studied in DMA. The detailed kinetics and mechanism of the sulfolane system are described i n Chapter IV. The catalytic behavior of the dimeric Ir(I) cyclooctene complex, [IrCl(C 0H 1.) in coordinating solvent was investigated also. It was used as a catalyst for hydrogenation of o l e f i n i c bonds (Chapter V) and the reaction of the dimer with molecular oxygen is described in Chapter VI. - 2 -Some efforts were made to synthesize iridium carbonyl complexes from the interaction of CO with iridium halides under mild conditions. Two new iridium carbonyl complexes (one of which could not be reproduced) have been prepared, and these studies are described i n Chapter VII. At the beginning of this work in early 1969 there was relatively l i t t l e reported regarding the catalytic aspects of iridium complexes in solution. However, since that time there has been a remarkable amount published on this topic. Some of these catalytic properties w i l l be reviewed bri e f l y in the following sections. 1.2 Homogeneous Hydrogenation Increasing attention has been paid in the last decade to homogeneously catalyzed reactions involving complexes of transition metals: hydrogenation, hydroformylation and other carbonylation reactions, polymerization, oxidation of alkenes, fixation of nitrogen, etc. A number of factors, for example, the mode of activation of the substrate, the nature of the active species, etc., are common to these processes. The investigation of the systematic features of these processes i s therefore of general importance. The problem of homogeneous hydrogenation is also of practical interest. For example, the use of molecular hydrogen in the polymerization of alkenes by means of Ziegler-Natta catalysts makes i t possible to control the molecular weight of the polymers formed (1), and homogeneous hydrogenation catalysts have been used i n fuel cells (2). Similarity to enzymatic systems (3) is another feature of some catalytic processes which has recently served to encourage their study. Although heterogeneous catalysts are currently - 3 -more prevalent in industrial use, research workers have long recognized the advantages of homogeneous over heterogeneous hydrogenation in terms of milder conditions of process, combined with high efficiency, selectivity, and stereospecificity, and convenience for kinetic studies. In recent years, the number of known homogeneous hydrogenation catalysts has increased considerably and this has made possible a deeper and more complete study of a number of problems relating to the mechanism of hydrogenation. In general, homogeneous hydrogenation with participation of molecular hydrogen involves two inter-related stages: (a) the activation of molecular hydrogen, and (b) the actual hydrogenation, including, activation of the substrate and the transfer of activated hydrogen to the substrate. Three distinct mechanisms by which hydrogen activation occur have been recognized (4) and these are exemplified by the following reactions. 1.2.1 Heterolytic Splitting (5) Ru(III)Cl, 3" + H. * Ru(III)HCl c 3" + H + + C l " (1.1) This involves essentially a substitutional process (replacement of a chloride ligand by a hydride derived from ri^) without change i n the formal oxidation number of the metal. Reactivity i s thus governed by the substitution l a b i l i t y of the complex, by the s t a b i l i t y of the hydride formed, and by the presence of a suitable base to stablize the released proton (6). - 4 -An o l e f i n molecule which has become coordinated to the metal may then undergo i n s e r t i o n into the metal-hydride bond, y i e l d i n g a a - a l k y l complex. For example, a Ru(II) c h l o r i d e c a t a l y s t system has been interpreted (7) i n terms of the following mechanism: H \ / rU I H — Ru — > - R u > -Ru-C " | c '\ y \ / N* H H + H + J x . ! 1 ^ — R u - + - C — C - (1.2) ' I I I 1.2.2 Homolytic S p l i t t i n g (8,9) 2Co(II) ( C N ) 5 3 ~ + H 2 2Co(III)H(CN) 5 3" (1.3) Here the hydride formation i s accompanied by formal oxidation of the metal, and r e a c t i v i t y i s c l o s e l y l i n k e d to the s u s c e p t i b i l i t y of the metal to oxidation. In t h i s p a r t i c u l a r case, an a l k y l complex formed v i a o l e f i n i n s e r t i o n reacts subsequently with a further mole of hydride complex to produce the saturated product, f o r example (10), [ ( N C ) 5 C o - a l k y l ] 3 ~ + [HCo(CN) 5] 3~ > 2[Co(CN) 5] 3~ + alkane (1.4) 1.2.3 Insertion (formation of dihydride) (11,12) t r a n s - I r ( I ) ( C 0 ) C l ( P P h o ) o + H_ * Ir(III)H.(C0)Cl(PPh_)_ 3 2 2 v 1 i 2. (1.5) - 5 -It i s a prototype of oxidative-addition, the oxidation number of the central metal increased by two. An olefin coordinated to such a dihydride may be reduced by consecutive transfer of two hydrogen atoms, via an o-alkyl hydride intermediate (13,14). A recent text on homogeneous hydrogenation covers the literature to 1972, and includes a survey of the reduction of both organic and inorganic substrates (15). Other articles of note include one review on selective hydrogenation of dienes and polyenes (16), asymmetric hydrogenation of olefins (17-19,34)and homogeneous catalysts supported on resins and polymers (20). These topics are a l l considered i n the text by James (15). ,1.3 .Catalytic Hydrogenation Using ..Iridium Complexes In 1962, Vaska and Diluzio(21) discovered that a synthetic complex of iridium(I) (11) reacted reversibly with hydrogen to give an isolable and stable dihydride adduct, formally an iridium(IIE) complex (see eq. 1.5). This i s now the prototype of the so-called oxidative-8 addition reaction of square planar d complexes which are also reactive toward a wide range of small gas molecules. The importance of these complexes to hydrogenation and homogeneous catalysis i s clear; such systems offer the opportunity of directly observing the electronic and stereochemical properties of a metal-gas-activated adduct and of obtaining an understanding of the factors that determine the reversible activation of small covalent gas molecules by transition metal complexes. Reviews that stress the importance of these oxidative-addition reactions in homogeneous catalysts generally, including hydrogenation include those - 6 -written by Vaska (22,23), Halpern (13,24-26), Collman (27,28), Collman and Roper (29), Carra and Ugo (30), Cramer (31), Strohmeier (32) and Wilkinson (33). The recent review ( 23 ) on the activation of gaseous molecules by platinum metal complexes, stresses the role of metal; Ir(I) and Rh(I) systems as well as 0s(0), Ir(I) and Pt(II) systems were compared. Studies on the hydrogenation of olefins and acetylenes using Ir(I) compounds, especially trans-[Ir(C0)XL2] (X = halogen, L = phosphines, phosphites, etc.) complexes have been reported by James and Memon (14), Vaska and coworkers (35,36), Strohmeier and coworkers (37-40) and Yamaguchi (41). James and Memon (14) observed very slow hydrogenation of alkenes in benzene solution of Ir(CO)X(PPh3)2 but noted enhanced activity for these substrates in DMA solution. The reduction of unsaturated carboxylic acid was studied in detail, and rate-laws and mechanisms were presented. An important finding was the slow dissociation of a phosphine ligand, with the formation of a "vacant s i t e " in the catalyst. This system appears to be the most studied and well characterized so far (see Chapter III). The relationship between various coordinating solvents and formation of the vacant site i s further described i n Chapter III. In recent years, a considerable number of iridium complexes have been reported as effective hydrogenation catalysts. The text by James (15) reviews these systems and covers the catalysts containing phosphine, phosphite, arsine, hydride, carbonyl and nitrosyl ligands as well as simple salts in both aqueous and non-aqueous systems. In general, the mechanisms and controlling factors for effective reduction - 7 -of unsaturated compounds using these catalysts remain to be f u l l y established. 1.4 Oxygen Ligand in Group VIII Transition Metal Complexes Molecular oxygen functions both as a ligand and as a reagent in transition metal organometallic chemistry. Recent interest i n the chemistry of molecular oxygen has concerned particularly both biochemists. interested i n biological oxygen transport and oxidase functions, and also, the industrial chemists interested i n developing homogeneous analogs to heterogeneous metal-catalyzed oxidation reactions. Both areas are involved to a considerable extent i n the study of the interaction of molecular oxygen with group VIII transition metal complexes, and partly so as model system for the more complex biological systems (42-44). Coordination of oxygen does not necessarily activate the oxygen for catalytic reaction, but the p o s s i b i l i t y of such activation does exist and i s the subject of considerable current research. Two reviews (45,46) giving a survey of oxygen compounds of platinum 8 10 metals particularly d and d systems have appeared recently. The 8 10 addition of molecular oxygen to d and d systems can be discussed on the basis of the Tr-bonding scheme of Chatt and Duncanson (47) and Dewar (48) I n i t i a l l y used to explain the coordination of the ethylene molecule to platinum in Zeiss's salt. Such bonding was further discussed by G r i f f i t h (49) for oxygenated haem, and has been discussed also by Ibers and coworkers (50) for complexes such as I r ^ O K l t P P h , ^ ^ ) . A a-bond is formed by donation from a f i l l e d Tr-bonding orbital on the ligand to an unfilled metal d or hybrid orbital. Back donation then - 8 -occurs, to offset the resultant dipole, from a f i l l e d metal orbital to * a TT -antibonding orbital on the ligand molecule. The two complexes Ir(CO)X(PPh 3> 2 (X = Cl or I) (50) show differences i n oxygen uptake properties,and oxygen adducts show differences in molecular geometry. The energy of the iridium orbitals i s varied by the change in the halide substituent. With the more electronegative chlorine atom the overlap between the metal and oxygen orbitals i s such little that^back-donation to oxygen occurs and the 0-0 bond is lengthened to o 1.30 A. With the less electronegative iodine substituent greater o overlap results in an 0-0 bond length of 1.51 A (the bond length i n 0 2» e _ _ o 0=0 i s 1.20 A, the peroxide ion bond length ~0-0~ is 1.49 A). The reactivities of different transition metals i n the same periodic subgroup, [M(2-phos>2]+ (M = Co., Rh, Ir; 2-phos = cis-Phg-P-CHCH-PPh2) toward 0 2 and H 2 have been compared by Vaska et a l (51). In these simple addition reactions, eq. (1.6), the reactivity of the activator complex and the s t a b i l i t y of the resulting adduct are generally observed to show the sequence third- > second-row transition metal. A comparison with the first-row metal has apparently not been possible thus far because of an absence of a suitable series of isostructural and isoelectronic complexes of a l l three metals, i.e. compounds which di f f e r only i n the metal, exhibit the same reaction path, and react at measurable rates. The study of the reaction (1.6), showed the reactivity of the [M(2~phos)2] complex and the adduct s t a b i l i t y i s in [M(2-phos) 2] + + XY ^ [(XY)M(2-phos) 2] + (1.6) M = Co, Rh, Ir; XY = H 2 > 0 2. - 9 -the sequence of Co ^  Ir > Rh (51). It was suggested (51) that the reactivity of the d M(I) complexes i s directly dependent on their ligand f i e l d stabilization energies. The most significant property of certain of the oxygen complexes is their a b i l i t y to oxygenate substrates under unusually mild conditions. Under mild conditions with no catalyst, molecular oxygen does not react with S0 2, C02, CS 2 > CO, N02, RSCN, RNC, RCHO, R2CO or PPho,. 0 xY8 e n complexes of group VIII transition metals, however, do react with a l l these substrates to give oxygenated substrates (45,46). A number of factors could contribute to the increase in reactivity of molecular oxygen upon coordination (45): (a) coordinated oxygen i s i n general diamagnetic and therefore, reactions with diamagnetic substrates to form diamaSrVg^ etic products are not hindered by the requirement of spin conservation; (b) the metal may hold oxygen and the substrate in cis positions, lowering the activation energy for oxidation of the substrate; (c) coordinated oxygen i s , in most cases, pa r t i a l l y reduced, and the increased electron density on the 0 2 may activate i t . Catalytic oxygenation by transition metal complexes via molecular oxygen complexes w i l l be more f u l l y discussed in Chapter VI. - 10 -REFERENCES 1. G. Natta, G. Mazzanti, P. Longi and F. Bernardini. Chim. e. Ind., Al, 519 (1959). 2. H.F. McDuffie, E.L. Compere, H.H. Stone, L.F. Woo, and CH. Secoy, J. Phys. Chem., 62, 1030 (1958). 3. R.F. Gould, ed. "Bioinorganic Chemistry", Adv. Chem. Series, Vol. 100, Amer. Chem. Soc, Washington, D.C., 1971. 4. J. Halpern, Ann. Rev. Phys. Chem., 16, 103 (1965). 5. J. Halpern and B.R. James, Can. J. Chem., 44, 671 (1966). 6. J. Halpern and J.B. Milne, Proc. Intern. Congr. Catalysis, 2nd, Paris, 445 (1960). 7. J. Halpern, J. Harrod and B.R. James, J. Amer. Chem. Soc, 88, 5150 (1966). 8. M.G. Burnett, P.J. Connolly and C.J. Kembell, J. Chem. Soc. (A), 800 (1967). 9. J. Kwaiatek, I.L. Mador and J.K. Seyler, Adv. Chem. Series, 37, 201 (1963). 10. J. Kwaiatek and J.K. Seyler, Adv. Chem. Series, 70, 207 (1968). 11. L. Vaska and J.W. Diluzio, J. Amer. Chem. Soc, J33, 2784 (1961). 12. P.B. Chock and J. Halpern, J. Amer. Chem. Soc, 88, 3511 (1966). 13. J. Halpern, Adv. Chem. Series, 70, 1 (1968). 14. B.R. James, and N.A. Memon, Can. J. Chem., 46, 217 (1968). 15. B.R. James, "Homogeneous Hydrogenation", Wiley, New York, 1973. 16. A. Andreetta, F. Conti and G.F. Ferrari, in "Aspects of Homogeneous Catalysis", ed. R. Ugo, Vol. 1, Carlo Manfred!, Milan, 1970, p. 204. - l i -l y. W.S. Knowles, M.J. Sabacky and B.D, Vineyard, Chem. Comm., 10, (1972) . 18. J.D. Morrison, R.E. Burnett, A.M. Aguiar, C.J. Morrow, and C. Philips, J. Amer. Chem. Soc, 93, 1301 (1971). 19. H. Hirai and T. Furuta, J. Polym. Sci. (B) , 9_, 459 (1971); i b i d . , 1,729 (1971). 20. J. Manassen, Plat. Metals Rev., 15_, 142 (1971). 21. L. Vaska and J.W. Diluzio, J. Amer. Chem. Soc, 84, 679 (1962). 22. L. Vaska, Acc Chem. Res., 1> 335 (1968). 23. L. Vaska, Inorg. Chim. Acta, _5> 295 (1971). 24. J. Halpern, Discuss. Faraday Soc, 46_, 7 (1968). 25. J. Halpern, Pure Appl. Chem., 20, 59 (1969). 26. J. Halpern, Acc. Chem. Res., .3. 386 (1970). 27. J.P. Collman, Acc. Chem. Res., 1., 136 (1968). 28. J.P. Collman, Trans. N.Y. Acad. Sci., 30, 479 (1968). 29. J.P. Collman and W.R. Roper, Adv. Organomet. Chem., 54 (1968). 30. S. Carra and R. Ugo, Inorg. Chim. Acta Rev., JL, 49 (1967). 31. R. Cramer, Acc. Chem. Res., 1, 186 (1968). 32. W. Strohmeier, "Problem und Modell der homogenenKatalyse" in C.K. Jorgensen, J.B. Neilands, R.S. Nyholm, D. Reinen, and R.J.P. Williams, Eds., Structure and Bonding (New York), 5^, 96 (1968). 33. G. Wilkinson, Bull. Soc Chim. Fr., 5055 (1968). 34. H.B. Kangan and T.P. Dang, J. Amer. Chem. Soc, 9_4, 6429 (1972); T.P. Dang and H.B. Kagan, Chem. Comm. (London), 481 (1971). 35. L. Vaska and R.E. Rhodes, J. Amer. Chem. Soc, 87^  4970 (1965). 36. G.G. Eberhardt and L. Vaska, J. Catal., 8, 183 (1967). - 12 -37. W. Strohmeier and T. Onoda, Z. Naturforsch., 24b, 461 (1969). 38. W. Strohmeier and R. Fleischmann, Z. Naturforsch, 24b, 1217 (1969). 39. W. Strohmeier and T. Onoda, Z. Naturforsch., 24b, 1493 (1969). 40. W. Strohmeier, W. Rehder-Stirnweiss, and R. Fleischmann, Z. Naturforsch., 25b, 1481 (1970). 41. M. Yamaguchi, J . Chem. Soc. Japan, Ind. Chem. Sect., 70.> 675 (1967). 42. L.H. Vogt, J r . , H.M. Faigenbaum and S.E. Wiberley, Chem. Rev., j>3_, 269 (1963). 43. J.H. Wang, Acc. Chem. Res., _3» 90 (1970). 44. E. Bayer and P. Schretzmann, Struct. Bonding,(New York), 2, 181 (1967) 45. V.J. Choy and C.J. O'Connor, Coord. Chem. Rev., 9_, 145 (1972/73). 46. J.S. Valentine, Chem. Rev., 73., 235 (1973). 47. J . Chett and L.A. Duncanson, J . Chem. S o c , 2939 (1953). 48. M.J.S. Dewar, B u l l . Soc. Chim. Fr., 18, C71 (1951). 49. J.S. G r i f f i t h , Proc. Roy. S o c Ser. (A), 235, 23 (1956). 50. J.A. McGinnety, R.J. Doedens and J.A. Ibers, Inorg. Chem., 2243 (1967). 51. L. Vaska, L.S. Chen and W.V. M i l l e r , J . Amer. Chem. S o c , 93_, 6671 (1971). CHAPTER II APPARATUS AND EXPERIMENTAL PROCEDURE II.1 M a t e r i a l s II.1.1 Iridium Compounds Ir(CO)Cl(PPh^)^ , trans-chlorocarbonylbis(triphenylphosphine)-i r i d i u m ( I ) was obtained from Johnson Matthey and Co. (m.p. 323-325°; vrri = 1950 cm"1; analysis = Calc. : C, 57.0%; H, 3.84%. Found: C, 56.3%; H, 3.90%). The properties of the purchased compound were i d e n t i c a l with those described i n the l i t e r a t u r e (1,2). [ I r C l ( C Q H 1 ) 0 ] _ , the orange-yellow u - d i c h l o r o t e t r a k i s ( c y c l o o c t e n e ) -8 14 2 2 d i - i r i d i u m ( l ) was prepared according to the method of J.L. Herde' and C.V. Senoff (3) by r e f l u x i n g I r C l ^ O ^ O and cyclooctene i n isopropanol and water. Y i e l d -\. 50% (Analysis: C a l c : C, 43.62%; H, 6.24%. Found: C, 42.92%; H, 6.25%). The i . r . spectrum was i d e n t i c a l to that reported (3). The compound was found to be quite a i r - s e n s i t i v e , and thus was stored i n a vacuum desiccator. Iridium t r i c h l o r i d e was obtained as the IrCl^'XH^O from Johnson Matthey and Co. and as the t r i h y d r a t e from Engelhard Industries. The s a l t s were stored over s i l i c a g e l i n a desiccator. - 14 -11.1.2 Organic Ligands and Substrates Maleic a c i d (Eastman Kodak, C P . grade) was r e c r y s t a l l i z e d from water or water/ethanol before use. The p u r i t y was checked by melting point determination. Cinnamic acid (Eastman Organic) and fumaric a c i d (Fisher S c i e n t i f i c ) were of reagent grade and were used as such. Cyclooctene, hexene-1 and d i e t h y l maleate were obtained from K and K Laboratories. These o l e f i n s were p u r i f i e d by passage through a 30 cm alumina column (to remove peroxides) and were used immediately a f t e r p u r i f i c a t i o n . The k i n e t i c r e s u l t s obtained using these p u r i f i e d substrates and the substrates without p u r i f i c a t i o n were i d e n t i c a l . Hence, a l l substrates were generally used without further p u r i f i c a t i o n . 11.1.3 Gases P u r i f i e d hydrogen was obtained from Matheson Company. The hydrogen was passed through a Deoxo c a t a l y t i c p u r i f i e r to remove traces of oxygen before use. Carbon monoxide was obtained as C P . grade from Matheson Co. P u r i f i e d nitrogen and oxygen were from Canadian L i q u i d A i r Co. 11.1.4 Solvents DMA (b.p. 165-166°) was obtained from Fisher S c i e n t i f i c Co. I P u r i f i c a t i o n involved s t i r r i n g over calcium hydride under nitrogen atmosphere f or 40 hours, followed by d i s t i l l a t i o n under a nitrogen atmosphere. The constan t - b o i l i n g f r a c t i o n was c o l l e c t e d d i r e c t l y i n t o Linde 4A molecular sieve and stored under nitrogen. - 15 -Sulfolane (tetrahydrothiophene-l,l-dioxide) was obtained from A l d r i c h Chemical Co. The solvent was p u r i f i e d i n the same manner as described f o r DMA. Nitromethane and py r i d i n e were obtained from Fisher S c i e n t i f i c Co., and DMSO was obtained from Baker Chemical Co. These solvents were p u r i f i e d by d i s t i l l a t i o n and stored over Linde 4A molecular sieve under nitrogen atmosphere. Other solvents - a c e t o n i t r i l e , acetone, DMF and benzene - were obtained from Fisher S c i e n t i f i c Co. ( c e r t i f i e d grade) and were used as such. II.1.5 Other Materials p_-Toluene sulphonic a c i d and triphenylphosphine were A.R. grade obtained from Eastman Kodak. The l a t t e r were r e c r y s t a l l i z e d from benzene-ethanol before use. L i C l was A.R. grade obtained from A l l i e d Chemical Co. A l l other chemicals used were of reagent grade. D i s t i l l e d water was always used when d i l u t i o n s or aqueous s o l u t i o n preparations were necessary. I I . 2 Gas Uptake Measurements II.2.1 Apparatus f o r Gas Uptake Measurements I A constant pressure gas uptake apparatus as shown i n Figure I I - l was used f o r k i n e t i c studies. The pyrex r e a c t i o n vessel (A), which could be cli p p e d to a metal rod shaken by motor ( I ) during the r e a c t i o n , was connected by a s p i r a l glass arrangement with tap (C) to the o i l manometer (D) through tap (H). The o i l manometer which consisted of a c a p i l l a r y U tube Figure I I - l . Apparatus f o r constant gas-uptake measurements. - 17 -f i l l e d with butyl phthalate (a l i q u i d of ne g l i g i b l e vapour 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 measuring burette was i n turn connected through an Edward's high vacuum needle valve (M) to the gas handling part of the apparatus, which consisted of a mercury manometer (F), the gas i n l e t (Y) and connections to the Welch Duo Seal rotary vacuum pump (G). The reaction f l a s k (A) was thermostated i n a s i l i c o n e o i l (Dow Corning 550 f l u i d ) bath (B). I t consisted of a four l i t r e glass beaker insulated by polystyrene foam on a l l sides and enclosed by a wooden box with a c i r c u l a r hole for observing the colour changes of the reaction mixture. The top of the o i l - b a t h was w e l l covered by poly-styrene foam. The gas burette was immersed i n a thermostated water bath made from a perspex rectangular tank. Both thermostat baths were operated using "Jumo" thermo regulators with "mere to mere" relay control c i r c u i t s and heating provided by 25 watt elongated l i g h t bulbs. These together with mechanical s t i r r i n g ensured temperature control to within +0.1°. A v e r t i c a l l y mounted t r a v e l l i n g telescope was used to follow the gas uptake. A lab-chron 1400 timer was used to record the time during the k i n e t i c experiments. 11.2,2 Procedure for a Typical Gas Uptake Experiment Since the iridium complexes i n solution were a i r - s e n s i t i v e , a reaction f l a s k with a side-arm f i t t e d with a hook at the end was used. The iridium complex was weighed into a small glass bucket suspended from the side arm which, after removal of a i r from the system, was rotated - 18 -to drop the bucket into the required amount of solvent. Substrate was added to the above solvent i f required. The r e a c t i o n f l a s k (A) was then connected by the s p i r a l and tap (C) to the gas-handling part of the apparatus at (0). The reactant s o l u t i o n was degassed by a l t e r n a t e cooling with pumping and warming. The reactant gas was admitted at a pressure somewhat les s than that required for the experiment and then taps (C) and (P) were closed. The whole system up to tap (H) was pumped down with taps (K), (L), (J) and (M) open. The f l a s k and s p i r a l arrangement were disconnected from (0) and transferred into the thermostated o i l bath with the s p i r a l connected to the o i l manometer through tap (H). Tap (H) was opened and a f t e r the a i r between tap (H) and (C) was pumped out, tap (Q) was closed. Gas was then admitted to the r e s t of the gas uptake apparatus up to tap (C), which was then opened so that the pressure i n the whole system was equalized. The r e a c t i o n pressure required was adjusted by using the mercury manometer. Tap (J) and the needle valve (M) were closed while the i n i t i a l reading of the mercury l e v e l i n (N) was taken. Taps (K) and (L) were closed and the timer and shaker were s t a r t e d simultaneously. As a r e s u l t of any gas uptake, the o i l l e v e l on the l e f t hand side of the manometer rose and to maintain zero d i f f e r e n c e i n l e v e l s , gas was admitted into the gas measuring burette through tap (J) and needle valve (M) to give a corresponding r i s e of mercury i n (N). The change i n height of the mercury was noted as a function of time. Since the diameter of (N) was known, the corresponding volume of gas used was found and an uptake p l o t of gas consumption i n moles l i t r e 1 against time could be drawn. - 19 -The use of a small volume of s o l u t i o n (y 5 ml), a r e l a t i v e l y large indented vessel (y 30 ml) and a high shaking rate ensured the absence of d i f f u s i o n c o n t r o l i n the rate of gas consumption. II.2.3 Gas S o l u b i l i t y Measurements The s o l u b i l i t y of a gas i n c e r t a i n solvents under s p e c i f i c tempera-ture and pressure conditions could be determined using the gas-uptake apparatus and a rea c t i o n f l a s k containing a stopcock i n i t s neck. The en t i r e system inc l u d i n g the re a c t i o n f l a s k containing a measured amount of solvent was evacuated at room temperature. The tap on the f l a s k was then closed, and the f l a s k was placed i n the o i l bath at the desired temperature. The system was then evacuated to the f l a s k tap and f i l l e d with gas to the approximate pressure desired. The f l a s k tap was then opened and the pressure adjusted immediatley to that required. Taps K and L were closed, the shaker s t a r t e d and the immediate uptake measured as described i n the previous s e c t i o n , allowing c a l c u l a t i o n of the gas s o l u b i l i t y . II.3 I d e n t i f i c a t i o n of Liquids and S o l i d Products D i s t i l l a t i o n under vacuum ( i . e . pumping through a l i q u i d nitrogen co l d trap) was used to separate the solvent plus l i q u i d organic products from the c a t a l y s t and n o n - v o l a t i l e products. The d i s t i l l a t e could then be investigated by gas chromatography using a s u i t a b l e column, or by i . r . or n.m.r. spectroscopy. The residue could also be studied using i . r . or n.m.r. when dissolved i n an appropriate solvent. - 20 -II.4 Instrumentation V i s i b l e and u l t r a v i o l e t absorption spectra were recorded using a Perkin Elmer 202 spectrometer; t h i s could be f i t t e d when necessary with a thermostated c e l l compartment. Matched s i l i c a c e l l s of 1 mm or 1 cm path lengths were used. A c e l l f i t t e d with a micro stopcock, and car r y i n g a q u i c k f i t j o i n t f o r f i t t i n g i n t o a side arm of f l a s k s containing a s o l u t i o n kept under vacuum or a p a r t i c u l a r gaseous atmosphere, was used to take spectra of such s o l u t i o n s . Infrared spectra were recorded on a Perkin-Elmer i n f r a r e d grating spectrophotometer 457 using KBr discs or n u j o l mulls. NaCl, KBr, C s l p l a t e s and l i q u i d c e l l s (NaCl, AgCl) of 0.1 mm or 1 mm path length were also used. E.s.r. spectra were recorded on a Varian Associates E 3 e l e c t r o n s p i n resonance spectrometer. A Beckman GC-2A chromatography u n i t with thermal conductivity type detector, and dinonyl phthalate, carbowax and pOrapak columns were used f o r a n a l y s i s of l i q u i d samples. Conductivity measurements were c a r r i e d out using a Thomas Serfass condu c t i v i t y bridge model RCM 15 B l with dip type conductivity c e l l s of c e l l constant of about 0.1 cm 1 . Melting points were determined on a Superior E l e c t r i c melting point apparatus. - 21 -REFERENCES L. Vaska and J.W. DiLuzio, J . Amer. Chem. S o c , 84_, 679 (1972); 83, 2784 (1961). P.B. Chock and J. Halpern. J . Amer. Chem. S o c , J38, 3511 (1966). J.L. Herde and C.V. Senoff, Inorg. Nucl. Chem. L e t t e r s , 7_, 1029, (1971). CHAPTER III SOLVENT EFFECTS ON HOMOGENEOUS HYDROGENATION CATALYZED BY VASKA'S COMPOUND, Ir(CO)Cl(PPh 3) 2 I I I . l Introduction The simple reversible hydrogen activating system involving the square planar trans-Ir(CO)XP 2 complexes (X = halide, P - P P h 3 ) at 25° and 1 atm hydrogen was reported i n i t i a l l y by Vaska and Diluzio (1), eq. (3.1). Kinetic and equilibria data for such systems have been reported subsequently by a number of groups (2-4,18). The rate law of the forward k^ values for Ir(CO)Cl(PPh 3) 2 increase somewhat with solvent polarity; for example, in a ratio of approximately 1:2:3 for the series toluene < chlorobenzene <C DMF (18). But the activation parameters are l i t t l e affected on using the different solvents and the mechanism is thought to be essentially the same (2,3). The reaction i s pictured as interaction of a Lewis base iridium complex and a Lewis acid, hydrogen, the dihydride being formed via a concerted mechanism involving a 3-centered transition state as shown in eq.(3.2)(the PPh^ ligands, normal to the plane of the paper, are not shown): 1^ H 2Ir(CO)XP 2 (3.1) reaction i s k ^ I r ] ^ ] , where ^  i s the second order rate constant. The - 23 -, „ , C l , H C l H Ir | — ^ Ir. ; "> Ir 1 0 C X H -1 OC H (3.2) The Ir(CO)X(PPh 3)2 complexes do act as c a t a l y s t s in,benzene s o l u t i o n for the hydrogenation of ethylene and acetylene, but are not very e f f e c t i v e (13). James and Memon (4), however, reported a more e f f e c t i v e c a t a l y t i c hydrogenation of maleic a c i d i n the coordinating solvent DMA using Ir(CO)X(PPh 3)2« The following slow d i s s o c i a t i o n e q u i l i b r i u m (absent i n benzene) was postulated f o r the DMA system. k l Ir(CO)ClP„ — I r ( C O ) C l P ( S ) + P (3.3) 2  I -1 I I S = solvent Dihydride formation s t i l l occurred r a p i d l y , however, according to eq. (3.1). The importance of the slow d i s s o c i a t i o n step i n making an a v a i l a b l e coordination s i t e f o r the maleic a c i d was discussed (4). I t should be noted that i n a recent paper, Burnett and coworkers (23) have questioned the conclusions of James and Memon (4), and considered that phosphine d i s s o c i a t i o n occurs a f t e r o l e f i n coordination. The important r o l e of the solvent molecule i n c a t a l y t i c hydrogenation of o l e f i n s was furth e r inv e s t i g a t e d i n the present studies using various solvents, and the r e s u l t s are described i n the following sections. The work described i n t h i s chapter represents a cursory preliminary study to discover what general solvent effect; might be evident, and which might be worthy of more d e t a i l e d i n v e s t i g a t i o n . - 24 -III.2 Properties of Ir(CO)ClP 2 III.2.1 Dissociation of P in Various Solvents The catalytic activity of Ir(CO)ClP 2 in DMA was thought to result from the dissociation of a P ligand. The fourth coordination site was considered to be occupied by a solvent molecule, eq. (3.3), and the existence of this "vacant" solvated site was proposed as essential for the catalytic activity for olefin hydrogenation (4). In the solvents studied in the present work, Ir(CO)ClP 2 gave the. same absorption peaks as those observed in the solid reflectance spectrum at 340 nm, 387 nm and 440 nm (4) (See Figure I I I - l ) . In the strongly coordinating solvents, pyridine, DMSO, DMF and CH^ CN , slow decreases in the i n i t i a l l y measured optical density (O.D.) at these wavelengths were observed under N2, suggesting some slowly established dissociation equilibria. On addition of a large excess of P to the i n i t i a l solution of Ir(C0)ClP 2, the same peaks as those observed for the solid were again recorded, but there was no change of O.D. with respect to time. Hence i t can be concluded that P dissociation occurs in these coordinating solvents, the e value at the strongest absorption peak, 387 nm for the undissociated complex was evaluated by using a large excess of added P. The dissociation constant K for eq. (3.3) was determined i n various solvents by analysis of the spectral data at 387 nm. III.2.la Pyridine The spectrum of Ir(C0)ClP 2 in pyridine at 70° under N 2 slowly changes with time and the O.D. at each of the 3 absorption maxima decreases to an equilibrium position,(Fig. I I I - l ) . The change of absorbance O.D 1.0 0.5 [Ir] = 3.33 x IO - 4 M; [P] = 0.83 M [Ir] = 3.48 x 10~4 M; [P] = 1.43 x 10_3M — ; [Ir] = 3.33 x 10"4 M; [P] = 0 350 400 500 wavelength, nm 600 gure I I I - l . Visible absorption spectra of Ir(C0)ClP 2 in pyridine at 70° under N2, c e l l path = 10 mm. O.D. ON overnight Figure III-2. Variation of optical density of Ir(CO)ClP 2 at 387 nm in pyridine at 70c [Ir] = 3.33 x 10 M; c e l l path = 10 mm. under N 2' - 27 -at 387 nm was followed and is shown in Fig. III-2. In the absence of the added P or with added P up to the ratio of [P]/[Ir] = 4, the f i n a l spectra show a continuum and have no characteristics of the Ir(CO)ClP 2 spectrum (Fig. I I I - l ) . Addition of a very large excess of P into the i n i t i a l solution of lr(CO)ClP 2 ([Ir]=3.33 x 10~ 4 M; [P] = 0.83 M) gave 3 an i n i t i a l spectrum for the undissociated complex for which e =3.7 x 10 although there was a subsequent very slight decrease i n the measured optical density (1.24 to 1.20). It i s reasonable to assume that P dissociation is complete i n dilute Ir(C0)ClP 2 solutions in the absence of added phosphine. The 3 continuum spectrum of Fig. I I I - l gives e^j (387 nm) = (1.1 +0.1) x 10 for the l i k e l y Ir(C0)ClP(py) species. At any particular wavelength, the absorbance of a solution containing species I and II, reaction (3.3), i s given by the expression: where O.D. is the optical density of the solution measured i n a 10 mm optical c e l l , and e^ . and are the molar extinction coefficientyof species I and II, respectively, at the given wavelength. Also, O.D. = e ].[I] + E l I [ I I ] (3.4) [ I r ] T = [I] + [II] (3.5) [ P ] T = [ P ] a + (3.6) where [Ir] and [P] are the total iridium and total free P concentrations, - 28 -respectively, and [P] is the concentration of P added to the solution. cl Combining eq. (3.4) and (3.5) gives the result: O.D. - £ T T [ I r ] T [I] = _ 1 1 l — (3.7) I II Knowing the extinction coefficient of the two species, their concentrations may be calculated from eq. (3.5) and (3.7), which allows K to be calculated using the relationship: [ H ] [ P ] T K = — j j j L (3-8) The dissociation constant K for reaction (3.3) in pyridine, at 70°, was then calculated for various concentrations of added P, and the -3 data are summarized i n Table I I I - l . In the [P] T range below 2 x 10 M, or higher than 0.2 M, the [II]/[I] ratio i s too high or too low respectively, that i s , the Ir(C0)ClP2 i s practically completely dissociated or completely non-dissociated, and hence the K values obtained in these ranges are not very reliable. The three better K values are obtained using a ' T -2 -2 [P] range from 0.1 M to 4 x 10 M, and the average K i s 4.5 x 10 M. III.2.lb DMSO, DMF and CH CN The Ir(C0)ClP2 i s only pa r t i a l l y dissociated i n these solvent systems which were studied in a similar manner to that described above for the pyridine system. The absorbance of Ir(C0)ClP2 in these solvents decreases slowly under ^ to an equilibrium position, and the f i n a l equilibrium spectra are shown in Figure III-3. No change in spectrum - 29 -Table I I I - l . Spectrophotometric study of equilibrium reaction (3.3) in pyridine solution. Dependence of O.D. on triphenyl-phosphine concentration at 70°. [ I r ] T x l 0 4 0.D.a [II]/[I] [P] T K x 10 2 b M M 387 nm M M 3.33 0 0.39 — 3.33x10 3.54 4.8xl0~ 4 0.36 83 8.3xlO~4 ^6.9 3.33 1.24xl0~3 0.38 17 1.57xl0~4 ^2.7 3.48 1.43x10"3 0.43 10 1.75xl0~3 VL.4 3.58 4.1xl0~ 2 0.83 0.97 4.1xl0~ 2 4.0 3.54 9.73xl0 _ 2 0.95 0.55 9.73xl0"2 5.3 3.41 0.109 0.98 0.39 0.109 4.2 3.33 0.83 1.20 3.7xl0 - 2 0.83 ^3.2 c e l l path i s 10 mm calculated according to eq. (3.8). O.D. 1 — 1 1 i _ 350 400 5 0 0 6 0 0 wavelength, nm Figure III-3. Visible absorption spectra of Ir(C0)ClP o in various solvents under N_ (c e l l path = 10 mm). - 31 -was observed on addition of excess P to the i n i t i a l solutions of Ir(CO)ClP 2 > and the molar extinction coefficient e (387 nm) for Ir(C0)ClP o was calculated as 3.7 x 10 3 in DMSO and 3.2 x 10 3 in CH„CN. The £j.(387 nm) values can also be obtained from extrapolation of the O.D. vs. time data for solutions containing no added phosphine; the data may also be used to give information about the kinetics of reaction (3.3), although these have not been studied in any detail. Figure III-4 shows such an optical density change for the complex in DMF; analysis of the data is a standard procedure involving the relationship (4) D 2-D D k, (D +D ) t l o g - ° L _ § — = 1 ° e • ( 3 a ) i 0 g (D-D )D 2.3(D-D ) t e o o e where D t is the optical density at time t and subscripts o and e refer to i n i t i a l and f i n a l equilibrium values. A plot of the log term against time i s also shown In Fig. III-4 and is reasonably linear. The rate constant k^ for the P dissociation reaction (3.3) was evaluated as 1.0 x 10 _1 o o sec at 70°. The ei(387nm) in DMF was evaluated as about 3.J x 10 . The e^j(387 nm) values for Ir(C0)ClP(S) are not obtained experimentally -2 in these solvents, and thus i n order to estimate K values, the assumption 3 has been made that e T T(387 nm) i s ^10 , since the value i n pyridine is 3 known to be (1.1+0.1) x 10 . This is a very approximate procedure but at least i t w i l l give K values within an order of magnitude. These equilibrium constants K for reaction (3.3) were then evaluated according to eq. (3.8), and are summarized in Table III-2. O.D. log D -D D o t e D (D -D ) o t e 1.2 0.8 •A H 0.4 -3 time x 10 ,sec 10 10 ,-4 Figure III-4. Optical density plot at 387 nm for P dissociation of Ir(C0)ClP 2 (3.46 x 10 M) in DMF at 70° under N2; 10 mm c e l l ; A , O.D.; O . corresponding log plot. Table III-2. Spectrophotometric study of equilibrium reaction (3.3) i n DMSO, DMF and CH,CN solvents. Solvent Temp. [Ir] x 10 4 [P] 1 Jadded O.D.a [II]/[I] [P] T K e I M M 387 nm M M 387 nm DMSO 35° 3.78 0 1.32 8.0 x IO - 2 2.8 x IO - 5 ^2.3 X i o " 6 35° 3.36 0 1.20 5.0 x IO - 2 1.6 x IO - 5 ^8.0 X i o " 7 3 5 o 2.46 0 0.89 3.4 x 1 0 - 2 8 x IO - 6 ^2.7 X IO" 7 35° 3.51 5.02 x IO - 3 1.30 1 x 1 0 - 4 5.0 x IO - 3 - 3.7 x 1 0 3 50° 2.44 0 0.63 0.7 1.0 x IO - 4 7.2 X IO" 5 80° 3.84 0 1.19 0.27 8.2 x IO - 5 2.2 X IO" 5 80° 3.84 0 1.12 0.41. 1.1 x IO - 4 4.5 X IO" 5 DMF k 35° 3.20 0 0.90 0.49 1.05 x IO - 4 5.1 X IO"5 70° 3.46 0 0.94 0.57 1.26 x IO - 4 7.2 X IO" 5 CH3CN 20° 3.48 0 0.75 0.90 1.65 x IO - 4 1.5 X IO" 4 20° 3.51 • 4.1 x 10" 2 1.12 3.6 x 1 0 - 3 4.1 x 10~ 2 - 3.2 x 1 0 3 10 mm c e l l path. Data may be unreliable (see text). - 34 -It should be noted that i n contrast to the data reported here f o r the DMF system, and since these studies were completed, Burnett and coworkers (23) have very recently reported no s i g n i f i c a n t changes i n the absorption spectra of DMF s o l u t i o n of Ir(C0)ClP2. These findings cast some doubt on the present r e s u l t s , and the discrepancy may be due to traces of peroxides i n the DMF that we used (see below), although the solvent v/as f r e s h l y d i s t i l l e d . The data of Burnett's group implies that equilibrium (3.3) l i e s w e l l to the l e f t (K < 1 0 _ 7 M - 1 ) . III.2.1c DMA James and Memon (4) had observed changes i n absorbance for DMA solutions of Ir(C0)ClP2 under nitrogen. The changes were small but were analyzed as o u t l i n e d above f o r the DMF system (III.2.2b). Burnett and coworkers (23) have also recently reported the studies i n DMA and reported no change over long periods i n the o p t i c a l system of DMA solutions of Ir(C0)ClP2 under vacuum; they suggested that the discrepancies might be a t t r i b u t e d to leakage of a i r i n t o the o p t i c a l c e l l i n the e a r l i e r work. We have repeated these studies and indeed the absorption spectra of such solutions do remain e f f e c t i v e l y constant under vacuum or nitrogen. Very samll decreases i n o p t i c a l density were measured at -4 387 nm (e.g. the O.D. of a 4.10 x 10 M DMA s o l u t i o n of I r ( C 0 ) C l P 2 changed from 1.17 to 1.15 over 24 hrs., see Table III-3) but the changes are within experimental error and are probably not s i g n i f i c a n t . However, i n the presence of excess phosphine these very small decreases i n absorbance were not observed, such small changes would imply a K value for phosphine d i s s o c i a t i o n of <v 10 ^ M. - 35 -The erroneous measurements of James and Memon (4) appear to be due to the presence of traces of peroxides in the DMA solvent, since unless this solvent was d i s t i l l e d immediately prior to use, significant slow decreases in absorbance were observed. Table III-3. Spectrophotometric data in DMA (freshly d i s t i l l e d ) solvent m vacuum at 35 . [Ir]xl0 4 [P] 1 Jadded O.D.a K° E I M M 387 nm M 387 nm 3.97 0 1.08 M. X i o " 5 3.84 1.15 x I O - 2 1.1 2.9 x 10 3 3.68 0 1.05 ^2 X i o " 6 3.68 1.01 x IO - 2 1.07 3.0 x 10 3 4.10b 0 1.15 0.4 X i o " 6 a 10 mm c e l l path b under 1 atm. ° for equilibrium (3.3); assuming ^ (387 nm) = 3.0 x 10 e T T(387 nm) = 10 3 - 36 - ' III.2.Id Acetone and Sulfolane In acetone and in sulfolane, no change of absorbance was observed for the Ir(CO)ClP 2 solution with or without added P. Hence i t can be concluded that no P dissociation occurs; Cj(387 nm) values for < 3 3 Ir(CO)ClP 2 were calculated to be 3.4 x 10 i n acetone and 3.7 x 10 i n sulfolane. The relevant data are summarized in Table III-4. Further details about the sulfolane solvent system are given i n Chapter IV. III.2.1e CH3N02 Under a N 2 atmosphere, the spectrum of Ir(C0)ClP 2 decreases rapidly i n intensity and reaches an equilibrium spectrum shown in Fig. III-3. The spectrum shows an absorption maximum at 387 nm and a small shoulder at 440 nm, but no maximum at 340 nm. The spectral changes of the Ir(C0)ClP 2 solution were found to be affected by both addition of P and L i C l . The variation of O.D. at 387 nm with time in the presence or absence of added P or Cl i s shown i n Fig. III-5. The results indicate that Cl as well as P i s being dissociated i n this solvent. Studies of the separate dissociations were not pursued. The absorbance at 387 nm at time = 0 was obtained by extrapolation of the data of Fig. III-5, and 3 e^(387 nm) was evaluated as about 3.0 x 10 . III.2.2 Reaction with Molecular ^-Formation of the Dihydride Complex, Reaction (3.1) It i s known (1,2) that the square planar complex Ir(C0)ClP 2 reacts reversibly with molecular H 2 to form octahedral Ir(III) dihydride complexes in which the hydrogen is^present as anionic hydride ligands. The activation parameters for the forward reaction are l i t t l e affected by - 37 -Table III-4. Spectrophotometric study of equilibrium reaction (3.3) in acetone, sulfolane. Solvent Temp. [Ir] x 10 4 [P] , O.D.b e„ added I M M 387 nm 387 nm Acetone 20° 3.59 0 1.20 3.3 x 10 3 20° 3.36 0.102 1.12 3.3 x 10 3 Sulfolane 35° 3.38 0 1.23 3.7 x 10 3 a 3 See Section IV. 4.1 b 10 mm c e l l path. O.D. 1.0 0.5 P - o A oo 10 _ 3 time x 10 , 15 50 sec Figure III-5. Optical density plots for Ir(C0)ClP 2 in CH3N02 under N 2 at 25°; 10 mm c e l l path; (O) [Ir] - 3.46 x 10"4 M; ( A ) [Ir] = 3.44 x 10 _ 4 M, [Cl -] = 4.33 x 10~ 2 M; (O) [Ir] = 3.59 x 10~4 M, [P] » 0.406 M. - 39 -using different solvents, like benzene, toluene, chlorobenzene and DMF (2,3,18). In the present work, the dihydride formation reaction (3.1) was investigated in a number of other solvents. At 35°, and 1 atm pressure, the yellow color of Ir(C0)ClP2 in various solvents fades quickly; i n each case the spectrum of the dihydride solution shows a continuum with l i t t l e absorption at 387 nm compared with the Ir(C0)ClP2 complex. The reaction (3.1) appears to go to completion i n a l l the solvents under these conditions. The data giving the molar extinction coefficient for the dihydride at 387 nm in different solvents are summarized i n Table III-5. Lower limits can be placed on the equilibrium constants (*H) for dihydride formation. For example, i f less than 10% of the iridium remains as Ir(C0)ClP2 after hydrogenation, thenKjjt^] > 10. Since the s o l u b i l i t y of H 2 in a l l these solvents is ^ (2-3) x 10~3 M (14), 1^ must be > 3 x 10 3 M _ 1. The value for the sulfolane system has been measured at higher temperatures 3 —1 and, for example, at 80° was found to be 2.2 x 10 M (see Section IV. 4.2). III.2.3 Formation of an Olefinic Complex with Maleic Acid (MA) Reactions that might be anticipated are shown in equations (3.10) and (3.11) K Ir(C0)ClP 2 + MA m. Ir(C0)ClP2(MA) f Ir(C0)ClP 9 Ir(C0)ClP(S) + P K 9 < Ir(C0)ClP(S) + MA — ^ Ir(C0)ClP(MA) + S II ^ III (3.10) (3.11) - 40 -Table III-5. Spectrophotometric study of dihydride formation, reaction (3.1), in different solvents. T = 35°C, p[H 2] ~ 760 mm Hg. Solvent [Ir] x 10 4 0.D.a E(Ir(C0)C1P 2(H 2)) M 387 nm 387 nm pyridine 3.62 0.13 360 DMSO 3.23 0.10 310 DMA 500b DMF 3.23 0.08 250 CH3CN 3.41 0.06 180 acetone 3.90 0.13 330 sulfolane 3.49 0.07 200 CH3N02 3.59 0.05 140 10 mm c e l l path from ref. 4. - 41 -Ethylene reacts with benzene solutions of Ir(C0)ClP2 according to eq. (3.10) (13), while MA was postulated to react with DMA solutions of the iridium complex according to eq. (3.11) (4). Burnett and coworkers (23) have indicated that processes such as (3.12) are important. K 3 Ir(C0)ClP 2 + MA m - Ir(CO)ClP(MA) + P (3.12) III.2.3a Pyridine Spectral changes observed for pyridine solutions of Ir(C0)ClP2 containing MA at 70° were identical with those observed i n the absence of the maleic acid (see Section III.2.1a). This was so i n either the absence or presence of added phosphine; the observed f i n a l optical densities were the same as those calculated for the phosphine dissociation reaction only (Table III-6). Hence i t can be concluded that reactions (3.10) and (3.12) do not occur, or else occur to a very small extent under the conditions noted. III.2.3b DMA This system has been studied previously i n this laboratory (4), but was brie f l y reinvestigated for purposes of comparison with other solvent systems studied in the present work. In the presence of MA, the decrease in O.D. of Ir(C0)ClP2 i n DMA i s significant and much greater than in a corresponding reaction without the olefin, showing that some complexing i s occurring (Table III-6). The equilibrated spectrum at the conditions noted showed a continuum below 450 nm with relatively l i t t l e absorption at 387 nm (e = 320). Addition of excess P completely inhibits complex formation, indicating that maleic acid forms a complex Table III-6. Spectrophotometric study of olefinic complex formation in different solvents. (10 mm c e l l path). Solvent Temp. [Ir] x 104 [MA] [H]/[I] 3 O.D.b O.D. olefin complexing M M M 387 nm calculated 387 nm observed constants pyridine 70° 3.46 0.223 4.99 x 10~2 0.202 1.12 1.11 K small m DMA 35° 35° 4.03 3.74 0 0.195 5.18 x 10 - 2 8.2 x 10"2 6.6 1.2 x IO - 2 1.10 0.13 1.10 K 3 > 0.02 m DMSO 35° 3.49 0 1.66 x IO - 2 5.2 x 10"2 1.24 1.23 K small m DMF 3 5 o 3.08 0 0.104 0.50 0.87 0.90 K small m acetone 35° 3.90 0 4.2 x 10 - 2 — _ 0.75 Kp = 2.2 x 10~3 35° 3.58 0 0.114 - - 0.39 Kp = 3.4 x 10~3 35° 3.58 0.11 0.114 - - 1.05 K 1 = 1.1 M"1 m sulfolane 35° K 1 = 15 M _ 1 C m CH3CN 35° 3.23 0 0.102 - - 0.07 K large m CH3N02 35° 3.34 0 0.110 - . - 0.03 -3 K calculated according to eq. (3.8) for phosphine dissociation only; O.D. (387 nm) = e T[I]+e T T[II] for reaction (3.3); c See Section IV.4.3a. 1 1 1 - 43 " with species according to either reaction (3.11) or (3.12), rather than-the original Ir(CO)ClP 2, reaction (3.10) (4). The reaction (3.11) or 3 (3.12) essentially goes to completion and the data indicate that (or KK 2) > 0.02 at 35°. m III.2.3c DMSO and DMF In the presence of maleic acid, the O.D. decrease of Ir(C0)ClP 2 i n the solvents, DMSO and DMF i s very close to that for the corresponding phosphine dissociation reaction without added olefin, indicating that the extent of olef i n complexing i s very small. Thus reaction (3.11) or (3.12) in these solvents i s not observable. The results are summarized i n Table III-6. III.2.3d Acetone On addition of MA into a solution of Ir(C0)ClP 2, the O.D. decreases and indicates that complexing i s occurring. Since there i s no P dissociation in this solvent i n the absence of added olefin, the olefin complexing reaction could be reaction (3.10). On addition of large excess of P into the Ir(C0)C1P2~MA solution, the O.D. increases and this could be consistent with subsequent P dissociation from the ol e f i n complex Ir(CO)ClP 2(MA), i.e. K 1 Ir(C0)ClP o + MA m ^ Ir(CO)ClP0(MA) (3.10) I IV K Ir(CO)ClP„(MA) —2-=^  Ir(CO)CIP(MA) + P (3.13) 2 ^ IV III - 44 -An alternative qualitative explanation of the observations l i e s in an equilibrium such as reaction (3.12). However analysis of the data gives 3 -1 -3 variable K values between %10 and ^10 , and this rules out such an m equilibrium. The value at 35° of K ^ for reaction (3.10) was thus m evaluated as 1.1 M 1 from the O . D . data of Ir(C0)C1P2-MA solution with the presence of large excess P. The P dissociation constant of reaction (3.13) was calculated using the following relationship: [Ir] = [I] + [IV] + [III] (3.14) where [ I r ] T i s the total Ir-concentration. [Ir]-, = [I] + K ^"[1] [MA] + J K K 1 [I] [MA] i m \ p m ([Ir] - [I] - K 1[I][MA]) 2 or K = ^— m (3.15) P K [I] [MA] Since the ratio of O.D. at 387 nm to that at 440 nm was found to be the same for solutions of Ir(C0)ClP_, r 0 - D - ( 3 8 7 n m ) =  1'2Z = 3 R i , 2* lO.D.(440 nm) 0.32 J ' H J » Ir(C0)ClP 9 +MA,[°-°-< 3f7 n m> = £ i i = 3 9 ] . 2 'L0.D.(440 nm) 0.10 and Ir(C0)ClP 2 + P . f f g l g g g - i f f - 3-8); the extinction coefficients at 387 nm for the species Ir(CO)ClP(MA) and Ir(C0)ClP2(MA) are probably close to zero; with these assumptions, values for reaction (3.13) were evaluated as 2-3 x 10 M which are considered reasonably consistent. The data are summarized in Table III-6. - 45 -III. 2.3e Sulfolane The complex-formation reaction (3.10) occurs and this i s then followed by Cl dissociation. The system is f u l l y discussed in Chapter IV, where the K value for reaction (3.10) is estimated to be 15 M ^ m at 35°. III. 2. 3f ^3CN_And_CH3N02 In the presence of MA, the yellow color of Ir(C0)ClP 2 i n CR^CN or CH.jN02 fades quickly and the spectra show a continuum below 400 nm, and show no characteristic maxima of Ir(C0)ClP 2. Hence any o l e f i n i c complex formation constants must be relatively large i n both these solvent systems. In CH^CN, on addition of large excess of P into the i n i t i a l solution of Ir(C0)ClP 2 and MA, the spectrum changes and shows now a broad peak at 360-370 nm with a less intense shoulder at 440 nm. Some further complex must be formed i n the solution, see Fig. III-6. In CH3N02 solution,both P and C l ~ dissociate from the Ir(C0)ClP 2 complex (section III.2.1e), and hence further details about olefin complex formation were not pursued. I - 47 -III.3 Catalytic Hydrogenation of Maleic Acid using Ir(CO)ClP 2 It has been shown in the previous sections that molecular hydrogen and maleic acid form complexes with Ir(CO)ClP 2 i n various solvents. Catalytic hydrogenation of maleic acid by molecular H 2 u s i n g Ir(CO)ClP 2 as catalyst had been reported earlier in DMA solvent (4); detailed kinetics were studied and a mechanism was proposed. Hence i t seemed like l y that in the other solvents, the molecular H 2 and maleic acid might be activated for similar hydrogenation reactions, and a cursory survey of such catalytic activity was thus made i n these other solvents. Since this work was completed Burnett and coworkers (23) have reported on the catalytic hydrogenation system i n DMF and i n DMA. The H 2 uptakes of Ir(C0)C1P2-MA systems i n various solvents were measured as described i n Section II.2; some linear rates of H 2 uptake, under the conditions noted, are summaried i n Table III-7, and the H 2 uptake plots are shown in Fig. I l l — 7 . No H 2 uptake was observed in the solvent systems of pyridine, acetone, CH^CN and CH^N02. A large amount of gas evolution was observed i n the formamide solvent system. The evolved gases were not identified but plausibly, formamide is being catalytically decomposed into NH^  and CO (19) (see Discussion). For the DMA and DMSO solvent systems, the H2-uptake plots are very similar i n shape (Fig. III-7). A slow i n i t i a l autocatalytic period i s - 48 -Table III-7. Catalytic hydrogenation of MA using Ir(CO)ClP 2 in different solvents. Solvent Temp. °C [Ir]xl0 3 M [MA] M P[H 2] mm Hg Hydrogenation rate -1 M sec Relative activity pyridine 60 3.20 0.17 660 0 0 DMSO 80 5.36 0.12 750 2.0xl0" 6 a 1 DMA 80 5.36 0.07 730 3.7xl0" 6 b 2 DMF 80 5.92 0.108 730 ^10~ 7 <0.1 formamide 80 4.20 0.106 700 0 c 0 acetone 30 2.66 0.14 500 0 0 sulfolane 80 5.36 0.12 760 1.4xl0~ 5 d 7 CH3CN 50 3.92 0.12 500 0 0 CH3N02 70 5.36 0.14 500 0 0 a for details, see Table III-8. b from ref. 4. c large amount of gas evolved ^ for details, see Chapter IV. absorbed M 0.12 0.08 0.04 sulfolane (80°) {DMA. (80°) DMF (80°) CH3N02, CH3CN, acetone, pyridine Figure III-7, 20 60 AO time x 10"J, sec H 2 uptake plot of Ir(C0)ClP 2 (5.36 x IO""3 M) and MA (0.12 M) i n different solvents under 1 atm H„. - 50 -followed by a long linear region, and f i n a l l y the rate f a l l s to zero at the complete hydrogenation of MA. The slope of the linear region was taken as the rate of hydrogenation of MA. In DMSO, the variation of hydrogenation rate was studied with respect to concentrations of Ir(C0)ClP2', MA, and H^ , and the data are summarized in Table III-8, and in Fig. III-8. The kinetic dependences in DMSO were found to be the same as those reported earlier in DMA (4), and a similar mechanism seems l i k e l y (see Section III-4). The rate of hydrogenation of MA in both systems was first-order in each of Ir(C0)ClP2» and MA at lower concentrations, but became less than one and approached zero order at higher concentrations. Addition of excess P into the DMA and DMSO systems stopped the hydrogenation completely. In DMF, very slow gas uptake does take place, the linear rate (Fig. III-7) being about one-tenth that for corresponding conditions i n DMSO. The kinetics were not investigated, although the hydrogenation was completely inhibited on addition of P ([P]:[Ir] = 8). In sulfolane, an unexpected and higher activity was observed, and the kinetic dependences for the linear uptake region (See Chapter IV) were somewhat different to those of the DMA and DMSO systems. The hydrogenation rate of MA in sulfolane is first-order at lower concentrations, and becomes zero order at higher concentrations i n both catalyst and MA; and i the rate i s nearly independent of the pressure. Rate enhancement by added H+, and inhibition by added Cl in this solvent suggested that ionic dissociation was involved. The detailed investigation of the sulfolane system i s described in Chapter IV. - 51 -Table III-8. Catalytic hydrogenation of MA using Ir(C0)ClP o in DMSO at 80°. [Ir] x 10 3 [MA] P[H 2] Linear hydrogenation M M mm Hg rate, x 10^ M sec 1 5.36 2.47 750 14 5.36 1.36 750 11 5.36 0. 78 750 8.1 5.36 0.50 750 5.9 5.36 0.31 750 4.3 5.36 0.25 750 3.9 5.36 0.21 750 3.0 5.36 0.13 750 2.1 5.36 0.12 750 2.0 5.36 0.093 750 1.7 5.36 0.070 750 1.2 5.36 0.012 750 0.21 5.36 0 750 0 5.36 0.070 750 1.2 3.31 0. 070 750 0.99 1.77 0.070 750 0.74 0.82 0.070 750 0.55 0 0.070 750 0 5.36 0.070 745 1.2 5.36 0.070 555 1.1 5.36 0.070 365 1.0 5.36 0.070 181 0.94 5.36 0.070 107 0.87 5.36 0.070 39 0.65 5.36 0.070 o a o a 5.36 0.070b 750 0 under N„ atm with added [P] = 1.5 x 10 M rate x 10 M sec for rate x 10 M sec 0 200 400 600 800 M Hg . I ^ ] 0 1 2 3 4 M x 10 3, [Ir] 0 0.5 1.0 1.5 2.0 M, [MA] Figure III-8. Dependences of hydrbgenation rate on [Ir(C0)ClP 2], [MA] and [H2] in DMSO at 80°. O : [Ir] = 5.36 x 10"3 M, [MA] - 0-070 M; A : [MA] - 0.070 M , p[H 2] = 750 mm Hg O : [Ir] - 5.36 x 10~3 M , ' p[H ?] - 750 mm Hg. - 53 -III. 4 Discussion The catalytic hydrogenation of maleic acid in DMA solution using Ir(CO)ClP 2 has been studied previously (4) and the mechanism originally proposed i s shown i n the following scheme Ir(CO)ClP 2 + H 2 Ir(CO)ClP 2(H 2) -1 K P + Ir(CO)ClP + MA K m Ir(CO)ClP.(MA) + H 2 Ir(CO)ClP + succinic acid Such an unsaturate mechanism leads to the rate law (4) rate = k 2K*[H 2][MA][Ir T] 1 + r[MA] + [PPh3] x (1+KH[H2])/K (3.16) which is consistent with the observed kinetics in DMA and in the DMSO system studied in the present work. In the gas uptake plots for the DMA and DMSO systems (Fig. III-7),,the slow i n i t i a l autocatalytic period may be attributed to a dissociation of P to produce the active catalyst Ir(CO)C1P. The vacant coordination site would presumably be occupied by a solvent molecule; the o l e f i n i c complex i s then formed subsequently from the solvated species. In benzene, where phosphine dissociation does not occur, catalytic hydrogenation of olef i n i s - 54 -extremely slow (4). The recent paper by Burnett and coworkers (23) modifies the above mechanism in that formation of the olefin complex is thought to occur prior to phosphine dissociation. Ir(CO)ClP„ + olefin Ir (C0)ClP o (olefin) (3.16) 2 •*= z Ir (CO) C1P2 (olefin) Ir (CO) C1P (olefin) + P (3.17) The same kinetic dependences result, and thus the kinetics are consistent with either mechanism. The mechanism of Burnett implies no i n i t i a l autocatalytic period, although such 'induction' periods were observed. However, these workers found no evidence for the slow solvolysis reaction (3.3), and hence did not invoke such a reaction. The induction periods were attributed to traces of oxidized impurities which give rise to free radical processes; the nature of these processes was not elucidated. Burnett and coworkers (23) also suggested that the rate determining reaction with hydrogen (k 2 step) might involve radical intermediates; a step which would be strongly solvent dependent. It seems clear that the solvent does play an important role i n the catalytic a c t i v i t i e s of the Ir(CO)ClP 2 systems. Just considering solvolysis of the complex, which undoubtedly occurs in some solvents ( i f not in the amides), when the Ir(CO)ClP 2 complex is dissolved, the solvent molecule can compete with the P group in terms of coordination to the metal. If the donor power of the solvent i s too weak, the P is not expected to be displaced; i f the donor power of the solvents is too strong, the solvent could form a stable Ir(CO)ClP(S) complex with - 55 -the catalyst, and block coordination of the incoming substrate group, such as maleic acid; olefin activation i s thus prevented. In these two extreme cases, l i t t l e catalytic activity would be anticipated. In order 'to make the catalyst system effective, the solvent with the "right" donor power must be chosen. Clearly similar arguments may be presented for process (3.17) which i s l i k e l y to be solvent dependent. Gutmann (5) has studied the donor properties of a series of 0-and N- containing solvents and has defined the donor number, DN, as the numerical quantity -AIL for the enthalpy change in the reaction SbCl 5 D,.. . J N + SbCl c . A S — * D-SbClc . ,v (3.18) (dissolved) 5 (dissolved) — 5 (dissolved) where D = donor solvent Gutmann found that the AG values determined for the same equilibria have a linear relationship with the AH values. The donor number therefore approximates to a molecular property of the solvent, which can easily be determined by experiment. The DN expresses the total amount of interaction with an acceptor molecule, including the contributions both by dipole-dipole or dipole-ion interactions and by the binding effect caused by the a v a i l a b i l i t y of the electron pair. To some extent, even certain steric properties of the solvent molecules may be contained I i n the value of the DN (5). In the present studies, the results showed that Ir(C0)ClP 2 reacts rapidly with 1 atm. molecular H 2 at 35° to form a dihydride complex essentially completely in a l l the different solvents. The dihydride i s almost certainly the 6-coordinate Ir(CO)ClP 2H 2 species. The values at - 56 -35° for formation of the dihydride in pyridine, DMSO, DMA, DMF, CH3CN, acetone, sulfolane and CH3N02 are a l l > 3 x 10 M . Values at 30° reported previously in benzene, toluene, and chlorobenzene were in the 3 -1 range (6-32) x 10 M (20). Earlier work has indicated that the 6-coordinate dihydride i s not involved i n the catalytic hydrogenation (4). Catalytic hydrogenation l i k e l y requires activation of both and the olefin (MA) and thus a further coordination site i s required; this can be accomplished by dissociation of a P ligand at some stage. The equilibrium constant K of reaction (3.3) for P dissociation, the o l e f i n complex formation constants for reaction (3.10)-(3.12), the relative catalytic activity for hydrogenation of MA, together with the donor number and dielectric constant of the different solvents are summarized in Table III-9. The importance of donor strength i s perhaps illustrated i n this table. The solvents are arranged i n the order of decreasing donating power according to Gutmann's donor number. Catalytic hydrogenation of MA was observed i n the solvents DMSO, DMA and DMF, whose donor numbers range from ^ 27 to ^ 30. The P dissociation constant K i n these -6 -5 * solvents i s very small but appears to be in the range of 10 to 10 M. It should be noted that previous workers (4) in this laboratory have reported that ej.(387 nm) = 3.8 x 103 for Ir(C0)ClP 2 in DMA. On assuming that en(387 nm) = 0 for the species Ir(C0)ClP(S), the P dissociation constant K of reaction (3.3) was evaluated as 2.6 x 10 - D and 1.3 x 10-6 M at 25° and 80° respectively, in DMA (4). However, these data were based on spectral changes which now appear to be erroneous. In the present study, using ei(387 nm) = 3.0 x 10 3 for Ir(C0)ClP 2, and on assuming en(387 nm) = 10 3 for the species Ir(CO)CIP(S), the P dissociation constant K of reaction (3.3) is s t i l l * 10~5-10-6 M at 35° in DMA,usingen(387 nm) = 0 yields a K value of 10~" M at 35°. The difference in £j(387 nm) values in the two studies may arise from using different batches of DMA, although the same solvent purification procedure was used in each case. - 57 -Table III-9. Catalytic hydrogenation of MA using Ir(CO)ClP 2 and related properties i n different solvents Solvent Donor Dielectric number constant K° M' Olefin complexing constants Relative catalytic a c t i v i t y ^ pyridine 33.1 12.3 4.5x10 (70°) -2 K (small) m 0 DMSO 29.8 45.0 3x10 (80°) K (small) m DMA DMF 27.8 26.6 38.9 36.1 VL0 u (35°) 7x10 (70°) -5 K > 0.02 m K (small) m < 0.1 formamide -acetone 17.0 sulfolane 14.8 CH3CN 14.1 100 20.7 42.0 38.0 0 0 1.5x10 (20°) -4 K ^ l . l M 1 m K ^ l S M - 1 m K (large) m 0^  0 7 0 CH3N02 2.7 35.9 Equilibrium constant for reaction (3.3). Equilibrium constant for reaction (3.10)-(3.12). Large amount of gas evolved. For MA hydrogenation at 80c - 58 -The same kinetic dependences observed for MA hydrogenation i n the DMSO and DMA systems suggest that similar unsaturate mechanisms outlined at the beginning of this section are involved. The hydrogenation rate measured i n DMF seems surprisingly slow considering the solvent characteristics are very similar to that of DMA. However, Venanzi (19) has noted that Ir(CO)ClP 2 catalyzes the decomposition of DMF into carbon monoxide and dimethylamine: HCOKfcH3)2 > CO + HN(CH 3) 2 (3.19) No details were reported. If this reaction occurs during the hydrogena-tion conditions, then the gas evolution could "mask" any-H2 uptake that might be occurring; the very slow uptake rate measured could result from a contribution of the decomposition reaction (3.19). The gas evolution measured during attempted hydrogenation i n formamide solvent i s almost certainly due to a corresponding decomposition to CO and NH.^. These catalytic decompositions have not been studied to any great extent, but CO loss from a 5-coordinate intermediate Ir(CO) 2ClP 2 complex (21) seems a distinct p o s s i b i l i t y . Catalytic decomposition of DMA i s negligible under the condition of hydrogenation since the gas-uptake for MA hydrogenation corresponds to a stoichiometric reduction of the double bond (4). The lack of hydrogenation in pyridine may be attributed to the strong donor strength of the solvent. The phosphine i s readily dissociated (K *\» 0.045 at 70°), but the maleic acid i s unable to displace the pyridine and does not coordinate to the Ir(CO)C1P(pyridine) 'complex. - 59 -No P dissociation occurs from Ir(C0)ClP2 in the weakly coordinating acetone solvent, although an ole f i n i c complex Ir(C0)ClP2 (olefin) i s formed and this dissociates a P ligand; the limited equilibria data indicate that under the hydrogenation conditions (Table III-7) at least some (<10%) of the iridium w i l l be present as the Ir(CO)CIP(olefin) complex. The lack of hydrogenation in acetone certainly indicates that the 5-coordinate olefin complex does not readily add hydrogen, although some reactivity via the 4-coordinate olefin complex would be anticipated. Considering the observations for the acetone system, the pyridine system could contain a 5-coordinate Ir(C0)ClP(py)2 complex that is unreactive toward MA. 5-coordinate species Ir(C0)ClP2(olefin) containing activated olefins have been isolated (22), and some cyanoethylene complexes do not dissociate P in toluene (22). In the most active sulfolane solvent system, the measured kinetic dependences di f f e r from those measured i n the DMA and DMSO solvent systems. No P dissociation from Ir(C0)ClP2 i s observed in sulfolane. On addition of MA, olefin complex formation according to reaction (3.10), i s followed by a Cl ligand dissociation, and hence this creates the necessary vacant coordination site for the incoming molecule. The kinetics and mechanism for the hydrogenation of MA in sulfolane are fu l l y discussed in Chapter IV. Relatively l i t t l e has been reported on the solvent properties of sulfolane. The donor strength of sulfolane toward the polarizable acid i I^ has been reported (10) to be the same order of magnitude as that of the weak donor benzene. Langford and Langford (12) have reported that sulfolane acts as a bidentate ligand in metal complexes, supposedly by coordination through both 0-atoms, and an unstable Co(sulfolane)„(CIO,), i 4 , complex was isolated. There i s no obvious relationship between catalytic activity and dielectric constant in the active solvent systems. - 60 -According to Gutmann1s donor numbers, the donor strength of CH^ CN is relatively small and based on the data for the acetone and sulfolane systems, P dissociation from Ir(C0)ClP2 in CH^ CN i s perhaps surprising. The observed value for the dissociation constant, however, l i e s between that measured in pyridine and in DMSO. We must bear in mind that the donor number i s the -AH value for the reaction of a solvent molecule with SbCl,., a non-transition metal complex with f u l l y f i l l e d inner d-orbitals. Hence contributions from any ir-bonding with the acetonitrile cannot be involved. In the Ir(C0)ClP2 complex, low lying d-orbitals are available at the Ir-atom, and hence the type of bonding in CH3CN-SbCl5 and i n CH3CN-Ir(C0)C1P2 may be completely different. The donor number used by Gutmann may not be a good criterion for the donor strength of some solvents toward a transition metal complex. +2 8 Drago et a l . (6) have used the Dq values of six-coordinate Ni (d ) complexes as one criterion for the donor strength of solvents. Some +2 of the results are l i s t e d i n Table 111-10. In the Ni(CH„CN), complex, j o the large Dq value indicates there i s considerable metal to ligand ir-bonding, and the donor strength of CH3CN i s considerably greater than that reported by Gutmann. The P dissociation constant K in CH3CN for reaction (3.3) l i e s somewhere between that i n pyridine and i n DMSO, and is consistent with the donor strength reported by Drago et a l . Graham (7) has also investigated the a-donating and iT-accepting properties of a series of ligands, including CH3CN and pyridine, in complexes of the type LMo(CN),.. CH3CN was reported to be a slightly weaker a-donor and i r -acceptor than pyridine. Cotton (8) also considers that CH3CN can function as a ir-acceptor too. Complexes bonded with acrylonitrile (CH7=CH-CN) - 61 -Table 111-10. Dq values of six-coordinate Ni (d ) complexes in different solvents'(6) D (cm ) q ^ 1000 773 769 1026 Solvents pyridine DMSO DMA DMF acetone sulfolane CH3CN CH3N02 - 62 -through the C=N bond have been reported (9), but such bonding is not very l i k e l y in the present system. The findings for the CH^NC^ system are interesting, since CH^ NC^  is a weak donor, the donor strength being very similar to that of nitrobenzene (5,11). But as shown in the results, P as well as Cl dissociation take place i n this solvent. It seems surprising therefore that the system is ineffective for catalytic hydrogenation. Further investigation of the separate phosphine and chloride dissociation would be necessary for more meaningful discussion. There is no simple relationship between the K values of reaction (3.3) and the relative catalytic activity of the solvent systems. Qualitatively, some systems with no measurable P dissociation (acetone K < 10 ^ M) or with considerable dissociation due to a strong -2 coordinating solvent (pyridine, K ^ 10 M) are found to be inactive; those systems with perhaps just about measurable K values (DMSO, DMA, DMF) have been found active probably via a similar mechanism. The -4 acetonitrile system has an intermediate K value (10 M) but i s inactive for hydrogenation; however this system behaves differently toward the olefin substrate compared to the DMSO, DMA and DMF systems. The sulfolane system has an immeasurably small K value but is active c a t a l y t i -cally: again, however, this system involves a quite different mechanism I involving active cationic intermediates (Chapter IV). As outlined previously, the reaction between Ir(C0)ClP 2 and MA i n various solvents can be represented by one of the following equilibria: K1 Ir(C0)ClP 2 + MA -^Z±_ Ir(C0)ClP2(MA) (3.10) - 63 -or K Ir(CO)ClP 2 - — ^ Ir(CO)ClP + P 2 (3.11) K Ir(CO)ClP + MA — ^ Ir(CO)ClP(MA) K 3 or IrCOClP^ + MA — ^ Ir(CO)ClP(MA) + P[perhaps via (3.10)] (3.12) 2 ^ — James and Memon (4) investigated the Ir(C0)C1P2_MA hydrogenation system in DMA, and detailed kinetic studies indicated that olefin complex formation went through route (3.11). Other workers (15,23) have also concluded that a Ir(CO)CIP(olefin) species was the active complex in amide solvents and in toluene, but considered that the P dissociated from an i n i t i a l l y formed Ir(C0)ClP2(olefin) complex. Wilkinson and coworkers (14a) have also suggested possible catalyst production by P dissociation from the dihydride complex, R^Ir(C0)C1P2» but no evidence for such a hydride route mechanism has been presented. Strohmeier and Fleischmann (16) have presented K ^  values for reaction (3.10) in toluene for dimethyl maleate, fumarate, maleic anhydride, ethyl acrylate, styrene and heptene-1 with various Ir(C0)X(PR.j)2 complexes (X = halide: R = C-H., - , C,H_, 0C,Hc). For each substrate, K 1  r ' 6 11 6 5 6 5 m increases i n the order I < Br < Cl, and K 1 also increases i n the order m R = C 6H 1 1< CgH5 < OCgH,.. Both trends indicate increasing s t a b i l i t y of the complexes with decreasing electron density at the Ir-atom. The st a b i l i t y also increases with stronger Tr-acceptor properties of the olefin, i.e., with electron density at the carbon-carbon double bond (16). 1 4 -1 At 20° in toluene, the K value varies from about 10 M for maleic m anhydride adduct with Ir(C0)Cl[P(0C 6H 5) 3] 2 to the smallest measured - 64 -% 3 x 10 ^  M for styrene adduce with the same complex. At 20° i n toluene, the K value for olefin complex formation between Ir(C0)ClP2 and maleic anhydride i s ^ 350 M (16). In the present study, olefin complex formation i s not observed in the pyridine, DMSO and DMF solvents. In DMA, addition of excess P stopped the olefin complex formation, which must therefore occur via equilibria (3.11) or (3.12). In acetone, there i s no P dissociation in a solution of Ir(CO)ClP 2; the spectral data are consistent with olefin complex formation occurring via reaction (3.10), followed by subsequent 1 P dissociation from the Ir(CO)CIP2(MA) species. The K m value at 35° (1.1 M ^) i s much lower than that for maleic anhydride bonding i n toluene. Olefin complex formation in CH^CN may go via any of reactions ,/ (3.10) to (3.12), and i t i s interesting to note that., on addition of P into a CHgCN solution of Ir(C0)C1P2-MA, a further new complex i s formed, Fig. III-6. Halpern and Chen (7) have reported that i n chloro-benzene or DMF, Ir(CO)Cl(PMe 2Ph) 2 may react with added PMe2Ph according to the following equations: Ir(CO)Cl(PMe 2Ph) 2 + PMe2Ph — > Ir(CO)(PMe 2Ph) 3Cl (3.20a) 11 PMe Ph I" Ir(CO)Cl(PMe 2Ph) 4 < - — - [Ir(CO) (PMe^h) 3] + C l " (3.20b) Similar reactions forming 5-coordinate species may occur in CH^CN solutions of Ir(C0)ClP 2 and MA; the phosphine could react with a Ir(C0)ClP 2(MA) or Ir(C0)ClP(MA) species. - 65 -References 1. L. Vaska and J.W. Diluzio, J. Amer. Chem. Soc, 8 4 , 679 (1962). 2. L. Vaska and M.F. Werneke, Trans. N.Y. Acad. Sci., _31, 70 (1971). 3. W. Strohmeier and T. Onoda, Z. Naturforsch., 23b, 1527 (1968). 4. B.R. James and N.A. Memon, Can. J. Chem., 46_, 217 (1968). 5. V. Gutmann, New Pathways i n Inorg. Chem., Cambridge University Press, 1968. 6. R.S. Drago, D.W. Meek, M.D. Joeston, L. LaRoche, Inorg. Chem., !_•> 124 (1963). 7. W.A.G. Graham, Inorg. Chem., ]_, 315 (1968). 8. F.A. Cotton, Inorg. Chem., _3> 702 (1964). 9. R. Jones, Chem. Rev., 68_, 785 (1968). 10. R.S. Drago, B. Wayland, R.L. Carlson, J. Amer. Chem. Soc, 85, 3125 (1963). 11. R.S. Drago, K.F. Purcell, Non-aqueous Solvent Systems, Edited by T.C. Waddington, Academic Press, 1965, p. 234, 12. CH. Langford and P.O. Langford, Inorg. Chem., 1, 184 (1962). 13. L. Vaska and R.E. Rhodes, J. Amer. Chem. Soc, 87^ , 4970 (1965). 14. (a) J.A. Osborn, F.H. Jardine, J.F. Young and G. Wilkinson, J. Chem. Soc A, 1711 (1966); (b) B.R. James and G.L. Rempel, Discuss. i Faraday Soc, 46_, 48 (1968); (c) W. Strohmeier and T. Onoda, Z. Naturforsch., 24b, 515 (1969). 15. W. Strohmeier and T. Onoda, Z. Naturforsch., 24b, 461 (1969); ibi d . , 24b, 1493 (1969). - 66 -16. W. Strohmeier and R. Fleischmann, Z. Naturforsch, 24b, 1217 (1969). 17. J.Y. Chen and J. Halpern, J. Amer. Chem. Soc, 93, 4939 (1971). 18. P.B. Chock and J. Halpern, J. Amer. Chem. Soc,«8_8, 3511 (1966). 19. L.M. Venanzi, Advances i n Chem., Amer. Chem. Soc, Vol. 98, p. 66. 20. B.R. James, Homogeneous Hydrogenation, Wiley-Interscience, New York, 1973, p. 293. 21. L. Vaska, Inorg. Chim. Acta, ,5, 295 (1971). 22. B.R. James, Homogeneous Hydrogenation, Wiley-Interscience, New York, 1973, p. 299. 23. M.G. Burnett, R.J. Morrison, and C.J. Strugnell, J. Chem. Soc, Dalton, 701 (1973). CHAPTER IV CATALYTIC HYDROGENATION OF OLEFINS USING Ir(C0)ClP o IN SULFOLANE SOLUTION IV.1 Introduction In Chapter III, a cursory study was described on the effects of solvents on the catalytic hydrogenation of MA using the Ir(CO)ClP 2 (P = PPh^) complex. Under similar conditions, the rate of hydrogenation in sulfolane was found to be about 4 times faster than that in DMA solvent. Detailed studies in the sulfolane solvent system revealed somewhat different kinetic dependences to those observed for the DMA and DMSO systems, possibly indicating that a different mechanism i s involved. The following sections give the details for the studies i n the sulfolane system. IV.2 Kinetic Measurements of Catalytic Hydrogenation of MA and DEMA Using Ir(CO)ClP 2 in Sulfolane In sulfolane solution, the Ir(CO)ClP 2 complex is a reasonably efficient catalyst for homogeneous hydrogenation of certain olefins at temperatures around 80° and H 9 pressure of 1 atm or below. The - 68 -hydrogenation of MA and DEMA were studied i n detail. Succinic acid (m.pt. is 185°) and diethyl succinate (v.p.c. column = dinonyl Phthalate, T = 220°, current = 200 mA, pr. of He = 10 lb/i n , time = 3.5 min) were obtained as the corresponding products. In the presence of substrates under H2, Ir(C0)ClP 2 dissolved i n sulfolane to give a pale yellow solution; no solids were observed during or after the completion of the catalytic hydrogenation. The so l u b i l i t y of H 2 a t 80° and 1 atm. -3 pressure of H^ , determined as 2.6 x 10 M, was assumed to obey Henry's Law. The parti a l pressure of sulfolane is ^  2 mm Hg at 80° and is considered negligible over the range of temperatures studied. A typical H 2 uptake plot for hydrogenation of MA i s shown i n Fig. IV-1. A slow i n i t i a l induction period was observed, followed by an extensive linear region. The rate then f e l l off slowly, and stopped at the point where the mole ratio of E^% olefin was close to unity, indicating the completion of hydrogenation. The hydrogenation rate was determined from the slope of the linear region i n terms of M sec ^ . IV.2.1. Rate Dependence on Concentration of Catalyst, Substrate and  Hydrogen. MA and DEMA were used as substrates i n these catalytic hydrogenation systems. The data showing the variation of hydrogenation rate with concentration of Ir(C0)ClP 2, MA and H 2 in the Ir(C0)C1P2-MA-H2 system are summarized in Table IV-1. Similar data for the Ir (CO) CIP^DEMA-H, system are summarized i n Table IV-2. The shape of the uptake plots were of the same form at a l l the conditions studied, and the kinetics i n each case were measured in the linear region, For the MA system, the - 69 -- 70 -Table IV-1. Kinetic data for the Ir(C0)ClP 2 catalyzed hydrogenation of MA at 80°. [Ir] [MA] p[H ] [HI linear rate 3 1 fi x 10 ,M M mm Hg x 10 ,M x 10 ,M sec" 0 0.117 760 2.6 0 1.33 0.117 760 2.6 4.9 2.93 0.117 760 2.6 9.5 5.36 0.117 760 2.6 14 9.95 0.117 760 2.6 18.8 2.93 0 760 2.6 0 2.93 2.9 x 10 - 2 760 2.6 3.4 2.93 8.8 x 10~ 2 760 2.6 9.0 2.93 0.124 760 2.6 9.5 2.93 0.147 760 2.6 10.1 2.93 0.488 760 2.6 12.5 2.93 0.294 760 2.6 11.2 2.93 0.294 380 1.3 15.3 2.93 0.294 202 0.69 14.5 2.93 0.294 46 0.16 16.9 2.93 0.294 0 0 0 2.93 0.117 559 1.9 10 2.93 0.117 380 1.3 11.7 2.93 0.117 361 1.2 10.9 2.93 0.117 103 0.35 11.8 2.93 0.117 54 0.19 10.5 - 71 -Table IV-2. Kinetic data for the Ir(CO)ClP 2 catalyzed hydrogenation of DEMA at 80°. [Ir] = 2.93 x 10"3M [DEMA] p[H ] [H ] linear rate 3 6 M mm Hg x 10 ,M x 10 ,M sec" 0.273 760 2.6 1.71 0.218 760 2.6 1.41 0.109 760 2.6 0.95 5.45 x 10~ 2 760 2.6 0.66 1.09 x 10~ 2 760 2.6 0.17 0 760 2.6 0 0.273 380 1.3 1.10 0.273 104 0.36 0.73 0.273 44 0.20 0.46 0.273 0 0 0 - 72 -dependences of hydrogenation rate on the concentration of Ir(C0)ClP2 catalyst, substrate and a r e shown in Fig. IV-2, Fig. IV-3 and Fig. IV-5 ,respectively. For the DEMA system, the dependences of hydrogena-tion rate on the concentration of substrate and on the concentration are shown in Fig. IV-4 and Fig. IV-6. For the MA system, the rate i s f i r s t order i n iridium at low catalyst concentrations but becomes less than one at higher concentrations; the hydrogenation rate i s f i r s t order i n substrate at low concentrations and approaches zero order at higher concentrations. The same substrate dependence i s observed in the DEMA system but the levelling-off region is not so pronounced as in the MA system. Under the same conditions, the hydrogenation rate of MA i s about 7 times faster than that of DEMA. A dramatic difference between the MA and DEMA systems is apparent i n the rate dependences on hydrogen concentration. For the MA system, the dependence was studied at two MA concentrations (Fig. IV-5). The series of experiments at 0.294 M MA showed that the hydrogenation rate decreased as the concentration (or pressure) increased from 50 mm to 760 mm Hg; at 0.117 M MA, a similar but less marked inverse dependence was evident. In the DEMA system, the hydrogenation rate showed a first-order dependence on concentration at low concentrations, and then became less than -3 ^  first-order as the concentration increased up to 2.6 x 10 M (or' 1 atm, pressure). No gas uptake was observed in the absence of the iridium complex. - 73 -20 L 0 2 4 6 8 10 [Ir] x 10 ,M Fig. IV-2. Variation of hydrogenation rate vs. concentration of Ir(C0)ClP_; [MA] = 0.117 M, p[H 2] = 760 mm Hg, T = 80°. - 74 -0.2 0.4 M, [DEMA] Fig. IV-4. Variation of hydrogenation rate vs. concentration of DEMA; [Ir] = 2.93 x 10~3, p[H„] - 760 mm Hg, T = 80°. - 75 -- 76 -- 77 -IV.2.2 Rate dependence on temperature. Hydrogenation rates were measured over the temperature range 75 -85° for both systems. The data for the MA system are summarized i n Table IV-3 and those for DEMA system are summarized in Table IV-4. Corresponding Arrhenius-type plots log(hydrogenation rate) vs. 1/T at these conditions are shown in Fig. IV-7 and Fig. IV-8; both give good straight lines. The A H ^ values calculated from the slopes are 30 ± 1 kcal mole for the MA system, and 27 i 2 kcal mole for the DEMA system. The effect of the small differences i n sol u b i l i t y on the hydrogenation kinetics over the temperature range were neglected. IV.2.3 Effect of addition of P (P = PPh 3) on hydrogenation rate. The catalytically active species for the Ir(C0)ClP2 system i n DMA solution i s produced at some stage (1,2) by the dissociation of a P group from the iridium, and the addition of P was found to stop the catalytic activity completely. The data showing the effect of addition of P on the hydrogenation rate i n sulfolane are summarized in Table IV-5 for the MA system and i n Table IV-6 for the DEMA system. The dependences are shown in Fig. IV-9 and Fig. IV-10. Even at very high [P]:[Ir] ratios the hydrogenation rate was decreased by only about 30% in the MA system; about an 80% reduction in rate was observed i n the DEMA system at the highest phosphine concentrations. The effect of added P on rate was also studied at a fixed ratio of [P]/[Ir] = 8 using different MA concentrations. The data are summarized in Table IV-7, and the variation is shown in Fig. IV-3. Again, the hydrogenation rates of MA were decreased by ^  30 - 40%, except at very low MA concentrations, where the reaction was completely Inhibited. - 78 -Table IV-3. Ir(C0)C1P2-MA-H2 system. Temperature dependence of hydrogenation rate [Ir] = 2.93 x 10~ 3 M; [H2] = 2.6 x IO - 3 M. [MA] T° linear rate M x 106,M sec" 0.294 75 4.5 0.294 80 11.0 0.294 82.5 11.9 0.294 85 16.5 0.117 75 3.9 0.117 80 9.5 0.117 85 16.5 Table IV-4. Ir(C0)ClP2-DEMA-H2 system. Temperature dependence of hydrogenation rate [Ir] = 2.93 x 10 - 3 M, [DEMA] = 0.273 M, [H,,] - 2,6 x 10~ 3 M. linear rate x I O ^ JM sec 1 75 0,77 80 1.71 82.5 i.82 85 2,43 - 79 -log (rate) -4.8 -5.0 -5.2 -5.4 -5.6 Fig. IV-7. 2.80 2.85 1/T x 10~ 2.90 Arrhenius-type plot for Ir(C0)C1P2-MA-H2 system; [Ir] = 2.93 x 10"3M, p[H 2] - 760 mm Hg; (A) [MA] » 0.117 M, (0) [MA] = 0.294 M. -5.2 Fig. IV-8. 2.80 - 2.85 2.90 1/T x 10 Arrhenius-type plot for Ir(C0)C1?2-DEMA-H2 system; [Ir] - 2.93 x 10 - 3 M, [DEMA] = 0.273 M9 p[H ] = 760 mm Hg. - 80 -Table IV-5. Ir(C0)C1P2-MA-H2 system. Effect of added P on rate of hydrogenation [Ir] = 2.93 x IO - 3 M, [MA] = 0.294 M, p[H 2] = 760 mm Hg, T = 80°. [P] [P]^ linear rate M ^ 1 x io6,M sec" 3.03 x 10 2.24 x 10 7.15 x 10 -3 I -2 I -2 1 7.7 24 11.0 9.1 7.2 7.0 Table IV-6. Ir(C0)ClP2-DEMA-H2 system. Effect of added P on rate of hydrogenation [Ir] = 2.93 x IO - 3 Ms [DEMA] = 0.273 M, p[H ?] =» 760 mm Hg, T = 80°. [P] [P], linear rate M x 10 ,M sec 0 - 1.71 2.82 x 10 - 3 1 1,25 2.98 x 10" 2 10 0.81 5.73 x 10 - 2 20 0.40 0.172 60 0.27 - 81 -rate x 10^,M sec 10 5 r t l r ] [MA] P[H ?] T 2.93 x 10 M 0.294 M 760 mm Hg. 80° -O-[P] x 10 ,M Fig. IV-9. Variation of hydrogenation rate vs. added [P] for Ir(CO)ClP2-MA-H2 system. rate x 10 , M sec ^  [Ir] = 2.93 x 10 M [DEMA] = 0.273 M p[H ?] = 760 mm Hg. t - 80° 0 0.1 0.2 [P],M Fig. IV-10. Variation of hydrogenation rate vs. added [P] for Ir(CO)ClP2-DEMA-H2 system. - 82 -Table IV-7. Ir(C0)C1P2-MA-H2 system. Effect of added P on rate of hydrogenation. [Ir] = 2.93 x 10~ 3 M, pfH,,] = 760 mm Hg, T = 80°; [P] = 2.3 x 10" 2 M, [P]/[Ir] = 8. [MA] linear rate M x 106,M s e c - 1 2.94 x 10" 2 0 0.117 5.0 0.294 7.2 0.479 7.9 - 83 -IV.2.4 Effect of added acid on hydrogenation rate. On addition of a strong acid (p-toluene sulfonic acid p-CH_C,.H.SO.H) j O H J to the sulfolane systems at [H ]:[Ir] mole ratios up to 40, the hydrogenation of both MA and DEMA were increased compared to the systems with no added acid. In the MA system, the rate at f i r s t increased with added acid to a maximum about twice the rate with no added H+, and then decreased somewhat at higher acidities (Fig. IV-11). For the DEMA system, the rate increased in a reasonably first-order manner with added acid to ^  0.05 M and then levelled off somewhat at higher acidities to a value 'v* 35 times the rate i n the absence of added acid (Fig. IV-12), The data are summarized i n Table IV-8 and Table IV-9, In a l l these experiments no hydrogenation of nCH^CgH^SO^H was observed. IV.2.5 Effect of added salts on hydrogenation rate. Due to the insolubility of oCH0C,H.SO.Na i n sulfolane, possible salt effects due to sulfonate i n the acid dependence studies (IV.2.4) could not be studied. Thus the effect of other salts on the hydro-genation rate was investigated. The changes of hydrogenation rate were followed as a function of variation of added L i C l and added LiNO^. The data are summarized i n Table IV-10 and Table IV-11 for the MA system, and the chloride and nitrate dependences are shown in Fig. IV-13. For the DEMA system, the data for the effect of added L i C l are summarized in Table IV-12, and the dependence is shown in Fig. IV-14. Figures IV-13 and IV-14 show dramatic differences between the MA and DEMA systems. Addition of about 1 mole L i C l per Ir in the MA system completely stopped the reaction, while in the DEMA system, no appreciable - 84 -Table IV-8. Ir(C0)C1P2-MA-H2 system. Effect of added acid (pCH_C,H.SO.H) on hydrogenation rate [Ir] = 2.93 x 10~JM, p[H 2] = 760 mm Hg, T = 80°. [MA] [H+] [H +]. linear rate M M x 10^,M sec 1 0.304 0 11.4 0.304 1.74 x 10~ 2 6 24.8 0.304 5.28 x 10~ 2 18 22.4 0.304 9.85 x 10" 2 33 18.3 3.46 x 10~ 2 0 5.5 3.46 x 10~ 2 2.72 x 10~ 2 9 12.7 3.46 x 10~ 2 0.116 40 12.1 Table IV-9. Ir(C0)C1P2-DEMA-H2 system. Effect of added acid (pCH^H^SO^) on hydrogenation rate [Ir] = 2.93 x 10"3M, [DEMA] = 0.273 M, p[H 2] = 760 mm Hg, T = 80°. [H +] [ H + ] / linear rate M [ I r ] x 106,M sec" 1 0 - 1.71 -2 4.73 x 10 1.6 3.71 2.78 x 10~ 2 9 25.7 1 -2 3.66 x 10 13 31.5 4.72 x 10~ 2 16 47.2 , « -2 6.98 x 10 24 50.0 0.102 35 50.0 0.119 40 58.2 » - 85 -rate x IO**,M sec 1 [MA] = 0.304 M [MA] = 3.46 x 10 -2 M 0.05 [H+],M 0.10 Fig. IV-11. Variation of hydrogenation rate on added acid (pCH 3C 6H 4S0 3H); [Ir] = 2.93 x 10 -3 M , p[H ] = 760 mm Hg, T = 80°; (0) [MA] = 0.304 M, (A) = 3.46 x 10 -2 M. - 86 -rate x 10^,M sec 1 - 87 -Table IV-10. Ir(CO)ClP2-MA-H2 system. Variation of hydrogenation rate on added L i C l [Ir] = 2.93 x 10" 3 M, [MA] = 0.293 M, p[H 2] = 760 mm Hg T = 80°. [LiCl] M [Cl~] [Ir] linear rate x 10 ,M sec 1 2.26 x 10 3.73 x 10 9.29 x 10 3.26 x 10" 0.138 -3 -3 0.8 1.2 3 11 47 11.0 2.8 0.3 'v-O 0 0 Table I V - l l . Ir(C0)C1P2-MA-H2 system. Variation of hydrogenation rate on added LiNO^ [Ir] = 2.93 x 10_3M, [MA] = 0.293 M, p[H 2] = 760 mm Hg, T = 80° M [LiNO 3] L N 0 r ] / linear rate x 10 ,M sec u - 11.0 1.02 x 10" 3 0.4 10.5 4.35 x 10~ 3 1.5 9.1 8.30 x 10~ 3 3.0 3.0 1.73 x 10" 2 6 1.4 3.16 x 10" 2 11 0 rate x 10^,M sec" 1 0 - 0.01 0.02 [LiCl] or [LiNOg], M Fig. IV-13. Variation of hydrogenation rate on added L i C l or LiNO^; [Ir] = 2.93 x 10"3K, [MA] • 0.293 M, p[H 2] = 760 mm Hg, T (0) [ L I N O 3 ]; (A) [LiCl] 0.03 80°; - 89 -Table IV-12. Ir(CO)C1P2-DEMA-H2 s y Stem. Variation of hydrogenation rate on added L i C l . [Ir] = 2.93 x 10_3M, [DEMA] = 0.273 M, p[H 2] = 760 mm Hg, T = 80°. [LiCl] [ C l ~ ] , linear rate 3.30 x 10 3.04 x 10 0.163 -3 -2 1 10 55 1.71 1.79 1.35 1.36 o to vO o a) 4J u I 0.04 0.12 0.08 [LiCl],M Fig. IV-14. Variation of hydrogenation rate on added L i C l for 0.16 Ir(CO)ClP2-DEMA-H2 system. - 90 -effect on the hydrogenation rate was observed even at relatively high concentrations of added L i C l . The effect on the ester hydrogenation rate by the addition of a number of other reagents—NaOCH^, (n-Bu).jN, NaClO^, p i c r i c acid—was also studied. The results are summarized in Table IV-13. A significant inhibition was observed on adding methoxide, other effects were quite small. IV.3 Catalytic Hydrogenation of other Olefins and Acetylenes. A number of other olefins and acetylenes were tried as substrates for catalytic hydrogenation at one set of conditions using the sulfolane system. The results are summarized in Table IV-14. Of the substrates l i s t e d , only a c r y l i c acid and fumaric acid were effectively hydrogenated, IV.4 Ir(C0)ClP 2 and i t s Properties i n Sulfolane Solution. Ir(C0)ClP2 i s an air-stable pale yellow crystalline s o l i d . Its reactions i n solfolane solvent with hydrogen, substrates (maleic acid and diethyl maleate), L i C l , and PPh^ were studied i n order to learn more of the catalytic hydrogenation processes. These studies are described below. A number of the studies were br i e f l y mentioned in Chapter III. IV.4.1 Physical properties in sulfolane. In sulfolane, at 35°, under N 2, the yellow solution of Ir(C0)ClP 2 3 3 gives absorption peaks at 340 nm (e = 3.1 x 10 ), 387 nm (e = 3.7 x 10 ) - 91 -Table IV-13. Ir(C0)C1P2-DEMA-H2 system. Effect of added salts and other reagents on hydrogenation rate. [Ir] = 2.93 x 10" 3 M, [DEMA] = 0.273 M, p[H 2] = 760 mm Hg, T = 80°. added salt concentration linear rate M x 106,M sec" 0 - 1.71 NaOCH3 3 x 10~ 3 0.71 (n-Bu)3N 3.66 x 10~ 3 1.35 (n-Bu)3N 3.40 x 10~ 2 1.24 NaC104 2.16 x I O - 2 1.60 pi c r i c acid 3.8 x 10~ 2 2.16 - 92 -Table IV-14. Hydrogenation of olefins and acetylenes using Ir(CO)ClP 2 in sulfolane. [Ir] = 2.93 x 10~ 3 M [substrate] ^ 0.1 M p[H 2] = 760 mm Hg. T = 80° substrate reaction rate 1-hexene 1-pentene 1- octene 2- hexene cyclo-hexene cyclo-octene 1.3- cyclo-hexadiene arcyclic acid (CH2=CHC00H) cinnamic- acid (Ph-CH=CH-C00E) 1.4- dicyano-butene-2 (CN)2C=C(CN)2 trans-s tilbene PhCECH PhC=CPh PhCECCOOH HOOCCECCOOH i Indene FA MA not dissolved + not dissolved + + 6.5 x 10 ^  M sec 1 1.9 x 10"6 M sec" 1 10 x 10~ 6 M sec" 1 - 93 -2 and 439 nm (e = 9 x 10 ), and the spectrum remained unchanged with respect to time. The absorption peaks are in the same position as those observed i n the reflectance spectrum of a powdered sample suggesting that the complex remains as Ir(C0)ClP 2 in solution. The same peaks were also observed in the spectrum of benzene solutions of Ir(C0)ClP 2 which similarly remained unchanged i n the absence of 0^° The i . r . spectrum of Ir(C0)ClP 2 i n nujol gives a very sharp and strong carbonyl peak at 1950 cm \ IV.4.2 Reaction with hydrogen to form a dihydride, At 35°, Ir(C0)ClP 2 in sulfolane solution undergoes a quite rapid reversible reaction with H 0 (t, ^  1 min at 0.5 atm of H_), the Cm ^ Z yellow color of the chloro complex fading; the spectrum of the dihydride shows a continuum below 400 nm with slight absorption at 387 nm (e = 200) compared with the i n i t i a l chloro complex. The equilibrium constant for equation (4.1) Ir(C0)ClP 2 + H 2 v Ir(C0)ClP 2(H 2) (4.1) was readily evaluated from the absorption data at 387 nm over the temperature range 60 - 80°, and the data are summarized i n Table IV-15, The'AH and AS values for reaction (4.1) were evaluated as AH = - 13 ± 1 Kcal mole - 1 and AS = - 22 ± 3 e.u. - 94 -Table IV-15. Equilibrium constant for dihydride formation. Ir(CO)ClP 2 + H 2 s ^ * Ir(CO)ClP 2(H 2) [Ir(CO)ClP 2(H 2)] *H = [Ir(CO)ClP 2][H 2] Temp [Ir] x 10 4 p[H 2] [H 2] x 10 3 O.D. a * H X 1 0 " 3 M mm Hg M 387 nm M"1 80 3.66 100 0.34 0.85 1.9 » 80 3.66 184 0.63 0.59 2.3 / average: 80 3.56 380 1.3 0.41 2.1 I 2.2 80 3.62 760 2.6 0.25 2.4 •' 70 3.64 100 0.34 0.65 3.5 » 70 3.56 188 0.64 0.44 3.7 / average: 70 3.62 380 1.3 0.32 3.2 1 3.5 70 3.59 760 2.6 0.19 3.7 ' 60 3.54 100 0.34 0.50 5.5 i average: 60 3.64 184 0.63 0.31 6.9 ) 6.2 10 mm c e l l path - 95 -The i . r . spectrum of the dihydride in sulfolane (prepared in situ) gave a carbonyl peak at 1980 cm \ and a sharp hydride peak at 2080 cm \ When deuterium was used instead of hydrogen, the i . r . spectrum of the Ir(III) dideuteride complex, prepared in s i t u , showed the carbonyl peak at 2000 cm 1 and the hydride peak was absent; no deuteride peaks were observable in the 1400 cm 1 region because of the presence of the solvent bands. The dihydride complex I^CXOCn^R^) was prepared according to the method of Vaska and coworkers (3) and i t s i . r . spectrum in nujol was studied. A carbonyl peak was observed at 1980 cm \ and two hydride peaks at 2100 and 2196 cm ^ . When the dihydride was dissolved in benzene, the i . r . spectrum gave the carbonyl peak and the hydride peaks at the same positions as those observed i n nujol. However, when the dihydride was dissolved in sulfolane, the i . r . spectrum, as i n the si t u system, gave only one hydride peak at 2080 cm 1 as well as the carbonyl peak at 1980 cm"1. IV.4.3 Reaction with substrates; MA and DEMA. IV.4.3a Formation of olefin complex. The i . r . spectrum of a sulfolane solution containing Ir(C0)ClP2 and -1 MA under N 2 gave the carbonyl peak at 2020 cm . The shifting of the carbonyl peak indicates coordination of the MA with the iridium complex, Coordination was also indicated by changes i n the UV/VIS solution spectra; the yellow color of the chloro complex faded on addition of MA. The formation of the olefin complex was followed by measuring the change in absorption at 387 nm. On addition of MA to the Ir(C0)ClP 9 solution, a - 96 -rapid drop of absorption at 387 nm was observed indicating rapid formation of an olefin complex. This was then followed by a slower decrease in absorption (Fig. IV-15), and this change is attributed to a slow dissociation of Cl from the iridium-maleic acid complex (see later discussion) . The process can be represented as follows: K 1 k Ir(CO)ClP 2 + MA Ir(CO)ClP2(MA) ^==i [Ir(CO)P2(MA) ] + c l (4.2) fast k_^ slow, (K^) Following the procedure outlined in Appendix IV-1, the values of K 1 and K.^  were evaluated, and the data are summarized in Table IV-16. In presence of added L i C l to the solution of Ir(CO)ClP 2, the spectrum of the above solution was identical to that of Ir(CO)ClP 2 without added L i C l . But, on addition of a small quantity of L i C l ([Ir]:[Cl~]=*1:2) into a solution of Ir(C0)C1P2-MA ([Ir] = 3.9 x 10-4M, [MA] = 5.7 x 10~ 2 M), the v i s i b l e spectrum gave a continuum from 350 to 750 nm, and no peak was observed at 387 nm, quite different to the rapidly established spectrum i n the absence of added chloride. Addition of P into the solution of Ir(C0)C1P2-MA did not affect formation of the ol e f i n complex or the subsequent Cl dissociation, see Table IV-16. As the equilibrium for the formation of the ol e f i n complex is fast, and the dissociation of Cl is slow, the i n i t i a l part of a first-order log plot (O.D. vs. time) for the slow O.D. decrease should yield values of k^, since the [Ir(C0)P 2(MA)] + and [Cl ] concentrations are i n i t i a l l y close to zero. From the slope of such plots (Fig. IV-16) the approximate -4 -1 values of rate constant k 1 were estimated to be ^ 1 x 10 sec . A - 97 -Table IV-16. Equilibrium constants for equilibrium K1 m 1 Ir(CO)ClP 2 + MA Ir(CO)ClP2(MA) v N [Ir(CO)P2(MA)] + Cl" fast slow Temp [Ir] x 10 M [MA] M O.D. a,c a,d 387 nm O.D 387 nm .1 m -1 1^ x 10_ M 35 3.54 1.36 x 10~ 3 1.26 1.21 8.5 6,6 35 3.56 2.29 x 10~ 3 1.26 1.19 18 3.4 35 3.54 1.56 x I O - 2 1.05 0.91 15 4,2 35 3.62 3.98 x 10 - 2 0.90 0.74 22 3.0 35 3.62 7.53 x 10 - 2 0.62 0.51 15 3,5 35 3.44 0.169 0.28 0.22 16 2.8 35 3.56 0.279 0.34 0.20 12 10 35 3.49b 4.67 x I O - 2 0.78 0.58 14 11 50 3.49 0.117 0.63 0.31 7.5 50 a 10 mm c e l l path b with addition of PPh3; [P] = 4.21 x 10 - 3 M c absorption shortly after addition of MA d absorption at equilibrium - 98 -O.D. at 387 nm 1.4 0 1 . 2 1.0 \-0.8 Fig. IV-15. 120 240 360 800 Time,min Change of absorption for Ir(C0)C1P2"MA in sulfolane under N 2, [Ir] = 3.54 x 10-4M, [MA] = 1.56 x 10_2M, T = 35°; 10 mm c e l l path. - 99 -log (A-A ) 120 240 Time,min. Fig. IV-16, First-order log plot of absorption change for the Ir(C0)ClP2-MA system, [Ir] = 3.54 x 10_4M, [MA] = 1.56 x 10_2M, T = 35°, 10 mm c e l l path, under N 2 A, A^ are the absorptions at time t and at equilibrium. - 100 -mathematical approach for a general solution for in systems of the type exemplified in eqn. (4.2) is shown in Appendix IV-2. It is impos-sible to estimate the value of k^ i n a simple manner since the general solution for such a sequence of reactions is very complicated. In the case of the DEMA substrate, a rapid decrease in absorption at 387 nm was again observed on addition of the DEMA to the iridium solution under N 2» however, there was no significant observable subsequent slower decrease i n absorption. The data were analyzed for K 1 in the m same manner as outlined for the MA complex, and the results are summarized i n Table IV-17. The K 1 value for DEMA i s about 10 times m smaller than that for the MA system under the same conditions. Any dissociation of Cl from the Ir(C0)C1P2(DEMA) complex is not observable in the v i s i b l e absorption studies. In contrast with the MA system, the presence of added L i C l did not affect the spectrum of Ir(C0)ClP2-DEMA solution, see Table IV-17. IV.4.3b Dissociation of Cl from the o l e f i n complexes. The presence of free Cl in the sulfolane solution of Ir(C0)ClP 2 and MA, was confirmed by addition of aqueous AgNO^ /HNO^  solution; a white cloudy suspension was observed instantaneously. Such a suspension-was also observed for the iridium solution containing the DEMA. No cloudy suspension was observed when aqueous AgNO^ /HNO^  was added to a sulfolane solution of Ir(C0)ClP 2, or MA or DEMA. An observation of interest i s that on addition of aqueous AgN0,/HN0 into a sulfolane J 3 solution of Ir(C0)ClP 2 under H,,, a dark brown solution was i n i t i a l l y produced, followed by precipitation on standing of a black solid, - 101 -Table IV-17. Ir(C0)C1P2-DEMA system Equilibrium constant for the equilibrium at 35° Ir(C0)ClP 2 + DEMA ^ Tr fCOlCIP^ fDEMA^  [Ir] x 10 4 [DEMA] O.D.a K 1 m M M 387 nm M"1 3.62 1.56 x 10~ 2 1.27 3.0 3.62 6.24 x 10~ 2 1,24 1.2 3.58 0.186 1.02 1.5 3.52 0.234 1.0 1.6 3.59b 1.09 x 10~ 2 1,29 2.5 a 10 mm c e l l path b on addition of L i C l , [LiCl] = 5.5 x 10~ 3 M - 102 -Reduction of the Ag T to Ag-metal i s probably occurring. The presence of free Cl in the sulfolane systems under hydrogenation conditions was further confirmed by potentiometric method using a specific chloride electrode. The apparatus set up (4) is shown in Fig. IV-17. A calibration curve using Ph^AsCl was determined and i s shown in Fig. IV-18. The calibration curve was determined twice, and gave consistent results. The data on the Cl dissociation for the Ir(CO)C1P2~DEMA-H2 system are summarized in Table IV-18. The potentials were "equilibrium" values taken after ^ 2 hr during which time the salt bridge was removed; the recorded values taken after that time s t i l l varied irregularly by ^ 5 - 10 mV for each individual experiment. The extent of dissociation i s small ($ 5%). No Cl dissociation was observed in solution of Ir(C0)ClP 2 or Ir(C0)C1P 2-H 2 (Table IV-18). The equilibrium constant for Cl dissociation from a Ir(CO)C1P2(DEMA) complex was estimated to be i n the range 10~ to 10~ 4 M (Table IV-18). IV.4.3c Conductance measurements on the hydride and the olefin complexes  i n sulfolane. The equivalent conductances of solutions containing the ol e f i n complexes and hydride complex were determined using the apparatus described i n section II.4, and the data are summarized in Table IV-19, IV.4.3d Possibility of proton dissociation from the hydride and o l e f i n  complexes. The possibility of proton dissociation from the dihydride complex Ir(C0)ClP ?(H 2) was investigated by potentiometric t i t r a t i o n . The - 103 -Table IV-18. Determination of Cl dissociation in the Ir(CO)ClP2-DEMA-H2 system in sulfolane at 80°, p[H 2] = 760 mm Hg [Ir] x 10 3 [DEMA] P[H 9] potential [Cl -] % Cl reading l M M mm Hg mV dissociation M 8.3 0.16 760 -85 -*• -95 4 x 10~ 4 5 3 x 10~ 3 11.7 0.27 760 -70 -> -77 2 x 10~ 4 1.7 3 x 10 4 8.26 0.27 760 -83 + -89 3 x IO - 4 3.5 1 x 10~ 3 2.93 0 o a -20 -* -26 ~o 2.93 0 760 -17 -»• -23 =^ 0 a under N 2 atm. b i s the equilibrium constant for the equilibrium Ir(C0)ClP2(DEMA) s N [Ir(C0)P 2(DEMA)] + + C l " for calculations, see Appendix IV-3. - 104 - ' potentiometer H 2 gas Calomel electrode f i l l e d with 0.1 N Et 4NCl Ir(C0)ClP 2 + DEMA in solfulane at 80° bath aqueous 0.1 N Me.NCI 4 Fig. IV-17. Apparatus set up for determination of [Cl~] and acid-base tit r a t i o n in sulfolane. (a) specific chloride electrode for [Cl~] determination; (b) glass electrode for acid-base tit r a t i o n . - 105 -mV -150 -100 -50 0 [Cl"*],M Fig. IV-18. [Cl ] calibration curve in the potentiometric method. - 106 -Table IV-19. Conductance measurements in sulfolane. Temp. [Ir] M [substrate] M X(equivalent conductance) , -1 2 -1 ohm cm mole 35 80 35 80 35 80 80 35 80 35 80 80 80 80 80 9.93 x 10 4.85 x 10 9.93 x 10 4.85 x 10 4.83 x 10 4.85 x 10 — sulfolane : 4.0 x 10 - sulfolane : 1.2 x 10" [LiCl] = 7.55 x 10" -4 L i C l : 5.5 [LiCl] = 7.55 x 10" -4 Li C l : 13. [Ph 4AsCl] = 5.02 x 10" -4 Ph.AsCl : 4 13. [Ph 4AsCl] = 5.02 x 10" •H 1. Ph.AsCl : 4 28 -4 [PPh3] = 4.57 x 10" Ph3P : =^ 0 ) -4 under N 2 IrC0ClP 2 : 0.38 1 -4 under N 2 IrC0ClP 2 : 0.73 >-4 1 atm. H 2 Ir(C0)ClP 2(H 2) : 0.22 1 1 atm. H 2 DIr(C0)ClP 2(H 2) : 0.17 [MA] = 0.204 MA : 2.3 x 10" -4 [MA] = 0.204 CIr(CO)ClP 2(MA) : 25 [DEMA] = 0.218 DEMA : 4.6 x 10" -4 [DEMA] = 0.218 dIr(C0)C1P2(DEMA): 1.0 -5 a X X b *H 1000 1000 k < ~ — c c R 1000„ LK. for more than C 1 J one species where c = concentration. k = c e l l constant = 0.1315 R = resistance. K = m ir = m 2.2 x 10 3 M"1 at 80°C and hence [Ir(C0)C1P 2(H 2)] can be evaluated. [IrCOClP2(MA)] [IrCOClP ] [MA] ~ 3 ; a t 8 0 ° ' e s t i m a t e d f r o m Table IV-16, hence [Ir(CO)ClP2(DEMA)] [Ir(CO)ClP2][MA] [IrCOClP2(MA)] can be evaluated. =^ 0.3; at 80°, assuming K (DEMA system) ^ 10 1 m times < K (MA system)—see text—hence m [Ir(C0)ClP2(DEMA)] can be evaluated. - 107 -titr a t i o n set up is shown in Fig. IV-17. A typical acid-base ti t r a t i o n curve in sulfolane for pCH„C,H.S0_H vs. 0.1 N CH_0Na in CH„0H is shown 3 6 4 3 3 3 in Fig. IV-19. The sulfolane solutions of dihydride formed i n si t u in the absence and presence of DEMA were titrated vs. 0.1 N CR^ONa, and the t i t r a t i o n curves are shown in Fig. IV-19. IV.5 Discussion As mentioned i n Chapter III, preliminary kinetic studies on the hydrogenation of MA i n sulfolane catalyzed by Vaska's compound suggested that a quite different mechanism was operating from that established previously i n DMA and DMSO. The more detailed kinetics described i n this chapter for the sulfolane systems confirm this, although the kinetics proved to be much more complicated than expected. The UV/VIS and i . r . spectra (v ) of Ir(C0)ClP„ in sulfolane are CO z identical with those i n the sol i d state or i n benzene; there i s no change of the v i s i b l e spectrum on adding phosphine or chloride, and thus the compound exists as Ir(C0)CJ"J?2 i n the absence of or olefins. IV.5.1 Dihydride formation Absorption of H 2 occurs according to eq. (4,1) . Ir(C0)ClP 2 + H 2 s ~ Ir(C0)ClP 2(H 2) (4.1) and the equilibrium constant and thermodynamic parameters (AH = -13 Kcal mole \ AS = -22 e.u.) are very similar to those reported i n other solvents—benzene, toluene, chlorobenzene, DMF, DMA [AH = -12 to -14 mV I —_____ ,—i ______— 1 ; , i : i i 0 0.2 0.4 0.6 0.8 1.0 [CH3ONa] x 10 4, mole Fig. IV-19. Potentiometric acid-base titration in sulfolane at 80°; O [p.CH3C6H4S03H] = 0.64 x 10 _ 4 mole A [Ir(C0)ClP 2] » 0.68 x IO - 4 mole, [DEMA] = 1.9 x IO - 3 mole, p[H-] =760 mm Hg • [Ir(C0)ClP-] = 0.67 x IO - 4 mole, p[H_] = 760 mm Hg. - 109 -Kcal mole AS = -23 to -26 e.u. (5,6)]. The close resemblance of these parameters shows that reaction (4.1) is solvent independent, although the forward rate appears to increase somewhat with solvent polarity (6). It i s somewhat surprising that i n sulfolane the dihydride (formed in si t u , or by dissolution of the synthesized dihydride) shows only a single hydride stretch at 2080 cm 1 . A dehydrochlorination reaction yielding Ir(C0)P2(H) (7) seemed a po s s i b i l i t y , but the conductance measurements show that the conductivity of Ir(C0)ClP 2 under i s very small and very similar to that measured under N^; t i t r a t i o n of the hydride system with base also gave no evidence for dissociation of a proton. The CO stretch (1980 cm ^) i s in the same position as that measured for the dihydride, and the consistency of the values with varying H 2 pressure, both give further evidence for the formation of a dihydride i n sulfolane. The pos s i b i l i t y that the dihydride product contains trans hydrides cannot be completely ruled out. IV.5.2 Olefin Complexation The addition of MA or DEMA to Ir(C0)ClJ? 2 i n sulfolane results in an i n i t i a l rapid color fading of the solution. This change i s not affected by the presence of excess phosphine, and i s attributed to coordination of the o l e f i n : K 1 Ir(C0)ClP 2 + olefin , m » Ir(C0)ClP 2 (olefin) (4.3) The reasonably consistent K 1 values over a wide range of acid or ester m concentrations are in agreement with such an equilibrium. At 35°, K 1 is ^  16 M 1 for the acid system, and ^  1.5 M 1 for the ester system. - 110 -Strohmeier and Fleischmann (8) have reported some corresponding data in toluene at 20° giving K 1 ^ 350 M 1 for maleic anhydride, and <\< 4,5 M 1 for dimethyl maleate, while Vaska (9) has reported a value of 1.1 M 1 for the ethylene system in chlorobenzene at 30°. The upward shift of the carbonyl stretch on complexing with MA (Av = 7,0 c m is comparable to that reported for complexation with other olefins containing electron-withdrawing groups (9). Coordination via the carboxylate groups of the MA seems highly unlikely; this would involve proton release. Evidences against this are a) the consistency of the K 1 values calculated according to eq. (4.3); b) comparison with the binding data for the other non-acid olefin substrates; c) proton release would give rise to an immediate production of ions and an immediate increase in conductivity, which was not observed; d) the natural a f f i n i t y of a 'soft' iridium centre for the 'soft' o l e f i n i c ligand. After the i n i t i a l rapid complexing of MA according to eq. (4,3), a subsequent much slower reaction takes place at 35° as indicated by a further slow fading of the Ir(C0)ClP2 color. Evidence indicates that this reaction involves dissociation of chloride. The independence of this fading on the addition of excess phosphine rules out phosphine dissociation. An increase in conductance during the slow fading i s consistent with production of ions. The system also gave an immediate precipitation of AgCl on adding Ag*—no such precipitation is observed with the iridium complex in the absence of olefin; this certainly demonstrates an increased l a b i l i t y of the chloride in the olefin complex. - I l l -The dissociation constant (K^) for the chloride ionization (4.4) was determined from the v i s i b l e K l + Ir(CO)ClP2(MA) ^ ^ [Ir(CO)P2(MA)] + Cl (4.4) -5 -4 spectra data to be ^  5 x 10 M at 35°, and ^  5 x 10 M at 50°. Extrapolation to 80°, the usual hydrogenation temperature, suggests -2 that would be of the order of ^  10 M which means complete ioniza--1 2 tion, and the conductivity data at this temperature (X ^  25 ohm cm mole ^) are consistent with the presence of a 1:1 electrolyte (Table IV-19). The calculation of this X value requires a knowledge of K 1 at m 80° for reaction (4.3); this could not be obtained experimentally since the subsequent chloride dissociation i s more rapid than at 35°, and the i n i t i a l O.D. decrease to a steady value due to reaction (4,3) could 1 - 1 not be measured. An approximate value of K ^ 3 M was determined m by extrapolation of the data at 35° and 50°. Formation of the ol e f i n i c adduct i s seen to be an exothermic process; and such i s invariably found to be the case for such coordination addition reactions (9). The ionization constant at 35°, taken together with the approximately -4 -1 determined forward rate constant (y 10 sec ), shows that chloride w i l l react with the 4-coordinate iridium cation with a second order rate constant of about 2 M 1 sec ^ . 1 Cationic complexes such as Ir(C0)P 2(C 2H 4) + (10), I r ( C 0 ) L 2 ( C 2 H 4 ) 2 + and Ir(CO)L 2(diene) 2 + (L = PMe2Ph) (11) have been isolated from alcohol solutions and undoubtedly chloride dissociation is more li k e l y in polar solvents. - 112 -An attempt to further confirm chloride ionization was made by adding excess L i C l to the Ir(COJCIP^-MA solution; suppression of the ionization reaction following the olefin complexation reaction was anticipated. However, on such addition ([Cl ]:[Ir] = 2), the vis i b l e spectral change indicated formation of some completely different species; addition of Cl to the Ir(C0)ClP2 complex in the absence of added olefin gave no change whatsoever in the absorption spectrum. Thus chloride apparently reacts with five coordinate Ir (COJCH^ (MA) to give some other complex. One possibility i s that the chloride promotes dissociation of carboxylic protons to give an anionic f i v e -coordinate complex and pa r t i a l l y dissociated HC1 in this solvent; another poss i b i l i t y i s chloride substitution with loss of a phosphine, The effect of the added chloride was not persued however. The use of DEMA as o l e f i n i c substrate as opposed to MA, seemed worthwhile since any complications resulting from the presence of the carboxylic acid groups (proton dissociation) would be eliminated. The ester was f i r s t used in fact after the complex dependence of MA hydro-genation rate on the added acid concentration was observed (Fig. IV-11). As mentioned above, coordination of ester according to eq. (4.3) was found, with an equilibrium constant (K 1) about ten times smaller than m that measured for the acid. In the case of the ester, however, no subsequent slower color fading (attributable to an ionization reaction such as eq. (4.4)) was observed at 35°, Nevertheless chloride was immediately precipitated by Ag + from the Ir(COjCH^-ester system, again indicating the possible presence of small amounts of free chloride (or at least an increased chloride l a b i l i t y ) . Conductivity measurements - 113 -(Table IV-19) indicated some production of ions (which must be chloride) at 80°, and potentiometric data (Table IV-18) with the specific chloride electrode also indicated slight chloride dissociation (< 5%) for the ester system under hydrogenation condition at 80° (no chloride was 1 detected in the absence of the ester substrate). Assuming that the K m value i s s t i l l about ten times smaller than that for the acid system at 80°, i.e. using K 1 (DEMA) ^ 0.3 M _ 1 , the value of K, (DEMA) for the m 1 -3 -4 ionization is found to be of the order of 10 to 10 M at 80°. Because of the K 1 and K^ values for the DEMA system, the changes in v i s i b l e absorption spectra for the ionization reaction are not detectable at 35°. It is important to note that addition of L i C l did not affect the v i s i b l e spectra of the Ir(CO)CIP2-DEMA solutions; this i s in sharp contrast to the MA system and tends to suggest that the 'chloride reaction' in this system involves the carboxylic acid functions. IV.5.3 Kinetic Data In DMA and DMSO solvents, dissociation of a P group from an Ir(C0)ClP2 complex at some stage appears responsible for production of the catalytically active species, and oxidative-addition of H_ to a 4-coordinate Ir(CO)C1P(olefin) complex i s the rate determining step (1, 2, 12); addition of triphenylphosphine ([P]:[Ir] ~ 5) completely inhibited hydrogenation. Under corresponding conditions i n the sulfolane solvent, addition of P to the MA system resulted an inhibition but only to the extent of about 35% decrease in rate (Fig, IV-9); i n the DEMA system, the rate was more strongly inhibited to the extent of 80% (Fig. IV-10). - 114 -IV.5.3a MA Substrate The data presented in the previous section show that there is con-siderable dissociation of chloride from the iridium complex during the hydrogenation of MA, and due to the fact that added P does not strongly inhibit the rate, a hydrogenation path involving a cationic [Ir(C0)P2(MA)] complex seems l i k e l y to predominate. The molecular dihydride Ir(C0)C1P 2(H 2) i s rapidly formed in the sulfolane system but i t does not appear to be directly involved i n the main catalytic hydrogenation reaction. There is no P, proton or chloride dissociation and the iridium remains 6-coordinate at least in the absence of ol e f i n . A rate deter-mining step such as k fast Ir(CO)ClP 2H 2 + olefin Ir (CO) C I P ^ (olefin) *-Ir(CO)ClP 2 + product (4.5) would give an overall first-order dependence on Ir (which is not observed) and would not account for observed i n i t i a l autocatalytic shape of the uptake curve. A process in which chloride loss occurs slowly, for example as the i n i t i a l step of the following processes, k l + Ir(CO)ClP 2H 2 ^ •*• » Ir(CO)P 2H 2 + Cl k =^7 ^ -^r^ ^ ^ olefin ^ , (scheme 4.6) fc2 Ir(CO)P 2 + + product-* [Ir(CO)P 2H 2 (olefin) ] + would account for the autocatalytic region, the less than first-order in Ir, and the presence of ionic chloride. However, such a mechanism - 115 -via this 'dihydride route' w i l l not account for the inverse dependence of H,, at higher pressures, and this i s most readily accounted for by production of an inactive Ir(CO)ClP 2H 2 complex. Further, since olefin coordination to Ir(CO)ClP 2 and a following chloride ionization was substantiated, hydrogenation by an unsaturate route involving mainly Ir(CO)P 2(MA) + is strongly favoured. Cationic iridium complexes such as Ir(diene)P 2 + have been reported as effective catalysts for hydrogenation of olefins (13). The following scheme accounts at least qualitatively for the kinetic and spectroscopic data of the MA system: Ir(CO)ClP 2 + H 2 ^ n ^ Ir(CO)ClP 2(H 2) + MA Km <V k l + Ir(CO)ClP2(MA) K k ^ Ir(CO)P2(MA) + Cl * ^ " " k 2 | H 2 IrCOP 2 + +"product < [Ir(CO)P 2(MA)H 2] + (scheme 4.7) MA coordinates to the Ir(CO)ClP 2 complex in a rapid equilibrium to form an olefin complex with a subsequent slow dissociation of chloride. This slowly established equilibrium accounts for the i n i t i a l slower hydrogen absorption in the uptake plot (Fig. IV-1). The chloride dissociation from the olefin complex provides a 'vacant' coordination s i t e (presumably occupied by a solvent molecule) for the subsequent rate determining attack by the H 2 molecule. The slow step l i k e l y involves oxidative-addition of hydrogen to give a dihydride; subsequent olefin insertion - 116 -into the metal hydride would yield an alkyl-hydride which could then reductively eliminate the saturated product with regeneration of Ir(CO)P 2 + cationic catalyst. The rate law for the above.scheme is given by -d[H ] rate = ^ - = k ^ ] [Ir(CO)P2(MA) ] (4.8) k ^ K * [H 2][MA][Ir] T ° r r a t e = ( k ^ t C l " ] + k 2 [ H 2 ] ) ( l + K^MA] + Kj j l ^ ] ) + k^tMA] (4.9) (for derivation, see Appendix IV-4). The rate law shows between zero and f i r s t order in both MA and total iridium, as observed. The less than f i r s t order in Ir arises because of the chloride dissociation equilibrium; in the limit at high total Ir concentration, a half-order dependence in catalyst can be expected, and indeed the data of Table IV-1 show a half-order dependence on Ir at the higher concentrations. The complex hydrogen dependence (Fig. IV-5) may also be explained qualitatively by rate-law (4.9); at very low [H2], k ^[Cl ] may be > k 2[H 2], which means that the chloride dissociation (K^) is established rapidly compared to the k 2 step, and the rate-law takes the limiting form. k k / [ H ][MA][IrL 1 z m I 1 / / i r» \ rate = (4.10) [Cl -](1 + K^[MA] + Kjjf^]) + kjK^lMA] The H 2 dependence w i l l then be between f i r s t and zero order, depending on the relative magnitude of K^[MA] and K^t*^] (since at 80° K 1 is very approximately 3 M-"1, and 1^ ^  2 x 10 3 M - 1, the K 1 [MA] and ^ [ H ] terms w i l l be of comparable magnitude), and the rate w i l l increase with - 117 -increasing H 2 concentration. At higher [H 2], k 2[H 2] may become > k_ 1[Cl ], and in the limiting case chloride dissociation becomes rate determining; the rate-law takes the form rate = k 1 K 1 [ H 2 ] [ M A ] [ I r ] T (4.11) k 2.[H 2](l + Km[MA] + B y i y ) + k^EMA] which can clearly give an inverse dependence on H2- Presumably the curve shown in Fig. IV-5 can then be represented by a rate equation of the type shown in eq. (4.9). A reasonable test of this explanation for the H2~dependence would l i e in examination of the chloride dependence of the rate, which at low [H 2] should be [Cl ] 1 (eq. 4.10), but should be independent at higher [H2] (eq. 4.11). Unfortunately added chloride gives a quite different complex in solution (see IV.5.2), and, judging by the almost complete inhibition of the hydrogenation rate at 1:1 added chloride, this complex must be completely inactive as a hydrogenation catalyst. It is worth noting that LiNO^ also inhibits hydrogenation although somewhat less markedly than L i C l (Fig. IV-13). Some interaction with the carboxylic group of the coordinated MA appears l i k e l y : -C00H + LiX - -COOLi + HX -COO~Li+ (4.12) The addition of salts would also give rise to primary salt effects, particularly, for example, in the ionization equilibrium and i t is d i f f i c u l t to estimate the importance of these. - 118 -It should be noted that an inverse dependence has also been observed for the hydrogenation of ethylene catalyzed by Ir(CO)P.jH in DMF (14), and this again results from 'tying up' available catalyst as an inactive 6-coordinate species Ir(CO)P.jH3. The addition of strong acid to the MA system ([H +]:[Ir] ~ 10) produces a significant doubling of the rate, which then decreases somewhat with further increase in added acid (Fig. IV-11). One can only speculate on the effects that give rise to this acid dependence. It should be noted that the rate of hydrogenation in DMA, showed only a slight decrease on the addition of strong acid (Table IV-20), and thus the acid dependence is probably a function of the use of the sulfolane solvent (the same appears to be true for hydrogenation of the DEMA substrate—see l a t e r ) . In view of the fact that the 'special role' of the sulfolane apparently involves chloride dissociation, i t seems feasible that added proton might enhance chloride dissociation, i.e. favour the equilibrium and hence increase the hydrogenation rate. Electrolytes such as LiNO^ form ion-pairs to some extent at concentra--2 tions of ^  10 M in sulfolane (15), and protons could enhance Cl dissociation i f HC1 remains undissociated to some extent. Salt effects could account for the less significant rate decrease at higher added acid. The salt effect due to added DCH_C,H.SO.Na could not be studied 3 6 4 3 due: to i t s complete insolubility i n sulfolane (addition of NaClO^ was found to give sl i g h t l y lower rates for the DEMA substrate). The enhancement of the MA hydrogenation rate by added acid and the previously discussed inhibition by added chloride, are both rationalzied in terms of equilibrium (4.13) and the product of partly associated HC1 - 119 -Table IV-20. Kinetic data for the Ir(CO)ClP 2 catalyzed hydrogenation in DMA. [Ir] = 5.36 x 10~ 3 M, p[H 2] = 730 mm Hg, T = 80° substrate [substrate] [pCH^H^SO^] hydrogenation rate M M x 10 6 M sec" 1 MA 0.063 0 1.25 0.061 0.024 1.0 DEMA 0.273 0 3.0 0.273 0.096 3.2 - 120 -i f + Cl ^ ^ HC1 (4.13) There was no evidence for proton dissociation from a dihydride species alone, or in the presence of the DEMA substrate, and any acid effect due to such proton dissociation equilibria i s considered unlikely. The 35% maximum rate decrease due to added phosphine could imply that some hydrogenation occurs via a path resulting from phosphine dissociation at some stage, i.e. via a neutral Ir(CO)C1P(MA) + path (see section III.4). Alternatively P could compete for a coordination site on the cationic Ir(C0)P2(MA) + complex, which would inhibit hydrogenation via the Ir(C0)P MA+ + H_ path. 2 It is d i f f i c u l t to attach much significance to the apparent activation energy of 30 Kcal mole \ since this must refer to a 'composite' rate constant. IV.5.3b DEMA Substrate The kinetics of the hydrogenation of this substrate in sulfolane are in general quite similar to those observed for the MA hydrogenation in DMA and DMSO. The substrate dependence and the hydrogen dependence are both 'normal', and are f i r s t order at low concentration and become less than f i r s t order with increasing concentration. Further, the hydrogenation rate is inhibited to a large extent (80%) on addition of phosphine, and i s essentially independent of added chloride. The data thus indicate that the major pathway involves reaction of R^, with a neutral Ir(CO)CIP(DEMA) complex. The olefin complexation data are consistent with the following reaction scheme: - 121 -Ir(CO)ClP 2 + H 2 s n > Ir(CO)ClP 2(H 2) + DEMA Ir(CO)ClP2(DEMA) 1 ^ Ir(CO)C1P(DEMA) + P + H2 (scheme 4.14) Ir(C0)ClP + product Ir(CO)ClP(DEMA)H2 The derived rate-law is entirely analogous to that given i n eq. (4.9) but with the k ^  [Cl ] term in the denominator replaced by k ^  [P], i.e. k 1 k ^ [ H 2 ] » ] [ I r ] T rate = : z (k n[P] + k_[H_])(l + KHDEMA] + K j H ] + k r [ » ] -1 2 2 m U 2 1 n^MAl (4.15) The plot of rate vs. DEMA concentration (Fig. IV-4) shows much less curvature than the corresponding plot for the acid system (Fig. IV-3), and this i s consistent with the K 1 value for the DEMA being an order m of magnitude lower than that for the MA (i.e. the contribution of the K^[olefin] term i n the denominator term of eq. (4.15) is much less for m the ester system). In contrast to the acid system, the hydrogen dependence (Fig. IV-6) does not become inverse at the higher pressures; this implies that the k ^[P] term i s greater than the k 2[H 2] term i n eq. (4.15)—the rate determining step is then the reaction of Ir(CO)C1P(DEMA) with E^ even at higher E^ pressures. This contrasts with the acid system where a k^ dissociation step (of chloride) becomes rate-determining. Again, - 122 -i f k^tH^] < k_-^[P]> the rate should show an almost direct inverse dependence on added P, but the data in Table IV-6 indicate that this i s not so (particularly at the lowest concentration of added phosphine). A possible rationalization for this can be found on consideration that part of the hydrogenation occurs via the chloride dissociation path. Slight chloride dissociation from the IrC0ClP2(DEMA) complex was observed, and the hydrogenation rate is retarded by about 20% (Table IV-12) on adding excess chloride. In terms of the equilibria, IrCOP2(DEMA)+ + C l " ^  IrC0ClP 2 (DEMA) IrCOClP (DEMA) + P I II (4,16) addition of P w i l l inhibit hydrogenation via II but w i l l enhance hydro-genation via I, and because of the two hydrogenation paths, a simple direct inverse dependence on P would not be expected. The data suggest that about 80% of the hydrogenation goes via II and 20% via I—an inverse phosphine dependence results because the major pathway i s via II. The acid dependence data (Fig. IV-12) are markedly different from those for the MA hydrogenation: the rate increases throughout with increasing added acid, and the enhancement of rate is very much greater (a factor of 'v 35 compared to 2 for the MA system). In DMA, addition of acid has essentially no effect on the ester hydrogenation rate (Table IV-20), so again the acid dependence behavior results solely from use of the sulfolane. As for the MA system the rate increase i s possibly attributable to enhanced chloride dissociation due to equilibrium (4.13), and hence promotion of the path catalyzed by [IrC0P9(DEMA)] , - 123 -It is readily shown that the acid dependence should be between f i r s t and zero order, as observed. The much greater acid effect in the DEMA system is consistent with the much smaller degree of chloride dissociation of this substrate. Addition of tributylamine in the DEMA system (Table IV-13) has relatively l i t t l e effect which shows that a proton dissocia-tion equilibrium involving a hydride species is not the cause of the effect of added acid. Addition of the weaker phenolic acid, p i c r i c acid, caused only a very slight increase i n hydrogenation rate. IV.5.4 General Conclusions Although the hydrogenation of MA and the DEMA in sulfolane were investigated quite exhaustively, i t has not proved possible to quanti-tatively analyze the kinetic data. The reactions are much more complicated than corresponding ones studied in DMA and DMSO. The main features are recognisable however. A cationic olefin species produced by loss of chloride is the major catalyst for the MA reduction, and oxidative-addition of to the cation is thought to be the rate-determining step. The possibility of some reaction occurring to a lesser extent via a corresponding path involving phosphine instead of chloride dissociation cannot be ruled out. For hydrogenation of the DEMA, the major pathway involves a phosphine dissociation step while the cationic pathway occurs i to a lesser extent. Ionization of chloride from the 5-coordinate IrC0Cl[PMe2Ph]3 complex has recently been discussed by Chen et a l . (16) who attribute weakening of the Ir-Cl bond to a build up of electron density at the Ir-centre by the 3 strongly basic phosphines. The present system is thought to involve a 5-coordinate Ir(CO)C1P,(olefin) - 124 -complex. The MA binding constant is ten times that of the DEMA and i t is the acid system in which the chloride dissociates more readily. Electronic arguments are, however, somewhat complicated due to the 'synergistic' bonding effect of the TT-acid olefin ligands (17). Of the substrates tested (Table IV-14) only the olefins 'activated' by electronic withdrawing groups were successfully hydrogenated. This presumably results from the much higher binding constants for such olefins to the Ir(I) centre. Other factors also play a role however. For example, such activated olefins w i l l make the subsequent rate determining dihydride formation more d i f f i c u l t due to an increased electronic promotion energy (18). It would seem that the olefin s t a b i l i t y constant has to be sufficiently large for a reasonable hydrogenation rate ( c f , in eq. (4.9)) but not too large since w i l l be l i k e l y decreased. The trans-cinnamic acid l i k e l y binds weakly to the Ir because of steric reasons. Acetylenes were not hydrogenated, although Strohmeier and coworkers (19) have reported that phenylacetylene is hydrogenated as effectively as dimethylmaleate using Ir(C0)ClP2 in toluene. Complexes with the iridium are almost certainly formed (20), although this aspect was not studied in the present work. It seems l i k e l y that the acetylene complexes which are generally more stable than corresponding o l e f i n i c complexes, do not activate (k 2 is small). Compared to the previously reported (Chapter III) iridium-catalyzed hydrogenation in benzene, toluene, DMA, DMSO and DMF, the sulfolane system is more active under corresponding conditions. The system in - 125 -each case involves a PPh^ ligand dissociation at some stage to form a square-planar iridium o l e f i n complex. Only in the case of sulfolane has chloride dissociation been observed, and this gives r i s e to the enhanced activity via the resulting cationic species. Even i n the 'unique' sulfolane system, the chloride-dissociation path becomes important only for certain o l e f i n i c substrates, for example, as i n the MA system compared to the DEMA system. The d i e l e c t r i c constant of sulfolane i s very similar to those of DMA and DMSO, but the coordinating power of the sulfolane is very much lower (Table III-9). Gutmann (21) has discussed the ionization of halides i n non-aqueous solvents, and concludes reasonably that such a reaction is more l i k e l y i n the solvent of higher donor number—an apparent contrast with the conclusions for these hydrogenation studies. The donor numbers have been measured relatively to acids such as SbCl,., 1^ and phenol (21), and not transition metals, although there are other data (22) showing that sulfolane forms relatively weak complexes (possibly chelates with bonding through both oxygens) with transition metal centres such as Co(II). Anion solvation must also be considered when discussing an ionization reaction; hydrogen-bonding solvents may undergo specific interaction with the anion," thereby promoting ionization. However, such specific interactions would seem no more l i k e l y in sulfolane than for i . • example in DMSO. The factors governing the chloride vs. phosphine dissociation are clearly quite subtle, but must involve to some extent here the differences arising from a -COOH as opposed to a -COOEt substituted on the complexed - 126 -olefin. Sulfolane is anomalous compared with other sulfur-containing solvents such as sulfites and sulfoxides in that i t does not follow a general linear relationship between the enthalpy of adduct formation with phenol (an acceptor acid) and the phenol frequency shift Av , OH and this was attributed to interaction of both sulfolane oxygens with the phenol proton (23): Similar interaction with the protons of the MA carboxylic groups is a possibility and presumably such electronic interactions (donation to the olefin complex) could lead to labilation of chloride attached to the iridium. - 127 -APPENDIX IV-I K 1 v IrCOClP 2 + MA ^ m > IrCOClP2(MA) 1 ^ [IrCOP 2(MA)] + + Cl" fast t = o, [IrCOClP-] = a , absorption = A at 387 nm 2 o o On addition of MA, the rapid equilibrium K 1 IrCOClP 2 + MA ^ m ^ IrCOClP2(MA) a l i s established in a very short time. (Fig. IV-14) t 'v, 0; [IrC0ClP 2] = a ^ absorption = A± ± [IrC0ClP 2(MA)] (A q - A.) Km = [IrCOClP 2][MA] = A1 x [MA] ( 1 ) Hence, K 1 can be evaluated, m After about another 3 hours, the chloride dissociation equilibrium K 1 K. IrC0ClP 2 + MA ^ m ^ IrCOClP2(MA) , 1 * IrC0P2(MA) + Cl~ ' a„ b c is established. On assuming that £[Ir(C0)C1P2(MA)] and e[Ir(C0)P 2(MA)] + are zero at 387 nm, t = °°, [IrCOClP.] = a , absorption = A - 128 -K [IrCOP2(MA)+][Cl"] 2 1 = [IrCOClP2(MA)] = b~ and K 1 = - — i f [MA] » a a [MA] then a = a + b + c o °° a m + K^a [MA] + C b 00 m ooL j 2. a r o + K^a [MA] + A1 a [MA] .£7 °° in 0 0 m 0 0 1 [a - a (1 + K^MA])] 2 o °° m K^cofMAl (2) Since a , a and [MA] are known, and values of K 1 can be obtained from ° m eqn. (1), the value of can be evaluated from eqn, (2), - 129 -APPENDIX IV-2 k k IrCOClP + MA ^ 2 ^ IrCOClP (MA) ^ 1 N IrCOP (MA)+ + C l " k If -2 -1 (A) (B) (C) K K, m 1 fast slow ^51 = k 2[A]b - k_ 2[B] - k x[B] + k ^ t C ] 2 (1) b = [MA] and [MA] » [A] A o i f 1 - JH- . [B] = K*b[A] ^ Km " [A]b ' ' ™ d[B] = ITbdfA] m substitute into (1) K^b = k 2[A]b - k_ 2 Kmb[A] - k x KH> [A] + k ^ t C ] 2 (2) Since K 1- ^ -m k_ 2 , (2) would become H A L = _,, r A l r ^ i 2 dt •k [A] + [C]' (3) K m a = i n i t i a l total IrCOClP- concentration o 2 = [A] + [B] + [C] = [A] + KH> [A] + [C] m .*. [C] = a - ([A] + KVA]) = a - [A] (1 + K^b) o m o m substitute into (3) - 130 -d l A l = dt k - i r -k [A] + - r ^ K b L m a - [A] (1 + K o let x = [A] dx dt = -k, x + \a - x ( l + K1!*) 1 x + K b m k (1 + K ^ ) 2 —J. m KXb m a 1 + K 1b m - x k n (1 + K^ ) ' ~ - i m K^ ( 2 a k (1 + K^b) 2 ^  -1 m K^ m K k (1 + K^b): -1 m ( i + nh)2 m KXb m k , (1 + K-hj) 2  -1 m m -! 2 a k ,(1 + K^b) ) a 2 k , , , o -1 m f , o -1 k. + ; > x + K m KXb m 2 , _ . = ax + Bx + Y where a -' k - l <1 + K m b > 2 m (4) 3 - - { k x 2 % k - l ( 1 + K m b ) KH» m a 2 v 0 -1 m - 131 -eqn. (4) can be rewritten as d F = K ( X + 2o" ) 4a + Y 2 2 = aX - co where X = x + — ; co = 2a Y) dX = dx dX _ v2 2 dX — - a X - co or — aX - co = d t (5) The general solution of (5) i s : t = t t=o 1 X - co// In 2/aco X + co //a t = t t=o Since B = m 7 (6) 4a [ k ^ b + 2a k . (1 + KH)]^ 1 m o - l m m m 4 k . (1 + K"H>)' —1 m As k -1 K, £2 4a K K b + 2 a (1 + K m l 0 4K 1K 1b(l + KH)2 m l m - 132 -- Y L m i b + 2a (1 + K \ ) o m ' J ^ K b d + K^b) 2 m l m o 1 K 1^ b m ± r%.\b + 2a (1 + K^b) "I L m l o_ m J 4K K b ( l + ^ b ) 2 m 1 m 2, a k., o 1 K^K-b m 1 = (1 + KH) m K ^ b m 1 x + 3_ 2a = x KlVb + 2a (1 + K-SJ) m 1 o_ m 2(1 + KH)2 m - 133 -APPENDIX IV-3 Calculation of K for DEMA system Ir(CO)ClP 2 + H 2 ^ ^ Ir(CO)ClP 2(H 2) (I) + DEMA (ID K m Ir(CO)ClP2(DEMA) ( I l l ) - [Ir(CO)P 2(DEMA)] + + Cl" (IV) + H Ir(CO)P 2 + product <r Ir(CO)P2(DEMA) (H2) [Ir] = [I] + [II] + [III] + [IV] [IV]' K K. [DEMA] m l K^[H ] [ I V ] 2 [IV] 2 . + -z-^ + + [IV] K K"[DEMA] x m Since: [ I r ] ^ , Kg, are known values K 1 is estimated to be ^ 0.3 M _ 1 m [IV] could be obtained from Table IV-18. Hence, K^ could be evaluated. - 134 -APPENDIX IV-4 IrCOClP 2 + H 2 -(I) + _• IrCOClP 2(H 2) (III) MA /II K m IrCOClP2(MA) v (II) \ (iv) : = i IrCOP„(MA) + + Cl" k-1 _ =7 2 H, IrCOP2 + S.A. <r [IrCOP2(MA)H2]" rate = d[H 2] = k-[H 2][IV] Apply the steady state theory = k_[H] - k^ECl-JtlV] - k 2[H_][IV] k_[H] = [IVlCk^ICl"] +k 2[H 2]) or [IV] = k ^ I I ] k ^ t C l ' ] + k 2[H 2] put (2) into (1) rate = k - k ^ i y i l l ] k ^ f C l " ] + k 2[H 2] - 135 -[ I r ] T [I] + [II] + [HI] + [IV] [II] \[H] + [II] + K^[I][H ] +• K^fMA] " * k ^ l C l ] + k 2[H 2] [II] V I I ] [H ] k [II] [Ir] = - j + [II] + - f + — (4) OMA] K^[MA] k .[Cl"] + k-[Hj m m —1 2. 2. , [HI [HI] for K = ; Kg = m [I][MA] " [H 2][I] (4) becomes W . k l [Ir,] = [II] -= + 1 + " ' + *• 1 t/ r w A i t-*-OMA] K"[MA] k [Cl ] + k.[H ] m m —1 2. 2 K^[MA][Ir] T(k .[Cl ] + k-[H.]) [II] = — T _ i 1 1 (1 + K*[MA] + V V ) ( k - l t C 1 ~ ] + k 2[H 2]) + k^lMA] put (5) into (3) ( 5 ) k.k 9Kj;[H-][MA][Ir] T . 1 2 m I T •  (k ^ f C l " ] + k 2[H 2]) (1 + K*[MA] + K__[H2]) + '^[MA] (6) - 136 -References 1. B.R. James and N.A. Memon, Can. J. Chem., 46^ , 217 (1968). 2. M.G. Burnett, R.J. Morrison and C.J. Strugnell, J. Chem. Soc, Dalton, 701 (1973). 3. L. Vaska and J.W. Diluzio, J. Am. Chem. Soc, 84_, 679 (1962). 4. D.H. Morman and G.A. Harlow, Anal. Chem., 39, 1869 (1967). 5. L. Vaska and M.F. Werneke, Trans. N, Y. Acad. Sci., 31, 70 (1971), 6. B.R. James, Homogeneous Hydrogenation, Wiley-Interscience, New York, 1973, p. 293. 7. J.F. Harrod, D.F.R. Gilson and R. Charles, Can. J, Chem., 47, 1431 (1969). 8. W. Strohmeier and R. Fleischmann, Z. Naturforsch., 24b, 1217 (1969). 9. L. Vaska, Acc. Chem. Res., _1, 335 (1968). 10. L. Vaska, Inorg. Chim. Acta, .5, 295 (1971). 11. A.J. Deeming and B.L. Shaw, J. Chem. Soc, A, 376 (1971). 12. W. Strohmeier and T. Onoda, Z. Naturforsch., 24b, 1493 (1969), 13. J.R. Shapley, R.R. Schrock and J.A. Osborn, J. Am. Chem. Soc, 91, 2816 (1969); M. Green, R.A. Kuc, and S.H. Taylor, Chem. Comm., 1553 (1970). 14. M.G. Burnett and R.J. Morrison, J. Chem. Soc, A, 2325 (1971), 15. R.C. Burwell and CH. Langford, J. Am. Chem. Soc, 81, 3799 (1959). 16. J.Y. Chen, J. Halpern and J. Molin-Case, J. of Coord. Chem., 2^, 239 (1973). 17. J.P. Collman and J.W. Kang, J. Am. Chem. Soc, 8£, 844 (1967). - 137 -18. J.A. Osborn, F.H. Jardine, J.F. Young and G. Wilkinson, J. Chem. Soc, A, 1711 (1966). 19. W. Strohmeier, W. Rehder-Stirnweiss and R. Fleischmann, Z. Naturforsch., 25b, 1481 (1970). 20. L. Vaska and R.E. Rhodes, J. Am. Chem. Soc, 8J_, 4970 (1965). 21. V. Gutmann, New Pathways in Inorg. Chem., Cambridge University Press, 1968. 22. CH. Langford and P.O. Langford, Inorg. Chem., 1, 184 (1962). 23. R.S. Drago, B. Wayland and R.L. Carlson, J . Am. Chem. Soc, 85, 3125 (1963). CHAPTER V CATALYTIC HYDROGENATION REACTIONS USING THE Ir(I) CYCLOOCTENE COMPLEX [ I r C l ( C 8 H 1 4 ) 2 ] 2 (1) IN DMA V.l Introduction A number of Ir(I) complexes are known to be active catalysts for hydrogenation of olefins (see Section 1.3). Following the studies in our laboratory (1) on the activation of hydrogen by the Rh(I) cyclooctene complex, [RhCl(CgH^) 2] 2, we decided to study the system involving the corresponding Ir complex (2). For comparison with the previously studied Rh system i n DMA solution, some kinetics of olefin hydrogenation for the Ir system were studied in the DMA solvent. Monoolefins and molecular oxygen were found to be catalytically reduced. The kinetic data for such systems and mechanistic interpretation are presented below. V.2. Catalyzed Hydrogenation Using [ I r C l ^ R ^ ) ^ in DMA V.2.1 Reaction of 1 with H 2 In DMA, complex _1 absorbs H 2 very quickly, at 25°, to a mole ratio of [H 2]/[Ir] = 2.5 + 0.2 ([Ir] = Ir(I) monomer = 2.09 x 10~ 2 M). The H 2 uptake plot is shown in Fig. V - l . On addition of a large excess of L i C l (0.2 M) the same ratio of H 9 absorption i s observed, (Fig. V - l ) . - 139 -- 140 -In both cases 5a black suspension of Ir-metal was observed after the reaction was completed. These experiments indicate that the complex is very vulnerable to H2 reduction, and that the presence of Cl does not stabilize the complex against reduction to the metal. The unsaturated cyclooctene ligands are believed to be hydrogenated to cyclooctane, and the Ir(I) i s reducted to Ir-metal: 2 5 H l / 2 [ I r C l ( C 8 H u ) 2 ] 2 -^h lr(0) + 2 CgH^ (5.1) The Ir-metal produced was found to be a very powerful heterogeneous hydrogenation catalyst. For example, hexene-1 and cyclooctene were readily hydrogenated by suspensions of the metal i n DMA at ambient conditions (1 atm H 2 at 25°). Benzene solutions of the complex rapidly absorb H2, again with metal production, but the uptake continues well beyond the 2.5:1 mole ratio, and this i s believed to result from hydrogenation of the benzene. In the presence of monoolefins, e.g. hexene-1, cyclooctene, maleic acid, fumaric acid, diethyl maleate, DMA solutions of 1 absorb H 2 at ambient conditions and homogeneous catalytic hydrogenation to the corresponding saturated product occurs. Metal was not formed u n t i l completion of the olefin hydrogenation, indicating that the olefins must coordinate and stabilize Ir(I) against hydrogen-reduction to the metal. - 141 -V.2.2 Catalyzed Hydrogenation of Hexene-1 In vacuum or under N^, an equilibrated solution of the complex _1 in 2 DMA gives absorption peaks at 385 nm (e = 6.8 x 10 ), 414 nm (e = 7.3 x 10 2) and 460 nm (e = 3.7 x 10 2), Fig. V-2. On addition of a 50-fold excess of hexene-1 at 25°, the change of absorption is very slow and small, Fig. V-2, indicating that any coordination of hexene-1 to the complex is relatively weak. Absorption spectrum of 1 i n solid state is shown in Fig. VI-6. A typical uptake plot for the hydrogenation of hexene-1 in DMA at 25° using 1_ i s shown in Fig. V-3. An i n i t i a l autocatalytic type region was observed i n the gas uptake plot; this region was then followed by an extensive linear region. The rate f i n a l l y started to f a l l off at the region where the ratio of ^rhexene-l was close to 1:1, and metal precipitation began to occur. The slope of the linear region i n the H^-uptake plot was taken as the maximum rate of hydrogenation of hexene-1, and was used to investigate the kinetic dependences. The data on the variation of hydrogenation rate with respect to concentration of hexene-1, H 2» and the catalyst, are summarized i n Table V - l . The hydrogenation rate i s first-order i n H_ concentration -3 up to 1.7 x 10 M (1 atm H2) (Fig. V-4), and i s f i r s t order i n catalyst -3 concentration up to 3.5 x 10 M, (Fig. V-5). The hydrogenation rate is I first-order i n hexene-1 at lower concentrations, becomes less than f i r s t -order at higher concentrations, and tended to zero order at about 0.1-0.2 M, (Fig. V-6). As the hexene-1 concentration goes higher than 0.2 M the reaction i s inhibited; the uptake plots are of the type shown i n Fig. V-7, and no linear region i s observed. O.D. 350 400 50CT 600 wavelength, nm Fig. V-2. Visible absorption spectra of [ I r C l C C g H ^ ) ^ recorded in DMA at 25°; 10 mm c e l l (A) [Ir] - 1.58 x 10" 3 M; (B) [Ir] = 1.58 x 10~3 M; [hexene-1] - 0.08 M. - 143 -- 144 Table V - l . [ I r C l C C g H ^ ) ^ Catalyzed hydrogenation of hexene-1:kinetic data at 25°C in DMA. [Ir] [hexene-1] p[H 2] [H2] linear rate 1 0 ' M M mm Hg x 103,M x 105,M sec" 1.12 0 760 1.7 0 1.12 0.158 760 1.7 5.2 1.12 0.081 760 1.7 4.4 1.12 0.032 760 1.7 2.7 1.12 0.016 760 1.7 1.6 1.12 0.081 412 0.92 2.8 1.12 0.081 224 0.50 0.89 1.12 0.081 103 0.23 0.38 1.12 0.081 0 0 0 3.32 0.081 760 1.7 9.3 2.37 0.081 760 1.7 6.5 1.12 0.081 760 1.7 4.43 0.62; 0.081 760 1.7 1.6 0.33 0.081 760 1.7 0.36 0 0.081 760 1.7 0 [Ir] x 10", M V-5. Dependence of linear hydrogenation rates of [Ir] concentration i n DMA at 25" ([hexene-1] - 0.081 M; p[H23 = 760 mm Hg). 10 CM O 01 u o M cfl 6 CM / [hexene-1] = 0.323 M O / / O P / / r p Ay Cr 0 8*. -3 time x 10 . sec [hexene-1] = 0.485 M Fig. V-7. H 2 uptake plot of [IrCltCgH.^)^ i n DMA at 25° with high concentration of olefin ,-3 ([Ir] - 1.12 x 10 M; p[H-] - 760 mm Hg). - 149 -The data for the rate dependence on temperature from 21° to 31°C are given in Table V-2 for the conditions noted. Differences i n solubility over the temperature range are very small (16) and were neglected. No effect on the hexene-1 hydrogenation rate was observed on addition of an excess of a strong acid (p_-toluene sulfonic acid), even up to the mole ratio of [H +]:[Ir] = 15:1. Conductance measurements (Table V-3), showed that there were practically no ionic species i n the solution during hydrogenation. n-Hexane was identified as the hydrogenation product by gas chromatography (column = AgNO^-triethylene glycol, temperature = 40°C, current = 2 250 mA, carrier gas = He, with pressure = 10 l b / i n , retention time = 10.2 min). In the presence of added Cl (LiCl) and of olefin concentration at conditions when the rate i s essentially independent of olefin concentra-tion, the reaction i s inhibited, The H^ gas uptake plots at various added Cl concentrations are shown in Fig. V-8, The uptake, plots are similar i n shape to those measured in the absence of added Cl (Fig, V-3) but the rates f a l l off rapidly. V.2.3 Selective Hydrogenation of a Mixture of Hexene-1 and  Cyclooctene In view of the fact that the starting catalyst complex contains coordinated cyclooctene, a study was made of the possible catalytic hydrogenation of this olefin. Indeed, cyclooctene i s catalytically hydrogenated by 1^  in DMA to cyclooctane, and at a rate independent of the concentration of cyclooctene i n the range of 0.03-0.16 M. A typical - 150 -Table V-2. Ir Catalyzed hydrogenation of hexene-1 in DMA. Temperature dependence of linear rate. [Ir] = 1.12 x IO - 3 M; [hexene-1] = 0.081 M Temperature [H2] mm Hg* linear rate x 10^, M sec 21.2 760 2.62 25 760 4.43 27 760 5.24 31 760 5.90 * H 2 solubility taken as -3 -1 1.7 x 10 M atm over the 21-31° temperature range. - 151 -Table V-3. Conductance measurements in DMA at 25°. c e l l constant = 0.1316 Reagents added M Resistance (ohm) Specific conductance ohm cm Equivalent conductance ohm cm equiv. [Ir] = 1.12 x 10 -3 -3 [Ir] = 1.12 x 10 [hexene-1] =0.08 -3 [Ir] = 1.12 x 10 [hexene-1] = 0.08 NaN0_ 2.7 x 10 4 a 2.7 x 10 4 a 2.8 x 10 4 a ^2.2 x 10 -4 b 4.9 x 10 ^0 = 0 1.9 x 10 -6 c -6 1.7 " 72 (17) under atm under EL, atm and.in a period of about 2000 sec C reported (17) to be 0.8-2.0 x 10~ 7 ohnf 1cm~ 1 based on monomeric Ir species - 152 -- 153 -H^-uptake plot i s shown in Fig. V-3. No autocatalytic period was observed i n the gas uptake plot, and the linear rates at 25° are summarized in Table V-4. The rate i s about 10 times slower than that of hexene-1 at a substrate concentration of 0.08 M. At completion of the hydrogenation, Ir-metal precipitation was observed. A gas uptake plot using a 1:1 substrate mixture of hexene-1 and cyclooctene (0.08 M i n each) is shown in Fig. V-9. The cyclooctene was selectively reduced f i r s t with i t s individual rate of hydrogenation (4.2 x 10~6 M sec" 1 at an [Ir] = 1.12 x 10~ 3 M, cf. Table V-4). After about 80% of the cyclooctene was selectively hydrogenated, the increased rate of hydrogenation signaled the start of hexene-1 hydrogenation. V.2.4 Catalytic Hydrogenation of Other Monoolefins Uptake plots for a number of olefin substrates were similar to those observed for the hexene-1 hydrogenation (Fig. V-3). Linear rates were again measured. The reduction rate of hexene-2 was about 0.3 times that of hexene-1 under the same conditions. Other olefins, e.g. maleic acid, diethyl maleate, fumaric acid, were also catalytically hydrogenated by J. to their corresponding saturated compounds with much slower rates (Table V-5). V.2.5 Catalytic Isomerization J. was found to be an effective catalyst for isomerization. When hexene-1 was shaken with a DMA solution of _1 under N 2 for about two hours at room temperature ([hexene-1] = 0.08 M), about 80% of the hexene-1 was isomerized into hexene-2. The isomerization was investigated - 154 -Table V-4. Catalytic hydrogenation of C gH 1 4 by 1 i n DMA. Temperature = 25°C. [Ir] x 10 3 p[H 2] [H2] x 10 3 f C8 H14^ 8 3 1 : 6 of hydrogenation M mm Hg M M x 10 6, M s e c - 1 1.12 760 1.7 0.153 4,4 1.12 760 1.7 0.077 4.6 1.12 760 1.7 0.031 4.5 - 155 -time x 10 3,sec Fig. V-9. H 2 uptake plot of [ I r C l ( C g H 1 4 ) 2 ] 2 and mixture of cyclooctene and hexene-1 in DMA at 25°. -3 ([Ir] = 1.12 x 10 M; [hexene-1] = 0.081 M; [cyclooctene] = 0.081 M; p[H 2] - 760 mm Hg). - 156 -Table V-5. Catalytic hydrogenation of monoolefins using [ I r C l ( C g H 1 ^ ) 2 J 2 in DMA. 3 3 [Substrate] Temp. [Ir] x 10 p[H ] [H ] x 10 Rate of hydrogenation -1 M °C M mm Hg M M sec [hexene-2] 25 1.12 760 . 1 . 7 1.2 x 10~5 = 0.081 [hexene-1] 30 2.4 760 1.7 1.4 x 10~ 4 = 0.15 [MA] = 0.18 30 4.1 760 1.7 6.4 x 10~ 6 [DEMA] 30 5.4 760 1.7 4 x 10 _ 6 = 0.1 <: [FA] = 0.1 30 5.0 760 1.7 «vl x 10~6 - 157 -by gas chromatography (column = dinonyl phthalate, temperature = 20°C, current = 280 mA, carrier gas = He with pressure = 7 lb/ i n , retention time for hexene-1 =8.3 min, hexene-2 = 10 min). V.3 Discussion The observations and kinetic data presented above are most readily interpreted i n terms of the mechanism outlined below:-Ir + olefin -^-Sfc Ir (olefin) (5.4) + H 2 T , saturated , r„ , , , . , ^ + product I^lr (olefin)] The ligands on the Ir complex are omitted here. The olefin i s coordinated to the complex i n a rapidly established equilibrium. The rate determining step i s thought to involve the oxidative-addition of H 2 to a square planar Ir(olefin) complex. Such a path for hydrogenation g catalyzed by d metal complexes has been described as the "unsaturate route" (3r5,8). The hydrogen transfer process i s usually considered to be stepwise involving an alkyl-hydride intermediate (26). Another possible path for hydrogenation, the "hydride route" (3-5), which can result i n the same kinetics observed for an unsaturate route, involves an i n i t i a l rapid coordination of the H^  molecule to the complex followed by a rate determining coordination of olefin. This mechanism was ruled out, since metal precipitation due to reduction by H^ was observed when olefin was not present. Hence the presence of olefin i s essential for stabilization of the complex; any hydride complex - 158 -reacting in a slow step with olefin would decompose to metal-The above scheme yields the rate law (see Appendix V-l)„ -d[H-] k_K[H 2][olefin][Ir]_ (5.5) rate dt 1 + K[olefin] ([Ir] i s total iridium as monomer) Such a rate law indicates that the hydrogenation rate i s f i r s t order in both total iridium and H 2„ The ol e f i n dependence w i l l be f i r s t order at low concentrations, and w i l l gradually decrease to zero order at relatively high olefin concentrations (depending on K). The rate law accounts for the kinetic observations. Eq. (5.5) can be rearranged, as and a more quantitative testing of the data can be made. By plotting rate ^ against [hexene-1] ^ for the hexene-1 system, at constant [H,,] and [Ir]_, (Fig. V-10), a very good straight line was obtained over ; the concentration range of hexene-1 from 0.05 to 0.2 M. The ratio of intercept to the slope gives at 25° a K value of 20 M \ and k^ was evaluated from the intercept as 35 M "'"sec The Arrhenius plot using the data l i s t e d in Table V-2, is shown in + Fig. V - l l . The activation parameters evaluated are AH = 1 3 + 1 Kcal mole \ AS^ = -9 + 3 e.u. These measurements were made using an olefin concentration where the limiting rate i s approached (Fig, V-6); the rate approximates to k.,[Ir] [H_], and the activation parameters refer _ J ^ _ _ 1 + I rate ^ [ o l e f i n ] [H2] [Ir]_, k - J - ^ H l r ] -(5,6) - 159 -- 16.0 -- 161 -to the rate constant. The parameters are similar to those that have been determined for oxidative addition of fi^ to Vaska*s type + -1 complexes, Ir(CO)ClP 2; AH values between 10.8 and 12 Kcal mole , and + AS values of ^  -20 e.u. have been measured i n a number of solvents (27). The negative activation entropy reflects mainly the loss of translational entropy of molecule upon coordination (28); a 3-centre transition state i n which both hydrogen atoms interact simultaneously with the metal ion, was proposed. The parameters for the cyclooctene system thus appear consistent with a rate-determining step involving oxidative addition of R ^ The f i r s t order dependence on [Ir],j, gives no evidence for a dimer — m o n o m e r equilibrium and indicates that either the dimeric species i s the active catalyst or else that the dimer dissociates completely to an active monomer under the experimental conditions. The latter seems more l i k e l y for the absorption spectrum i n DMA solution i s different from that observed i n the solid state (Fig. VI-8); there are very few clear cases of hydrogenation through dimer catalytic species (18). Complex 1^  i s readily cleaved by donor ligands, such as PPh^ or olefins (6,7), and a strong coordinating solvent (such as DMA) or the added o l e f i n (such as hexene-1 or cyclooctene) are li k e l y to cleave the chloride bridge in [IrCKC-H, . ) _ ] . . I The nature of the I_r species present in the DMA (Scheme 5-4), i s not yet established. There i s no evidence for ionic chloride i n DMA solutions of jL. A K value of 20 M 1 implies that about half the total iridium exists as I_r at 0.05 M hexene-1; in contrast to the system in the absence of olefin no metal i s formed under H9, and thus the species - 162 -written as Ir must already contain coordinated hexene-1. The catalyst then i s l i k e l y to be monomeric in DMA, and K could refer to an equilibrium such as: (DMA) xIrCl(olefin) y + ol e f i n ^ (DMA) x_ 1IrCl(olefin) y + 1 (5.7) (where y is 1 or 2 and x+y = 3) The mechanism outlined i n (5.4) i s also thought to be involved in the cyclooctene system, but since the rate of hydrogenation is independent of the cyclooctene concentration, the ol e f i n complex must be f u l l y formed over the substrate concentration range used. That i s , K[olefin] >> 1, and the K value for the cyclooctene system must be much greater than that for the hexene-1 system. Eq. (5.5) can then be simplified into: d [ H ] rate = - " «__[_-] [Ir]_ (5.8) which applies over the cyclooctene concentration range 0.03-0.16 M. From the slope of the uptake plot shown in Fig. V-3, k^ was calculated to be 2.4 M ''"sec ^ at 25°C. Since the cyclooctene hydrogenation i s zero-order in olefin at the lowest concentration used, 0.031 M, (and K[olefin] >> 1), I K must be > 100. The small change in the spectrum of the DMA solution of _1 under N_ i n the presence of a large excess of hexene-1 indicates that complex formation between hexene-1 and the Ir cyclooctene complex i s very weak. This indicates that the cyclooctene i s s t i l l strongly coordinated to the - 163 -Ir(I) complex even at the start of the hexene-1 hydrogenatlons. This conclusion i s consistent with the findings discussed above on the determined K values, and i s further supported by the kinetic observations that the cyclooctene i s selectively hydrogenated f i r s t in a mixture of added hexene-1 and cyclooctene. In the gas uptake experiments with hexene-1, the autocatalytic period observed is almost certainly due to the slower hydrogenation of the cyclooctene which i s present i n the catalyst. A similar autocatalytic plot resulted when the solution mixture was equilibrated under an atmosphere prior to hydrogenation; this shows that the "autocatalysis" i s not due to some slow equilibration process. In the hydrogenation of the 1:1 mixture of hexene-1 and cyclooctene (Fig. V-9), hydrogenation of the latter substrate occurs at i t s slower individual rate of hydrogenation (4.4 x 10 ^ M sec ^ i n the mixture; 4,6 x 10 ^ M sec ^ for cyclooctene alone). These results agree with the findings reported by Candlin and Oldham (8) that individual rates of hydrogenation are the wrong c r i t e r i a to predict which substrate w i l l hydrogenate the most rapidly i n a mixture. About 80% of the cyclooctene i s selectively hydrogenated before the start of hexene-1 hydrogenation. The equilibrium constant K for the two ol e f i n systems is the determining factor for this selective hydrogenation. For the unsaturate route, since the K value for cyclo-octene i s much greater than that of hexene-1, i t is predictable that the cyclooctene should be hydrogenated f i r s t , as i t i s more strongly coordinated to the catalyst. Such a relationship Is also observed for the hydrogenation of alkynes and dienes using the RhCl(PPh 3) 3 catalyst (8). The hydrogenation for each substate is stepwise involving one molecule - 164 -of hydrogen at a time, since alkynes and dienes are bonded more strongly to the complex than the monoalkenes. It i s not immediately clear why the cyclooctene coordinates so much more strongly than the hexene-1. No literature data are available concerning s t a b i l i t y constant studies on iridium(I) complexes. Some selective binding data have been reported for a rhodium(I) system (29), but this involved only lower olefins (up to C^). Formation constants for o l e f i n complexing with silver(I) in ethylene glycol have been reported (19), and that for cyclooctene (14.4 M "*") i s somewhat greater than that for hexene-1 (4.3 M . It i s interesting to note, however, that this study reports a binding constant of > 1000 M ^ for the trans-cyclooctene system, the difference with the cis-cyclooctene (the usual commercially available form) being attributed to r e l i e f of steric strain i n the trans form on metal coordination. The factors affecting binding of olefins to square planar d systems remain to be elucidated. During our studies, van der Ent and coworkers (6) published reports on the use of benzene solutions of the iridium complex 1 for hydrogenation of hexenes. The addition of one or two equivalents of triphenyl-phosphine per Ir increased the ac t i v i t y of the system. These workers found the resulting phosphine systems were catalytically active for hydrogenation of hexene-1. The maximum hydrogenation rate was found at the ratio of [PPh^J/flr] = 2, suggesting that the catalytically active phosphine species was chlorobis(triphenylphosphine)iridium(I), IrCl(PPh 3) 2, and the intermediate was considered to be IrCl(PPh 3)_(olefin). At the ratio of [PPh 3]/[Ir] = 3, no hydrogenation of hexene was observed and this was attributed to a lack of coordination sites for both the - 165 -olefin and U^. The position of an iridium-chloride stretching vibration at 301 cm ^ of an isolated IrClCPPh^)^(C^H^) complex suggested a square-planar structure with chloride trans to ethylene (6). In the presently studied system, no PPh^ was added, and a square-planar multi-olefin complex i s believed to be the active species. Both phosphines and olefins possess the iT-bonding a b i l i t y to assist the stabilization of metals in low oxidation states. Van der Ent and coworkers (6) reported that in benzene solution, the hydrogenation rate of hexene-1 (0.27 M) using [ I r C l ( C 8 H 1 4 ) 2 ] 2 ([Ir] = 4.6 x 10 - 4 M) was ^  1 x 10 4 M sec \ at 25° and 1 atm H2« Under the same conditions, i n the presently studied DMA solution, the hydrogenation rate i s estimated as -4 -1 MD.5 x 10 M sec . In benzene, the iridium cyclooctene complex remains dimeric (2), while i n DMA the complex i s almost certainly monomeric. The similar hydrogenation rates observed in the two solvent systems are thus probably fortuitous. It should be noted that an attempt in the present work to hydrogenate hexene-1 i n benzene-solution (25°, 1 atm H2) resulted i n traces of metal and a very rapid heterogeneous hydrogenation. Presumably just as an active site can be blocked by a coordinated PPh^ group, i t could also be blocked by an olefin group. The kinetic observations in the present system show that the hydrogenation rate decreases significantly when the concentration of hexene-1 exceeds 0.2 M. It i s very l i k e l y that further ol e f i n coordinates to the iridium complex, and this blocks the way of the incoming E^ molecule; hence the hydrogenation rate i s inhibited. The olefin complex, I r C l ( C 2 H 4 ) 4 has been isolated from a heptane suspension of [IrCl(CgH^ 4) 2] 2 (7), but is stable only below 30° under an C aH A atmosphere. - 166 -Osborn and coworkers (9) have reported that the halide and not the diene ligand in complexes such as [Ir(diene)Cl]^ can be replaced by phosphine (L) to give cationic species such as [Ir(diene)L,2] + i f the reaction i s carried out i n certain polar solvents. The cationic species are also catalysts for homogeneous hydrogenation through i n i t i a l dihydride formation (9). Further studies on cationic complexes of the type Ir(diene)L n + (L = PPh3; n = 2,3) have been reported recently (20,21). Other non-phosphine-containing hydrogenation catalysts [Ir(diene)(NCCH^)^] + have been also isolated (10,21). However, the conductance measurement in the present work showed no evidence for chloride dissociation from the [IrCl(C 0H . ) . ] . complex in DMA. The diene o 1 4 2 z ligands are presumably more strongly attached than the monoene. The hydrogenation rate of hexene-2 is about 3 times slower, and that of cyclooctene i s about 10 times slower, than that of hexene-1. The slower hydrogenation rates of internal alkenes and cycloalkenes could be due to steric problems concerning coordination or transfer of the hydrogen molecule (22,23). It i s interesting to compare the hydrogenation rate ratio of hexene-1:hexene-2 of the [IrCl(C_H1.)„]„ O 1 4 JL 2, system to that of the I r C l ^ P h ^ ^ i n benzene which gives the ratio of about 180 (6). The greater selectivity i n the phosphine complex may result from i t s more st e r i c a l l y crowded nature and an associated greater d i f f i c u l t y in hydrogen transfer. The hydrogenation rates of substrates containing electron-withdrawing substituents, e.g. maleic acid, fumaric acid, and diethyl maleate, are much slower than that of hexene-1. Since the rate-determining step in the system is thought to be the oxidative-addition of molecular H_ t o a n o l e f i n complex, substrates containing electron-withdrawing substituents w i l l decrease electron density at the - 167 -metal centre and therefore decrease the ease of formation of an Ir(III) dihydride intermediate, thus slowing down the hydrogenation rate. A similar olefin reactivity pattern has also been observed for various olefins in hydrogenations catalyzed by HRuClCPPh^)^ (11). The inverse chloride effect (Fig. V-8) could be due to a competition between hexene-1 and Cl in the solution for coordination to the iridium. As the concentration of hexene-1 decreases due to hydrogenation, the chloride competes more effectively for coordination and the reaction rate decreases markedly, Ng has reported (1) the kinetics of the [RhCl(C0H,.)„]„-catalyzed hydrogenation of MA in 0.5 M LiCl-DMA solution at 80°; the addition of chloride was necessary to dissolve the rhodium dimer. The rate of hydrogenation was first-order i n both [Rh] ([Rh] as monomer) and [H^], and was independent of [MA], i.e., rate = k2[Rh][H.2], The rate-law then i s of the same form found here for the iridium system i n the absence of chloride. For the present system the bimolecular rate constant for hexene-1 hydrogenation i s 35 M '''sec * at 25°, and under similar conditions (Table V-5), the MA hydrogenation rate i s some 20 times slower. For the rhodium-catalyzed MA hydrogenation, k2 was 2.0 M ^sec ^  at 80°. Thus the Rh-catalyzed system in chloride media i s probably much less efficient than the corresponding Ir-catalyzed chloride-free system. The activation parameters for the Rh system (AH^ = 18 kcal mole ^ , AS^ = -7 e.u.; cf. values of 13 kcal mole ^ and -9 e.u, in the present work) were again thought to be consistent with oxidative-addition of H2 to a square planar Rh^-olefin complex. Oxidative-addition of H2 i s usually favoured for Ir'*' systems compared to corresponding Rh* systems, and this has been attributed to the lower electronic promotion - 168 -energy in the larger third transition series and the somewhat increased strength of the Ir-H bond compared to Rh-H (4). It i s d i f f i c u l t to make direct comparisons in the cyclooctene systems because of the differences in hydrogenation conditions regarding the chloride concentrations, particularly since added chloride inhibits the iridium systems markedly (Fig. V-8). The rhodium solutions did not catalyze cyclooctene hydrogenation; metal was again precipitated under i n the absence of the stabilizing maleic acid ligand. Effective catalysts for hydrogenations are generally found to be catalysts for isomerizations under atmospheres. [IrCl(PPh_)_] (6) and [Ir(diene) (F~Ph^)^]+ (10), for example, are both reported to be catalysts for isomerization as well as hydrogenation under H_. A commonly postulated mechanism (24) involves the alkyl intermediate usually required for the catalytic hydrogenation mechanisms RCH0CH=CH„ =* RCH0CH-CH_ RCH=CH-CH„ (5.9) 2 | 2 2| 3 - | 3 MH M MH The hydride intermediate may be generated by reaction of the metal olefin complex with molecular hydrogen. The DMA solution of 1 was found to be a catalyst for isomerization of hexene-1 to hexene-2 under a purely atmosphere. Plausible mechanisms (24) include: (a) a T i - a l l y l metal hydride intermediate:-RCH-CH-CH. — - RCH / I 2 •«= ^ CH • CH2 RCH=CHCH3 (5.10) M MH M - 169 -or (b) a carbene intermediate:-RCH„CH=CH„ * RCH_CCH„ RCH=CHCH„ (5.11) 2 | 2 K — 11| 3 x — j J M M M ' V.4 Catalytic Reduction of Molecular 0,, A number of workers (13,14,25) have recently reported on the reactivity of platinum metals in solution toward ^/C^ mixtures. The interest i n such studies stems from their relevance to the design of fuel c e l l s , and also from a potential method for hydrogen peroxide synthesis. The [IrCl(C_H1.)„]„ complex was also shown to be reactive toward o 14 2. 2. oxygen (Chapter VI). Hence molecular 0^ was tried as a substrate for catalytic reduction using a DMA solution of 1_. In the gas-uptake experiments, the p a r t i a l pressures of and 0^ were kept constant. The gas-uptake plots were linear, as shown i n Fig. V-12 and the limited data are summarized in Table V-6. Water was identified as the only product of the reaction by gas chromatography (column = carbowax, temperature = 2 100°, current = 230 mA, carrier gas = He with pressure = 10 l b / i n , retention time = 13 min). Tests for the presence of hydrogen peroxide (KI, K^CrO^) were negative. The overall reaction occurring in the presence of catalytic species 1 i s presumably: 2H2 + 0 2 > 2H20 (5.12) - 170 -Table V-6. [IrCl(C H ^) ] 2 catalyzed reduction of 0 2 in DMA. [Ir] x 1 0 3 M p [ o 2 ] nun Hg P[H 2] mm Hg temperature °C slope M sec k o M sec -1 1.12 160 600 25 4 x 10" 6 2.7 x i o " 6 1.34 240 480 64 6.4 x 10~ 5 4.3 x i o ~ 5 time, sec Fig. V-12. H. uptake plot of [IrCl(C 0H..).]_ in DMA at 64° with mixture of L O 14 2 / -H 2 (480 mm Hg) and 0 2 (240 mm Hg), ([Ir] = 1.34 x 10 M). - 171 -The pseudo-zero-order rate constant was calculated from the slope of the gas uptake plot, d[H.O] d[H,] - d[H + 0 ] 1- = -4 — = k (5.13) dt dt 3 dt o X V z where k Q = [Ir] [H^] [0^] ; the kinetic dependences, x, y and z were not investigated, however. The pale yellow solution obtained after reaction gave a continuum spectrum down to 350 nm (e (384 nm) = 460, e (414 nm) = 310, e (460 nm) = 200), and i s quite different to that measured for the [ I r C l ^ - I L ^ ^ ] -solution i n vacuum (Fig. V-2). This pale yellow solution was apparently stable to a i r , since no color change was observed, and the solution retained about 80% of i t s catalytic activity after one day i n a i r . Further, the solution was unreactive toward H_ alone (60°, 1 atm). The i . r . spectrum of the solution showed DMA and H^O peaks only; no 0-0 bands or hydride bands were observed. A blank reaction, i.e. without the presence of _1, showed no gas uptake for °2^2 m ^ x t u r e ° Although heterogeneous catalysis of the reaction (5.12) by metal surfaces was among the f i r s t recognized catalytic reactions at large (30), few systems using homogeneous catalysts have been reported. The f i r s t reported system involved Cu(II) i n aqueous acid solution at elevated temperatures and under high pressure of the gases (12), The f i r s t homogeneous catalytic combination of molecular a n d 0_ to form under ambient conditions appears to be that reported by Hui (13) using a Ru(III) chloride complex i n DMA solution. The rate law was k[Ru(III)][I^] and a mechanism incorporating rapid attack of a hydride by 0 2 was postulated, although details of this process were not discussed. - 172 -Vaska (14) has reported that some triphenylphosphine complexes of platinum, iridium, rhodium and ruthenium in toluene solutions catalyze the oxygen hydrogenolysis to R^ O; complex kinetics were observed (rate = [complex]^" [KUJ [0 ]) and mechanisms were not proposed. A recent paper by Reed and Roper (25) reported the use of [IrH(NO)(PPh^)^] as a catalyst. Hydroxide intermediates were shown to be involved. The a b i l i t y to form either individual 0^ and/or H9 complexes seems essential for catalytic formation of E^O from gaseous 0^ and E^* F o r t n e series of complexes [IrX(C0) ( P P h ^ ] (X = Cl, Br, I), the sequence of catalytic activity for ^-C^ combination was found to be Cl > Br > I (14) % the same trend is found for reversible loss of oxygen from the [(0^)IrX(CO)(PPh^)^] complexes (15), and also for both homogeneous catalysis of the ^-D^ exchange equilibrium and para hydrogen conversion in toluene solution (14). In DMA solutions of 1_, during a catalytic ^2^2 c o m D i n a t i ° n « t n e complex species i n solution must contain coordinated H90, OH or 0^ and possibly hydride, since the species, unlike the i n i t i a l complex J., is stable against H^  reduction to the metaL The formation of an 0^ complex of 1_ i n the presence of 0.2 M LiCl/DMA has been studied (see Chapter VI). A plausible mechanism involves a fast formation of an oxygen complex., followed by a rate-determining oxidative addition of molecular E^ . This corresponds closely to the olefin hydrogenation mechanism suggested earlier, with the oxygen playing the role of the ol e f i n in stabilizing the iridium (I) against reduction to metal, However, the nature of the hydrogen transfer step to give water and regenerate the catalyst species - 173 -is not easy to visualize. Since the uptake is completely linear, the new "stabilized" iridium catalyst must be formed in the very early stages of the reaction. An alternative mechanism involving an i n i t i a l slow formation of hydride followed by rapid reaction with molecular 0^ (which prevents metal formation) also seems feasible. Whether the catalytic activity for hydrogenolysis of 0^ i s caused by traces of co l l o i d a l metal formed by reduction of complex 1 i s a c r i t c a l question, since the complex i s very vulnerable to hydrogen reduction. No absolute proof can be provided for the homogeneity of the catalysis, although the observation of non-reactivity toward for the iridium solution at the end of the hydrogenation is more consistent with homogeneous conditions. More work i s s t i l l i n progress i n our laboratory for a better understanding of the mechanism of these mixed H2^°2 s v s t e m s ' possible production of H^ O^  by such process remains a distinct p o s s i b i l i t y . - 174 -APPENDIX V-l Ir - + o l Ir(ol) v I r + sat. product (S.P.) rate = ^ f ^ 1 = \ [Ir (ol) ] [H.] k ^ I I r H o l H H - ] As K [Ir(ol)] [Ir][ol] [Ir_] = total Ir - [Ir] + [Ir(ol)] - [Ir] + K[Ir][ol] [Ir_] [ — 1 1 + K[ol] rate = k^ I o l H H -nir . , ] 1 + K[ol] - 175 -REFERENCES 1. B.R. James and F.T.T. Ng, Chem. Comm., 908 (1970); F.T.T. Ng, Ph.D. Thesis, University of British Columbia, July 1970. 2. J.L. Herde and C.V. Senoff, Inorg. Nucl. Chem. Letters, ]_, 1029 (1971). 3. F.H. Jardine, J.A. Osborn and G. Wilkinson, J. Chem. Soc. (A), 1574 (1967) 4. J.A. Osborn, F.H. Jardine, J.F. Young and G. Wilkinson, J. Chem. Soc. (A), 1711 (1966). 5. S. Montelatici, A. van der Ent, J.A. Osborn and G. Wilkinson, J. Chem. Soc. (A), 1054 (1968). 6. H. van Gaal, H.G.A.M. Cuppers and A. van der Ent, Chem. Comm., 1694 (1970). 7. A. van der Ent and T.C. van Soest, Chem. Comm., 225 (1970), 8. J.P. Candlin and A.R. Oldham, Discuss. Faraday Soc, 46_, 60 (1968). 9. J.R. Sharpley, R.R. Schrock and J.A, Osborn, J. Amer. Chem. Soc, 91, 2816 (1969). 10. M. Green, T.A. Kuc and S.H. Taylor, Chem. Comm., 1553 (1970), 11. L. Markham, Ph.D. Thesis, University of British Columbia, March 1973. 12. (a) H.F. McDuffie, E.L. Compare, H.H. Stone, L.F. Woo and CH. Secoy, J. Phys. Chem., 62, 1030 (1958); (b) T,V, Berlina and V.A. Tulupov, Katal. Reakts. Zhidk. Faze, Tr, Vses. Konf., 1962, 330 (1963). 13. B.C.Y. Hui, Ph.D. Thesis, University of Br i t i s h Columbia, September 1 1969. 14. L. Vaska and M.E. Tadros, J. Amer. Chem. Soc, 92, 7099 (1971). 15. L. Vaska, Accounts Chem. Res., JL, 335 (1968), preprint of paper by L. Vaska, L.S. Chen, and C.V. Senoff, to be published, 16. G.L. Rempel, Ph.D. Thesis, University of British Columbia, p. 170 (1968). - 176 -17. G.R. Lester, T.A. Gover and P.G. Sears, J. Phys. Chem., 60, 1076 (1956). 18. B.C. Hui, W.K. Teo and G.L. Rempel, Inorg. Chem., 12, 757 (1973). 19. M.A. Muhs and F.T. Weiss, J. Amer. Chem. Soc, 84, 4697 (1962). 20. L.M. Haines and E. Singleton, J. Organomet. Chem., Z5, C83 (1970). 21. M. Green, T.A. Kuc and S.H. Taylor, J. Chem. Soc., (A), 2334 (1971). 22. D. Evans, J.A. Osborn and G. Wilkinson, J. Chem. Soc.,(A), 3133 (1968). 23. C. O'Connor and G. Wilkinson, J. Chem. Soc., (A),2665 (1968). 24. C.W. Bird, Transition Metal Intermediates in Organic Synthesis, Logos Press, London, 1967, p. 69. 25. CA. Reed and W.R. Roper, J. Chem. Soc, Dalton, 1014 (1973). 26. J. Halpern, Ann. Rev. Phys. Chem., 16, 103 (1965). 27. B.R. James, Homogeneous Hydrogenation, Wiley Interscience, New York, 1973, p. 292. 28. L. Vaska and M.F. Werneke, Trans. N.Y. Acad. Sci., 31, 70 (1971). 29. V. Schurig and E. Gil-Av, Chem. Comm. (London), 650 (1971). 30. G.J.K. Acres, Platinum Metals Rev., 10, 60 (1966) and references therein. I d CHAPTER VI ACTIVATION OF MOLECULAR OXYGEN BY SOLUTIONS OF u-DICHLOROTETRAKIS-(CYCLOOCTENE)DI-IRIDIUM(I), 1. VI.1 Introduction The reactions of metals or their compounds with small gaseous molecules play a v i t a l role in both l i f e and industrial processes. Synthetic reversible oxygen-carrying complexes have been of great interest in recent years as model compounds for the study of reversible oxygenation mechanisms involved in much more complex natural oxygen-carriers, e.g. hemoglobin. Vaska's (1) discovery of the reversible molecular oxygen carrier Ir(C0)Cl(PPh 3) 2 i s of great importance, since the 1:1 oxygen adduct lr(0 2)(CO)Cl(PPh 3) 2 can be recrystallized and is extremely stable and well characterized (2-4). Few kinetic studies of . the formation and catalytic activity of molecular oxygen complexes have been reported (5). James and coworkers (6-8) have reported on the reaction of [RhCl(C_H .) ] with molecular oxygen to give a 1:1 metal-cs i-H 2 2. i • oxygen species in 0.2 M LiCl/DMA solution. Some kinetics of catalytic oxygenation reactions with the cyclooctene compound were also studied, Studies involving the reactivity of the corresponding iridium complex, [IrCl(CQH1.)„]„, (9) toward molecular oxygen are reported i n o 14 2 2 this chapter. - 178 -VI.2 Experimental ... . Results for the DMA System VI.2.1 Kinetics and Stoichiometry The orange solid [IrCl(C 0H..)_]_ i s f a i r l y oxygen-sensitive and decomposes slowly to unidentified green products when exposed to a i r . The compound i s soluble i n many common laboratory solvents such as benzene, chloroform, and acetone, as well as DMA, and a l l the i n i t i a l l y yellow-orange solutions turn dark green i n a i r . Some quantitative data in the DMA system w i l l f i r s t be presented. The iridium complex was insufficiently soluble i n the pure solvent for quantitative gas absorption measurements, but i t readily dissolved i n the presence of added L i C l . The solutions were usually made 0.2 M in chloride. At 25°, such solutions of [IrCl(CQH.. , ) 0 ] 0 i n 0.2 M LiCl/DMA were found to absorb oxygen at a l l pressures i n a 1:1 ratio of 0 2 : l r ([Ir] = monomer), the reaction being accompanied by the color change to dark green. The oxygen uptakes were followed as described in Section II.2. The [ I r C l ( C g H 1 4 ) 2 ] 2 complex was slowly dissolved i n a 0.2 M LiCl/DMA solution by shaking under nitrogen for about 2000 sec. Oxygen was introduced and the rate of oxygen uptake was measured and plotted in terms of mole ratio of [0 2]:[Ir] ([Ir] = monomer). A typical uptake curve i s shown in Fig. VI-1. The solutions absorb oxygen i n a f i r s t -I order manner to an [0 2 ] : [ l r ] ratio close to unity yielding dark green solutions, a 1:1 ratio was measured for systems i n which the 0 2 pressure varied from 79-312 mm Hg. The uptake does not level off completely, and a very slow oxygen uptake was evident over long periods. A good linear plot of log [Ir] against time was obtained from the rate data, assuming that the loss of the starting iridium complex was proportional to the 0 9 consumption (see - 179 -time x 10"J,sec Figure VI-1. O2 uptake plot of [ I r C l t C g H i ^ j ^ i n °- 2 M LiCl/DMA solution at 25°, p[0 2] = 310 mm Hg (3.10 x 10~3 M), and the corresponding log plot ( A ) : [ I r ] = 1.52 x 10 - 2 M; ( Q ) [Ir] = 0. - 180 -inset of Fig. VI-1). From the slope of the log plot, a pseudo-frist-order rate constant k^, and hence the i n i t i a l rates ( i n i t i a l concentration of [Ir] x kp x 2.303) could be evaluated. The data derived from these measured slopes are summarized in Table VI-1 for experiments in which the concentrations of [Ir] and [0 2] were varied. The rate dependences on the concentration of [Ir] and [0 2] are shown in Fig. VI-2 and Fig. VI-3. Both plots give good straight lines, showing that the reaction -2 rate i s first-order in [Ir] (up to 3 x 10 M) and in 0 2 (up to 7.6 x _3 10 M or one atm. pressure). In the absence of the [IrCl(C DH .)„]_ complex, o 1 4 / i. a very slow reaction between 0 2 and 0.2 M Lid/DMA solution was observed -"the blank reaction", Fig. VI-1. 'The rate of this blank reaction (2.5 x 10 ^ M sec *) i s practically independent of 0 2 pressure from 40 to 800 mm Hg (10). Free cyclooctene ([cyclooctene]:[Ir] = 1.4) was detected i n the i n i t i a l lr-complex-0.2 M LiCl/DMA solution by v.p.c. (Carbowax column, 200 mA, 130°, pressure gauge setting = 10, retention time = 10 min). An i n i t i a l lr-complex-0.2 M LiCl/DMA solution was shaken in a i r for about 1 hr, and the resulting dark green solution was injected into the same v.p.c. column under the conditions mentioned above. Again, free cyclooctene was detected ([cyclooctene]:[Ir] - 1.3). Cyclooctenol and cyclooctenone were not detected in the f i n a l dark green solution (the retention times for cyclooctenol and cyclooctenone are > 30 min in the above column under the same conditions). The data on the effect of added LiCl are summarized in Table VI-2, and the rate dependence is shown in Fig. VI-4. First-order uptake plots to a 1:1 Ir/0 2 mole ratio were obtained at each of the chloride concentrations studied. Without the presence of added chloride, the - 181 -Table VI-1. Reaction of [IrCl(C-H.. ).]_ with oxygen; kinetic data at 25° i n 0.2 M LiCl/DMA. [Ir]xl0 p[0„] [0 ]xl0 i n i t i a l rate M sec 5 M . mm Hg M x 10 2.78 312 3.10 1.9 1.52 312 3.10 1.0. 0.85 312 3.10 0.50 0.39 312 3.10 0.31 0 312 3.10 0.025 1.52 758 7.58 2.5 1.52 576 5.71 1.8 1.52 179 1.86 0.57 1.52 79 0.88 0.32 1.52 0* 0 0 in a nitrogen atmosphere. - 182 -[Ir] x 10 , M Figure VT-2. Dependence of i n i t i a l rate on [Ir] in [ I r C l t C g H j ^ ^ ^ - ^ " LiCl/DMA system at 25°; [LiCl] = 0.2 M; p[0 2] - 310 mm Hg, - 184 -Table VI-2. Oxygenation of [IrCl(C-H ) ] i n DMA at 25°; kinetic data for variation of L i C l concentration. [Ir]xl0 2 p[o2] [ O j x l O 3 [LiCl] I n i t i a l rates x 10^ M mm Hg M M yr -1 M sec 1.52 312 3.10 0 solid not dissolved 1.52 312 3.10 8.5 x 10~ 3 1.6a 1.52 312 3.10 1.7 -2 x 10 1.2b 1.52 312 3.10 2.6 x 10~ 2 i . o b 1.52 312 3.10 3.4 x 10~ 2 l . l b 1.52 312 3.10 10.8 -2 x 10 l . l b 1.52 312 3.10 21.4 x 10" 2 i . o b Shaken under N 2 for over 10 hr before a l l [Ir] dissolved. Shaken under N for ^  2000 sec before 0 was f i l l e d in the reaction flask. z 1 [LiCl], M Figure VI-4. Dependence of i n i t i a l rate on [LiCl] in the [IrClCCgHj^^J 2~ 02-DMA solution at 25°; [Ir] = 1.52 x 10~ 2 M; p[0 2] =312 mm - 185 -complex was essentially insoluble in DMA. Once the mole ratio of Cl :Ir exceeds about unity, the rate of 0^ uptake is practically independent of the added chloride concentration. For the data at high chloride concentration the rate-low can be written simply as: rate = k 2 [ l r ] [ 0 2 ] , where k 2 is the second-order rate constant. Values of k 2 can be obtained from the slopes of the i n i t i a l rate plots shown in Fig. VI-2 and VI-3. The average k 2 value at 25° was evaluated as 0.23 M^sec" 1. The Arrhenius plot from the rate-temperature data gives a f a i r l y good straight line over the limited temperature range studied of 22.5° to 29.7°, Table VI-3, Fig. VI-5. The activation parameters evaluated + - I t for k 2 are AH = 6 + 2 kcal mole and AS = -37 + 7 e.u. The oxygen reaction of the complex JL appears to be essentially irreversible. The dark green product solution (Ir0 2) does not yi e l d the original orange color solution under pumping at room temperature for 15 min, and does not absorb further 0 2 after pumping and re-admitting 0 2. However, the dark green solution does change into orange gradually i n a sealed e.s.r. tube under a i r at room temperature over a period of 12 hr. This orange solution reverted to the dark green color when the solution was exposed to a i r again. The dark green oxygenated solutions were found to absorb H 2 at 25° to a mole ratio of H 2:Ir = 1.6, and gave brown solutions (II), These subsequently absorbed 0 2 to a mole ratio of 0 2 : l r - 0.1 to regenerate a dark green solution (III). When this dark green solution (III) was retreated with H 2 at 25°, no gas uptake was observed, although a dark brown solution (IV) resulted. The above processes are shown in the - 186 -Table VI-3. Variation of k with temperature. 2 3 [Ir]xlO p[0 2] [O 2]xl0 Temperature k 2 M mm Hg M °C M"1sec~ 1.52 312 3.10 29.0 0.257 1.52 312 3.10 27.0 0.231 1.52 312 3.10 25.0 0.215 1.52 312 3.10 22.5 0.203 3.30 3.35 3.40 T" 1 x 10 3 Figure VI-5. Arrhenius plot of [IrCl(CoH 1 4) 2] 2-0 2-0.2 M LiCl/DMA system; [Ir] = 1.52 x 10~2 M; p[0 2] = 312 mm Hg. - 187 -following scheme for the 0.2 M LiCl/DMA solutions. [Ir] + 1.0 o 2 dark green solution + 1.6 H 2 orange (I) + o . i o 2 dark green solution (III) ^2000 sec' no uptake dark brown or black solution (Scheme 6.1) Treatment of the original orange solution of 1 with a mixture of hydrogen (p[H 2] = 260 mm Hg) and 0 2 (p[0 2] = 500 mm Hg) at 25° resulted in an absorption of ^ 2.6 moles of gas per Ir. Treatment of the complex _1_in 0.2 M LiCl/DMA with H 2 alone under these conditions results i n precipitation of metal (see Chapter V), VI.2.2 Spectroscopic Measurements The kinetic results and the 1:1 stoichiometry for 0 2 uptake without the production of oxidation products (such as cyclooctenone) give good indirect evidence for the formation of a 1:1 molecular Ir(I ) - 0 2 complex. However, isolation of any such 0 2 complex from solution has been unsuccessful, and the identification of the product i n solution has proved to be very d i f f i c u l t and somewhat inconclusive. Spectroscopic studies on the system are described below. VI.2.2a Infrared _2 A 3 x 10 M solution of _1_ in 0.2 M LiCl/DMA was exposed to air at room temperature, and the i . r . spectra were followed as the solution - 188 -turned green. No peaks appeared at the 800-900 cm x region where the characteristic v n stretch i s commonly found (1). Other peaks observed were those of the DMA solvent. A glue-like brown solid residue was obtained after removal of the solvent by pumping from the oxygenated green [IrCl(C0H.. . )„] -0.2 M LiCl/DMA o 14 2 2. solution. This solid did not mix with Nujol very well; however, the nujol i . r . spectrum showed the same peaks as DMA, plus one broad peak at 3400 cm ^ which could be attributed to hydroxide. VI.2.2b Visible The orange [IrCl(C0H.. .)_]_ solid i n nujol (applied on a f i l t e r paper) o 14 2 2. gives two absorption peaks at 420 nm and 470 nm, and a continuum below 390 nm (Fig. VI-6). The orange solution of 1_under vacuum gives exactly the same two peaks as observed i n the so l i d state at 420 nm (e = 600) and 470 nm (e = 250), Fig. VI-6. On bubbling 0 2 through the solution at room temperature, the above two peaks gradually disappeared with the development of a broad peak at 620 nm (see Fig. VI-6, curve c) which gives rise to formation of the green color. The rate of increase of the peak intensity at 620 nm (e = 360) and a corresponding log plot are shown in Fig. VI-7 (log [A -A ] vs. time, where A and A are the m t m t absorptions at the maximum and at time t respectively). A second order rate constant derived from the slope of the log plot for production of the -1 -1 green color, assuming the rate-law = k2l"Ir][02], i s 0.26 M sec which i s essentially the same as the value obtained from the gas-uptake -1 -1 -3 experiments at 25° (0.23 M sec ). A second experiment at 1.6 x 10 M [Ir] gave a k,, value of 0.30 M ^sec ^ at room temperature. 400 500 600 700 wavelength, nm Figure VI-6. Visible absorption spectra of [IrClCCgEL^)^ in 0.2 M LiCl/DMA at room temperature under 0 (1 atm ); [Ir] - 2.55 x 10~ 3 M; 10 mm c e l l path. - 190 -O.D. Figure VI-7. Variation of O.D. at 620 nm for [IrCl(C 8H 1 4) 2] 2-O 2-0.2 M LiCl/DMA solution at R.T. and the corresponding log plot; [Jr] = 2.55 x _3 10 M; p[0,] - 1 atm.; 10 mm c e l l path. - 191 -The intensity at 620 nm then decreased very slowly on further exposure to 0^, and the f i n a l absorption was shifted down to 590 nm (e = 180). An approximate isosbestic point i s observed at 490-500 nm throughout the i n i t i a l development of the green color (Fig. VI-6). VI.2.2c ESR No esr signal was observed at liquid temperature for a solution of 1_ ([Ir] = 1.5 x 10~ 2 M) i n 0.2 M LiCl/DMA when kept under vacuum or under nitrogen. When the solution was exposed to a i r at R.T. for a few minutes and then cooled to liquid N_ temperature, three signals developed. These signals were detected at 3305, 3245 and one around 2980 Gauss and correspond to g values of 1.975, 2.011 and 2.190 (Fig. VI-8). The relative height of these signals increased quickly with exposure time at R.T. up to a maximum and then reached an equilibrium value i f l e f t i n a i r overnight (Fig. VI-9). The data show that the diamagnetic solution of 1_reacts with 0_ to give some paramagnetic species which then apparently slowly "decomposes". The solution turned green as the esr signals developed and remained green as the signals faded. The orange solution formed f i n a l l y i n a sealed esr tube with limited oxygen, was diamagnetic. No esr signals were observed for a 0.2 M LiCl/DMA solution when exposed to a i r and then cooled to liquid N^ temperature. VI.2.3 Oxidation of Substrates by 0.2 M LiCl/DMA Solution of 1 The system i s not an effective catalyst for oxidation of olefin, e.g. cyclooctene and hexene-1 were not oxidized in a 0.2 M LiCl/DMA solution under the conditions of [Ir] = 1.52 x 10 M, 1 atm 0 , 25° and - 192 -Gauss Figure VI-8. Esr spectrum of [IrCl(CgH 1 4) 2] 2-O 2-0.2 M LiCl/DMA after exposed to air for 6 min and cooled to liquid N 2 temperature; [Ir] = 1.52 x 10~ 2 M. relative intensity at 3,245 G Time, min Figure VI-9„ Esr measurement at 3245 G for [IrCl(C gH 1 4) 2] 2-0.2 M LiCl/DMA solution under air at liquid N, temperature; relative intensity vs. time for exposure to a i r at room temperature; [Ir] - 1.52 x 10~ 2 M. - 194 -[olefin] = 0.1 M. However, cumene (isopropyl benzene) was oxidized under mild conditions (25°, 1 atm 0 2 > [Ir] = 1.52 x 10~ 2 M, [LiCl] = 0.2 M, [cumene] = 0.1 M), and the 0,,-uptake, over a period of 12 hrs, corresponded to about 80% oxidation of cumene (one 0 2 per cumene). The oxidation products were not identified, but are probably phenol and acetone (11). VI.3 Reactivity of 1 toward 0 2 in Benzene, Chloroform, and Ethylene  Chloride Solutions The 02~uptake measurements using orange benzene solutions of the complex 1_gave a total absorption corresponding to [ 0 2 ] : [ l r ] ([Ir] = monomer) ratio of 0.5 to 1.2. A green color developed as the uptake proceeded. The 02-uptake plot In benzene i s very similar in shape, to that observed in the 0.2 M LiCl/DMA solution (Fig. VI-1). The corresponding log plot based on the 02~uptake data gave a f a i r l y good straight line (Fig, VI-10). The i n i t i a l reaction rates were evaluated i n a manner similar to that used for the 0.2 M LiCl/DMA system, and the results are summarized in Table VI-4» Since the mole ratio of 02-uptake ([0 2]/[Ir]) i s so inconsistent, quantitative analysis of the resultswas not attempted, although the rate seems to be first-order i n [Ir] and possibly first-order i n 0,,. In benzene, complex 1_ i s found to be a dimer (M. wt. = 878 determined by Galbraith Lab. Inc.; calculated for dimer, M. wt. 896), which agrees with the finding of Senoff and Herde* (9). On addition of cyclooctene into benzene solutions of 1^, no extra 00-uptake corresponding to oxidation of cyclooctene was observed - 195 [0 2]/[Ir] log [Ir] -3 time x 10 ,sec Figure VI-10. 0 2 uptake plot (O) of [trCKCgH^) ] i n benzene at 25°; p[o 2 ] 216 mm Hg, and the corresponding log plot ( A ) [Ir] - 1.52 x 10 -2 M. - 196 -Table VI-4. Kinetic data for the oxygenation of [IrCl(C0H..)„]„ in o 14 2 2 benzene. Temp. °C [Ir]xl0 2 M P t o 2 ] mm Hg M [ 0 2 ] / [ I r ] b I n i t i a l rate x 10"* M -1 M sec 25 0.804 216 0 0.65 0.74 25 0.815 216 0 0.49 -25 1.52 216 0 0.90 1.2 25 1.52 216 0 0.80 1.6 25 1.52 216 0 0.65 1.0 25 2.32 216 0 0.70 2.3 25 1.52 374 0 0.90 2.0 25 1.52 666 0 0.45 -27.7 1.52 216 0 0.65 1.8 30 1.52 216 0 1.2 1.4 33 1.52 216 0 0.53 --25 1.52 216 0.26 1.05 1.2 25 1.52 216 0.50 0.75 1.1 Cyclooctene. iTotal mole of 0 2 absorbed i n terms of mole ratio.[0 ]/[Ir], ([Ir] = monomer). - 197 -(Table VT-4). Cyclooctene was detected to a [cyclooctene]:[Ir] ratio of ^1 in the oxygenated green solutions by v.p.c. (carbowax column, current = 200 mA, T - 130°, carrier gas = He with pressure = 2 10 l b / i n , retention time 5.5 min). No cyclooctenol or cyclooctenone was detected by the same v.p.c. column under the same conditions (retention time for both cyclooctenol and cyclooctenone > 30 min). Hence, benzene solutions of 1^  do not catalyze the oxidation of cyclooctene. Removal of the solvent, from the oxygenated green solutions gave a green solid residue, which gave strong and sharp i . r , bands at 800, 1260 cm 1 a broad peak at 1000-1100 cm \ as well as a small but sharp -1 -1 peak at 860 cm . The i . r . band at 860 cm may possibly be attributed to 0-0 stretching, while that at 1260 cm"1 could be 6(0~H) of hydro--1 peroxide (8). I.r. bands at 2900 and 1500 cm corresponding to the coordinated cyclooctene, were also observed. Partial elementary analysis of the green residue showed C: 26.4%; H: 4,31%, A solid of composition IrCl(C gH 1 4)(0 2) analyzes theoretically for 26,0% C and 3.9% H. However sometimes, other batches of green solid, prepared by the same ™1 method, gave no i . r . peaks in the range of 800-1300 cm , Also, a benzene solution of l_was exposed to a i r for various time periods, and the solvent was then vaporized; the resulting green residue again gave no i . r . peaks i n the range of 800-1300 cm The green residue does not dissolve i n ether, acetone, H^ O or CCl^, but dissolves in strongly coordinating solvents such as DMA, DMF and pyridine. The complex 1_ i s very soluble in chloroform and ethylene chloride. These orange solutions are very sensitive to a i r , and rapidly turn green. After removal of the solvents, the resulting green residues showed - 198 -no characteristic v(O-O) stretching frequencies in the i . r . range of 800-900 cm"1. VI.4 Discussion The observed kinetics, and the 1:1 stoichiometry for 02~uptake i n the I r ( I ) - 0 2 system do not give direct evidence for a dimer monomer equilibrium in DMA solution. However, complex 1_is readily cleaved by donor ligands, such as PPh^ or olefins (12), and i t seems certain that the DMA-insoluble dimeric complex _1_ breaks up into monomeric anionic species in the presence of added chloride. Surprisingly, however, the two peaks i n the v i s i b l e absorption spectrum of 1_ i n 0.2 M LiCl/DMA are at the same wavelengths as those observed in the solid state, and this suggests that the environment of the Ir(I)-metal ion i s very similar both i n the solid state and in 0.2 M LiCl/DMA solution. The following equilibrium seems plausible: l / 2 [ I r C l ( C 8 H 1 4 ) 2 ] 2 ° - 2 D ^ L i C 1 > [ I r C l 2 ( C 8 H 1 4 ) 2 ] - (6.2) The free cyclooctene, which was detected i n the i n i t i a l solution of 1 in 0.2 M LiCl/DMA by v.p.c, may result from a decomposition of Ir-species i n the v.p.c. column. The cyclooctene i s considered to be reasonably l a b i l e in solutions of the [IrCl(C_H1.)„]. complex since O 1 4 / 2. the dimer has been used for the synthesis of other iridium(I) compounds (30), but equilibria of the type shown in eq. (6.3) seem unlikely l / 2 t I r C l ( C 8 H 1 4 ) 2 ] 2 [ I r C l x ( D M A ) y ] ( x - 1 ) - + 2C 8H 1 4 (6.3) - 199 -considering the v i s i b l e spectra data. From the observed kinetics of the present system, the most l i k e l y reaction occurring is simply k2 Ir(I) + 0 2 ^—> l r ( l ) ( 0 2 ) (6.4) where the ligands on Ir(I) are omitted. The rate law for the reaction, rate = k 2 [ l r ] [ 0 2 ] , agrees with the observed kinetics, and consistent second order rate constants are obtained from both gas-uptake measurements and spectral data for formation of a green t -1 IrCV. product. The activation parameters for k 2 (AH = 6 kcal mole , AS^ = -37 e.u.) are very similar t o those reported by Vaska and coworkers (13,14) for the formation of oxygen adducts from Ir(CO)ClL 2 t - I t (L = phosphines) complexes (AH = 8.5 to 10.8 kcal mole , AS = -32 to -42 e.u.) in chlorobenzene. The second-order rate constants for formation of the Ir(C0)ClL 2(0 2) adducts range from 0.12 to 49 M^sec" 1 at 40° (13), while k 2 i s 0.23 M *sec * at 25° for the present system. Choy and O'Connor (5) have studied the uptake of oxygen by [Ir(CO) (PPh^H^HClO^], t - I t and the activation parameters were AH = 8 kcal mole , AS = -34 e.u. i n acetone, and AH*" = 10 Kcal mole \ AS**" = -27 e.u. i n -1 -1 dichloromethane; the second-order rate constants were 0.33 and 0.61 M sec in these solvent systems respectively. The similarity of the activation parameters indicates that formation of the oxygen adducts goes through essentially the same mechanism. The negative activation entropy corresponds to the loss of translational entropy of the 0 2 molecule during the formation of the activated state. - 200 -The rate of oxygen adduct formation i n the present work is independent of the added Cl concentration, as long as the [Cl ]:[Ir] ratio exceeds unity. The same species such as [IrC^CCgH^^] i s believed to be involved under these conditions. At lower added Cl concentrations ([Cl ]:[Ir] ^  0.5), a much longer time is necessary to dissolve complex _1; the enhanced 02~uptake rate may be due to a different species, presumably a monochloro complex i n solution. The oxygenated solution of _1 subsequently absorbs another 1.6 mole ratio of H 2 t o 8^ve a dark brown murky "solution" (Scheme 6.1), Such a stoichiometry could be accounted for by reactions (6.5) and (6.6), Ir(I)(0 2) + H 2 ^ Ir(I)(H 2) + 0 2 (6.5) Ir(I) + 2C gH 1 4 2 , 5 E2y lr(0) + 2CgH1(. (6.6) Coordinated 0 2 is f i r s t replaced by molecular H2, which then hydrogenates cyclooctene, and reduces Ir(I) to metal. These reactions would give an apparent gas uptake mole ratio [H 2]:[Ir] - 1.5, The brown solution II (Scheme 6.1) shows only a par t i a l a b i l i t y to absorb 0, which shows that the i n i t i a l Ir(I) species have been decomposed, When a mixture of H 2 and 0 2 i s introduced into solution _1, a 2,5 mole ratio of gas uptake occurs - this could again result from reactions (6,5) and (6.6), or solely reaction (6.6). A catalytic formation of H20 is not observed. Without the presence of added Cl , molecular 0 2 i s catalytically hydrogenated to H20 by complex_1_ (see Section V,4), It is l i k e l y that the added Cl coordinates to the Ir(I) (cf. eq. 6.2), and inhibits the coordination of both molecular H„ and 0_. - 201 -The absence of an esr signal i n a DMA solution of 1_ kept under vacuum, and the buildup of the esr signal on exposure to a i r i s due to some paramagnetic species formed in the presence of 0^ i n the a i r . The origin of these signals has not been identified exactly but i t i s caused by the reaction of 0^ with Ir ( I ) . The exact nature of the possible Ir(I ) ( 0 2 ) species i s not known but superoxide species such as Ir(II ) 0 2 , i n which electron transfer has taken place from Ir(I) to coordinated O2, are believed to be involved. Such electron transfer has been demonstrated for a number of metal-molecular O2 complexes. There are a number of reports on the esr signals of Co(II ) -02 complexes (for example 15-18). An eightline spectrum arising from the interaction of an unpaired electron with a single Co-nucleus (I = 7/2) was observed for systems involving cobaloximes(II) (15), Co(II) (3-CH_0-salen) (16), Co (II) (vitamin B ^ ) (17) and Co (II) (acacen) (18). These workers (15-18) concluded from the esr spectra that the unpaired electron i s largely associated with 0^ i n mononuclear species and that the Co(II) oxygen complexes are best formulated as Co(111)02 * species. An oxygenated 0 .2 M LiCl/DMA solution of [RhCl(C 0H..).]. (6) cooled o 14 2. / to 77°K showed esr signals. It was proposed (6) that the species In solution was Rh(II ) . . .02 *; the interaction between Rh(II) and 0^ • i was thought to be relatively weak, and the esr spectrum was essentially the superimposition of those of 0^ • and a d^ Rh(II) species. A small hyperfine s p l i t t i n g of the 0^ signals by the Rh-nucleus (I = 1/2) was observed. - 202 -The only esr report on Ir(II) compounds (19) is that of CP2I1:. The g values of Cp 2Ir (g H = 2.033, gj. = 2.001) are very close to those of the corresponding analogous Rh complex, Cp2Rh (g(, = 2.033, gj_ = 2.003), and a small hyperfine coupling with the Ir-nucleus (I = 3/2) was detected. No esr signals has been reported for Ir(I)(0 2) complexes; the 02~adduct of Vaska's compound, Ir(CO)Cl(PPh 3) 2 > gives no esr signal (32). On comparison with the esr data for the Rh(I)0 2 system (6), the esr signals of I r ( I ) 0 2 i n the present work may be attributed to the superimposition of those of 0^~- (15,18) and those of Ir(I I ) . No s p l i t t i n g by the Ir-nucleus (I = 3/2) i s observed. The following equilibria, believed to exist i n solution, could account for the esr spectral changes: Ir(I).+ 0 2 Ir(II)...0 2". " I r ( I ) ( 0 2 ) M (6.7) When the Ir(I) i s exposed to 0 2» an equilibrium takes place with relatively fast formation of Ir ( I I ) . . . 0 2 •, which can be detected by esr. The Ir(II)Q 2 • species then changes slowly to some Ir(I)(0 2) species which i s diamagnetic. Fig. VI-9 i s thought to show the concentration change of I r ( I I ) 0 2 • with respect to time. It i s believed that a slow solvent oxidation by way of a free radical path (scheme 6,11) occurs, with regeneration of Ir(I). As for the corresponding Rh system i t i s possible that decomposition to Ir(II) i s a competing processes (Scheme 6.11). It i s worth commenting on the relative reactivity of corresponding Ir and Rh systems toward oxygen. Vaska and coworkers (20) have reported the f i r s t direct comparison of the reactivity of planar d complexes of the - 203 -cobalt triad toward oxygen. The univalent cationic complexes, [M(2-phos)2]BPh4 (M = Co, Rh and Ir; 2-phos = cis-Ph2PCHCHPPh2) reacted with oxygen with second-order rate constants 1.7 x 10 , 0.12 and 0.47 M '''sec 1 respectively for Co, Rh and Ir (see also Section 1.4). The results show the reactivity order of Co >> Ir > Rh. The Ir(C0)ClL 2 (L = phosphines) complexes undergo oxidative-addition toward 0 2 with second-order rate constants ranging from 0.12 to 49 M ^sec 1 (13). The reaction rate i s proportional to the basicity of the metal complex which i s a function of the electron-donating properties of the ligand (14). The analogous Rh complexes, Rh(C0)ClL 2, do not form oxygen adducts and thus the order of reactivity for these isoelectronic complexes toward 0 2 is again Ir > Rh. James and coworkers (6,8) have reported on the reaction of [RhCl(C gH 1 4) 2] 2 with 0 2 i n 0.2 M LiCl/DMA solution. The following equations were proposed to account for an observed 1:1 rapid oxygen uptake followed by a subsequent much slower absorption: [RhCl(C 8H 1 4) 2] 2 - £ g p > 2[RhClx(DMA) (6,8) [RhCl (DMA) 1 ( X 1 ) _ + 0. O.RhCl (DMA) ( x ~ 1 ) " (6.9) 7 2 2 x y The slower reaction was due to a subsequent oxidation of substrates, e.g. cyclooctene, and DMA solvent: [0 2RhCl x(DMA) y] ( x 1 ) " + substrate [RhClx(DMA)y] ( x 1 ) _ + product (6.10) - 204 -Reaction (6.10) was fast and complete within one minute under 1 atm. of 0^ at 35-80° (7). Under the same conditions, the analogous Ir-complex 1_takes about one hour for the 1:1 uptake reaction with oxygen. The g order of reactivity toward 0 2 for this pair of d -metal complexes i s g thus Rh > Ir, the opposite of the other two d series mentioned above. However, there are differences i n the series of complexes. The [M(2-phos) 2] + complexes are univalent cations, the [M(C0)C1 (PPh.^] complexes are neutral, while the [MC1(C QH1.).]. complexes are very l i k e l y o 1 4 L 2 negatively charged anions in solution. Further more, the Rh species i n (x-D-solution is believed to be [RhCl (DMA) ] , while the Ir-complex i s x y believed to be [IrCl_(C QH ) ] . Hence i t i s very d i f f i c u l t to make a 2. o 1 4 / direct comparison for the three series of systems. Qualitatively the cyclooctene ligand w i l l be much less basic than DMA, and such ligand effects could well explain the relatively low activity of the [IrCl-CC-H^)-]" species. Catalytic oxidations using molecular oxygen are of intense current interest (5). Two mechanistic classifications of such reactions have emerged. One involves an overall s p l i t t i n g of a coordinated 0 2 molecule with subsequent oxygen atom transfer; the second involves formation of peroxide and radical chain process. Examples of the f i r s t type are found in the work of Halpern and coworkers (21) and Ugo and coworkers (31) who have reported on the homogeneous catalytic oxygenation of PPh^ to triphenylphosphine oxide, OPPh^, using a benzene solution of Pt(PPh^) 4 with excess of PPh^ under an 0 2 atm. Rate-laws and mechanisms were proposed. Ru phosphine complexes (22) are also reported to be effective catalysts for oxidation of PPh^ to OPPh^ by a similar oxygen atom transfer - 205 -mechanism. Peroxide radical mechanisms have been invoked by Japanese workers (23) who have reported on the oxidation of styrene and of PPh^ catalyzed by various Vaska-type iridium complexes, Ir(C0)X(PPh.j)2 (X = Cl, Br and I). The oxidation of cyclohexene catalyzed by low-oxidation state transition metal phosphine complexes has been similarly discussed (24-26). In the present work, cumene is catalytically oxidized by 1_ i n DMA under mild conditions (1 atm 0^ at 25°), while other substrates, e.g. cyclooctene and hexene-1, are not catalytically oxidized. Stern (11) has suggested that the autoxidation of cumene catalyzed by Palladium(O) phosphine complexes proceeds via a hydroperoxide free radical reaction initiated by hydrogen abstraction by a molecular oxygen complex, and James and coworkers (6) invoked the same mechanism for the Rh(I) catalyzed oxidations outlined i n scheme (6.11). The same reaction scheme is believed to be involved i n the present Ir(I) system. Ir(I) + 0 2 Ir(II) 0 2 Ir(II)...0 2 • + RH •> oxidation product R* + 0 2 — > R02» RO 2 + RH RO„H + R-(Scheme 6.11) - 206 -Consistent with a postulated slow catalytic oxidation is the observation that the green oxygenated DMA solution of JL kept in a sealed esr tube changed into orange over a period of about 12 hours. The solvent i s believed to be catalytically oxidized slowly: Ir(I)(0 2) + solvent > oxidized products + Ir(I) Ir(I) + 0 2 > Ir(I)(0 2) (Scheme 6.12) After a l l the 0 2 i s consumed, the catalytic cycle w i l l cease and the 'Ir' deoxygenated species w i l l remain i n solution. The fact that further exposure to a i r regenerates the green solution indicates that the deoxy-genated 'Ir' species i s the i n i t i a l starting complex believed to be [IrCl 2(CgEL^) 2] . The non-oxidation of the cyclooctene i n the Ir system could be related to the stronger bonding of the olefin compared to the Rh system, where the cycloolef i n i c ligand i s readily oxidized to cyclooctenone. The very slow 0 2 uptake by the "blank" DMA solution, i.e. , without the presence of Ir-complex JL, and the subsequent slow uptake beyond the 1:1 stage in the presence of the complex, are almost certainly due to the same catalytic oxidation of the solvent by traces of peroxide present (10). The rate of the "blank" reaction i s very similar to that reported by Ng (10) for the corresponding [RhCl(C0IL .)-] system. The Rh catalyzed solvent oxidation occurred at a rate of —6 —1 —2 ^ 4 x 10 M sec (02~uptake) at 10 M and 80°; extrapolation of —8 —1 the reported data to 25° gives a rate of VL0 M sec compared to a measured rate of ^ 2 x 10 7 M sec * for the iridium system studied here - 207 -at corresponding conditions. In benzene solution, complex _1_ remains as a dimer, and the 0^ uptake in this solvent gives mole ratio of [0 2 ] : [ l r ] = 0.5-1.2 in a rather irreproducible manner (Table VI-4). The i . r . data for the isolated green complex sometimes indicated the presence of peroxide and/or hydroperoxide groupings consistent with a scheme such as (6.11). The elemental C and H analysis i s consistent with the formation of IrCl(C 0H .)(0 ); however, some variation i n oxygen content would not o 14 z greatly affect the C, H elemental analysis, although the data are consistent with only one C_H-. moiety per Ir. It might be mentioned o 14 that oxygenation of RhCl(PPh 3) 3 is very solvent dependent and rapid uptakes of between 1 to 3 moles 0^ per mole Rh are observed, besides subsequent slow oxidation of solvent i n the case of alcohol solvents (27). Complexes isolated include Rh0(Ph»P),,Clo0_(C.H-)„ from benzene, 3 2 z D 6 o 2 Rh 0(Ph 0P).Cl o0 c from methanol, and [Rh(Ho0)(Ph.P)Cl(O)] from ethanol, z j 4 z J z .3 n although the characterization of these complexes does not appear to be well substantiated. A variety of 1:1 molecular oxygen complexes, both solvated and non-solvated have also been formulated as resulting from 0^ absorption by solutions of RhCl(PPh 3) 2 (29). Clearly the chemistry of the formation of apparently "simple" molecular oxygen complexes remains to be more f u l l y elucidated. In some related [Co(II)(salen)]-0 2 systems (28), the number of moles of 0 2 absorbed for each mole of Co changes from 0 to 1 depending on the pressure of 0 2 > as well as temperature; equilibria such as the following are involved. - 208 -Co(II) + 0 2 ^ K v Co0 2 (6.13) Co0o + Co (II) -k—*• Co0oCo (6.14) At lower temperaturesK i s relatively large and k i s small and a 1:1 uptake i s approached at 1 atm although less than 1 mole i s absorbed at lower pressures depending on K; at higher temperatures, reaction (6.14) occurs and the overall stoichiometry i s 0.5:1. Whether reactions of these types occur in the Ir(l)-0 2~benzene system and whether some kind of dimer ^± monomer equilbrium is involved remains to be established. It i s worth noting that the following reaction scheme would also account for the observed reaction kinetics for the Ir(I)-0 2-LiCl-DMA system: I r ( I ) + o,-*-* l r ( 0 9 ) . ^ > p r ° d u C t ( 6 ' 1 5 ) 2 ^  1 2 slow decomposition The general rate-law for 0 2 uptake i s k 2K[Ir][0 2]/(1 + K[0 2J); i f K[0 2] « 1, i.e. the K equilibrium l i e s well to the l e f t , the rate-law takes the limiting form k 2K[Ir][0 2] and the rate i s f i r s t order in [Ir] and [0 2]. However such a scheme i s considered less l i k e l y . A rate-determining decomposition of an l r ( 0 2 ) complex would have to involve formation of the green solution accompanied by production of an esr signal. Plausible slow steps include the following l r ( 0 2 ) S l ° W > Ir(II) + 02". (6.16) or l r ( 0 2 ) > Ir(II)(0 2H) —5> Ir(I) + 'C^ H (6.17) - 209 -The ' i n i t i a l ' overall 1:1 stoichiometry implies that reactions subsequent to formation of (Ir(II) +02 *) or Ir(II)(02H) must be much slower than their formation. A slow regeneration of Ir(I) is indeed indicated by the experiments involving a limiting O2 supply (see p. 185). However, the measured activation parameters for the 1:1 uptake stoichiometry are very similar to those reported for systems that definitely involve formation of Ir(I)-02 complexes by a rate-determining step involving O2 uptake. Further, the low intensities of the esr signals argue somewhat against stoichiometric formation of paramagnetic species. Formation of a 1:1 Rh(I)02 complex in a closely related Rh 3 - 1 system (K ^ 5 x 10 M at room temperature) adds indirect support for l i k e l y formation of a similar Ir02 complex. For such Ir-complex formation, 4 -1 K must be > 10 M at 25° (since the 1:1 stoichiometry is reached even at the lowest oxygen pressure of 80 mm Hg), and thus far, where data are available (20), Ir systems bind O2 more strongly than corresponding Rh ones. - 210 -REFERENCES L. Vaska, Science, 140, 809 (1963). S.J. La Placa and J.A. Ibers, J. Amer. Chem. Soc, 87_, 2 5 8 1 (1965). J.A. McGinnety, R.J. Doedens and J.A. Ibers, Inorg. Chem., 6^, 2243 (1967). J.A. McGinnety, N.C. Paynes and J.A. Ibers, J. Amer. Chem. Soc, 91, 6301 (1969). V.J. Choy and C.J. O'Connor,Coord. Chem. Rev., £, 145 (1972/73). B.R. James, F.T.T. Ng and E i . Ochiai, Can. J. Chem,, 50, 590 (1972). B.R. James and F.T.T. Ng, Chem. Comm., 908 (1970). B.R. James and E i . Ochiai, Can. J. Chem,, 49, 975 (1971). J.L. Herde and C.V. Senoff, Inorg. Nucl, Chem, Letters, ]_, 1029 (1971). F. T.T. Ng, Ph.D. Thesis, University of Br i t i s h Columbia, July 1970, p. 189. G. E. Stern, Chem. Comm., 736 (1970). H. van Gaal, H.G.A.M. Cuppers and A. van der Ent, Chem, Comm,, 1694 (1970); A. van der Ent and T.C. van Soest, Chem. Comm., 225 (1970). L. Vaska and L.S. Chen, Chem. Comm., 1080 (1971), L. Vaska, Accounts Chem. Res., 1_, 335 (1968). G.N. Schranzer and L.P. Lee, J. Amer. Chem. Soc, 92, 1551 (1970). A. Misono, S. Koda and Y. Uchida, Bull, Chem, Soc, Japan, 42_, 580, 3470 (1969); A. Misono and S. Koda, Bull, Chem, Soc, Japan, 42, 3048 (1969). J.H. Bayton, N.K. King, F.O. Looney, and M.E, Winfield, J, Amer,, Chem. Soc, 91, 2775 (1969). B. M. Hoffmann, D.L. Diemente and F. Basolo, J, Amer, Chem, Soc, 92, 61 (1970). - 211 -19. H.J. Keller and H. Wawersik, J. Organomet. Chem., 8, 185 (1967). 20. L. Vaska, L.S. Chen and W.V. Miller, J. Amer. Chem. Soc, 93, 6671 (1971). 21. J.P. Birk, J. Halpern, and A.L. Pickard, J. Amer. Chem. Soc, 90, 4491 (1968). 22. B.W. Graham, K.P. Laing, C.J. O'Connor, and W.R. Roper, J. Chem. Soc, Dalton, 1237 (1972); S. Cenini, A. Fusi and G. Capparella, J. Inorg. Nucl. Chem., 33, 3576 (1971). 23. K. Takao, Y. Fujiwara, T. Imanaka and S. Teranishi, Bull. Chem. Soc, Japan, 43, 1153 (1970). 24. J.P. Collmann, M. Kubota and J.W. Hosking, J. Amer. Chem. Soc, 89, 4809 (1967). 25. V.P. Kurkov, J.P. Pasky and J.B. Lavigne, J. Amer, Chem. Soc, 90, 4743 (1968). 26. A. Fusi, R. Ugo, F. Fox, A. Pasini and S. Cenini, J. Organomet. Chem., 26, 417 (1971). 27. R.J. Augustine and J.V. Peppen, Chem. Comm., 497 (1970). 28. E i . Ochiai, J. Inorg. Nucl. Chem., 35, 1727 (1973). 29. (a) M.C. Baird, D.N. Lawson, J.T. Mague, J.A. Osborn and G. Wilkinson, Chem. Comm. (London), 129 (1966); (b) J. Blum, J.Y, Becker, H. Rosenman and E.D. Bergnman, J. Chem. Soc, B, 1000 (1969); (c) M. Takesada, H. Tamazaki and N. Hagihava, Bull, Chem. Soc Japan, 41, 270 (1968); (d) J. Blum, H. Rosenman and E.D. Bergmann, Tetrahedron Letters, 3665 (1967). 30. A. van der Ent and A.L..Onderdelinden, Inorg. Syn., 14_, 92 (1973). 31. R. Ugo, G. La Monica, F. Cariati, S. Cenini and F, Conti, Inorg. Chim. Acta, U, 390 (1970). 32. B .R. James, private communication. CHAPTER VII SYNTHESIS OF SOME IRIDIUM CARBONYL COMPLEXES VII.l Introduction Few metal carbonylVcan be prepared as simply as was tetracarbonyl nickel(O) by Mond, Langer and Quincke in 1891 (1). Most preparative methods for metal carbonyls require high pressure equipment. Clearly the development of alternative syntheses which operate at atmospheric pressure i s desirable. Since the oxidation state of a metal atom i n a metal carbonyl is lower than that in the complex from which i t i s derived, a l l carbonylations of metal compounds are reductions. Therefore, most syntheses depend upon reducing a transition-metal compound i n the presence of CO under pressure. Common reducing agents are sodium, aluminum alkyIs, or carbon monoxide i t s e l f , and the latter sometimes mixed with hydrogen. Such reactions are termed as "reduction carbonyla-tions" (2). Mechanistic studies on these syntheses are lacking, but i t is reasonable to assume that successive reduction steps are accompanied by simultaneous coordination of CO groups(3). In recent years, syntheses of a number of platinum metal carbonyls using one atmospheric CO pressure have been reported (4-7). Following success i n this laboratory in synthesizing a number of ruthenium and - 213 -rhodium carbonyls and chloro carbonyls from commercially available chlorides using ambient CO pressures (4), some corresponding but preliminary studies using iridium chlorides have been carried out, and are presented in this chapter. VII.2 Tetraphenylarsonium tetrachlorodicarbonyliridate(III) The product of carbonylation of IrCI^*311^0 (in solid state) was f i r s t reported by Hieber et a l . (8) as "IrX(C0) 3". Carbonylation of RhCl3*3H20 leads to the well-characterized [Rh(CO) 2Cl] 2 or Rh(C0) 2Cl 2~ species (9,10) depending on the conditions. Chatt et a l . (11) prepared a series of iridium(I) phosphine complexes by adding phosphines to a carbonylated 2-methoxyethanol solution containing Ir(III) or Ir(IV) halo-complexes. Forster (12) has later reported that the carbonylation of IrX 3'3H 20 in methoxy-ethanol/H20 media leads to [IrX 2 (CO),,]-. In the present work about 0.3 gm of IrCl 3"3H 20 was dissolved in 200 ml of 0.1 N HC1 solution. The solution was refluxed at i> 80° while a CO stream at atmospheric pressure was bubbled through. The solution was i n i t i a l l y dark-brown but changed into light-brown after about half an hour. After one day, some black solid precipitated (which was identified as iridium metal) and a pale yellow f i l t r a t e resulted. After f i l t r a t i o n , the yellow solution was pumped to dryness. The1 dark brown oi l y solid obtained was washed with benzene and then redissolved in 50 ml of H20 in a i r . Ph^AsCl (y 0.3 gm) was added to the solution after warming. Orange crystals precipitated on cooling overnight. The crystals were collected and washed with H20, On drying in vacuo, the crystals disintegrated into an orange powder. - 214 -Partial elementary analysis of the orange powder gave the following results: [Ir(CO) 2C1 4]"[Ph^As] +-2H 20. C a l c : C: 38.6%; H: 2.97%; C l : 17.6%. Found: C: 39.20%; H: 3.03%; Cl: 17.69%. m.pt. (in a i r ) : 242°C. The i . r . (in nujol) showed stretches at 2055 and 2070 cm"1, v T at 310, 320 cm - 1 and a broad and small IrCl v. „ at 3300 cm"1. Yield ^ 5%. U—n Since carbonylation of Rh(III) halides to give Rh(C0) 2X 2~ (X = halides) in aqueous media apparently proceeds (13) through reaction of water with a Rh(III)-carbonyl, eq. (7.1), a similar mechanism Rh(III) + 3C0 + H20 ) Rh(I)(C0) 2 + C0 2 + 2H + (7.1) might be involved for the iridium system, although Forster has suggested (12) that [ I r ( C 0 ) 2 X 4 f i s in equilibrium with [Ir(C0) 2X 2]~ and X 2 i n alcoholic media, eq. (7.2). Ir(C0) 2X 4~ =^==- Ir(C0) 2X 2" + X 2 (7.2) I.r. evidence was presented for the existence of the iridium (III) dicarbonyl in solution (v :2140 and 2088 cm ^) although this complex was not isolated. No evidence was obtained for such an equilibrium in the present work. Reduction of Ir(III) carbonyls according to eq, (7,1) could be much slower for the highly substitution inert Ir(III) centre (14). The corresponding tetraiodocarbonyl complex has been made previously (v C Q:2110, 2068 cm - 1) (12). - 215 -VIII.3 An Iridium Carbonyl [Ir(CO)J ? 2 n The synthesis of Ir^CCO)^ w a s f i r s t reported by Hieber and Lagally (15) who prepared the carbonyl from IrX-j and CO in the presence of a halogen acceptor (Cu or Ag) at high pressure (350 atm ) and temperatures. The synthesis and yield were greatly improved (16) on using a water-2-soluble form of IrX„. A polynuclear [Ir.(C0) 1 C] anion and a mono-j o l j nuclear [Ir(CO).] anion, as well as [ Ir,. (CO)., , ] were later obtained H o lb (17) from I r 4 ( C 0 ) 1 2 in THF by reaction with 1 atm CO. The following studies describe a possible route for synthesis of what appears to be a new iridium(O) carbonyl cluster compound, although the results are somewhat inconclusive due to failure to reproduce the synthesis. About 0.25 gm of (NH.)_[IrCl.] was dissolved in 25 ml of water and a constant stream of CO was bubbled through at R.T, The i n i t i a l l y dark brown solution slowly became pale yellow and after one day, some blue solid was observed floating at the top of the yellow solution, The blue solid was fi l t e r e d , washed with water and dried i n vacuo. Elementary analysis gave the following results: [ I r ( C 0 ) 2 ] n : C a l c : C: 9.7%; H: 0%; C l : 0%, Found: C: 10.01%; H: 0%; Cl: 0%, Yield: ^30%. The i . r . spectrum (in nujol) gave CO frequencies at 2020(S), 2050(S) and 2105 (S) cm"1 with a shoulder at 2060 cm"1. When heated i n air , the solid became black at 170° but did not melt even up to 360°; i t possibly decomposes into iridium metal. The pale yellow f i l t r a t e resulting from the reaction was strongly acidic, indicating that protons were produced in the carbonylation process, as expected. The blue carbonyl did not dissolve in most common laboratory solvents, e.g. - 216 -CHCl^, CCl^, acetone, cyclohexane, benzene and ethanol. The evidence of the formulation of the blue compound as [Ir(CO)„] 2 n is certainly weak; the compound contains no chloride, and apparently no hydride although a hydride carbonyl cluster cannot be entirely ruled out. The carbon analysis for [IrCCO^J^ is 0.3% low, and i t is interesting to note that the carbon analysis reported for Ir^CCO)^ was also 0.3% low! A quite plausible structure has been suggested by King (18) for a cluster i n which n = 12, and i n which there are no bridging carbonyls (as indicated by the i . r . data): an icosahedron of iridium atoms with two terminal carbonyls bonded to each iridium atom allows each iridium to acquire the necessary nine electrons to form the favored rare-gas electronic configuration since five edges come together at each vertex (the five iridium-iridium bonds then account for 5 of the required electrons). It i s unfortunate, however, that the blue carbonyl synthesis could not be reproduced. Different batches of (NH,) .[IrCl^.], and different 4 2 6 cylinders of CO as well as conditions of varying pH were tried but a l l attempts to reproduce the synthesis fa i l e d . K^IrClg was also tried as starting material, but this was found to be completely unreactive towards CO i n aqueous solution at temperature up to 80°. i ^ - 217 -REFERENCES 1. (a) L. Mond and F. Quincke, J . Chem. News, 63, 301 (1891); (b) L. Mond and C. Langer, J. Chem. Soc, 1090 (1891). 2. H.E. Podall, J. Chem. Ed., 38, 187 (1961). 3. E.W. Abel and F.G.A. Stone, Quart. Rev., 24_, 498 (1970). 4. B.R. James and G.L. Rempel, Chem. Ind. (London), 37, 1036 (1971). 5. J.L. Dawes and J.D. Holmes, Inorg. Nucl. Chem. Letters, _7, 847 (1971). 6. P. Chini and S. Martinengo, Inorg. Chim. Acta, _3, 315 (1969). 7. C.W. Bradford, Plat. Met. Rev., 16, 50 (1972). 8. W. Hieber, H. Lagally and A. Mayr, Z. Anorg. A l l g . Chem., 246, 138 (1941). 9. J.A. McCleverty and G. Wilkinson, Inorg. Synth., 8, 211 (1966). 10. B.R. James, G.L. Rempel and F.T.T. Ng, J. Chem. Soc, (A), 2454 (1969). 11. J. Chatt, N.P. Johnson, B.L. Shaw, J. Chem. Soc, (A), 604 (1967). 12. D. Forster, Inorg. Nucl. Chem. Letters, _5, 433 (1969). 13. B.R. James and G.L. Rempel, Chem. Comm., 158 (1967). 14. F. Basolo and R.G. Pearson, Mechanism of Inorganic Reactions, John-Wiley, New York, 1967, Chapter III. 15. W. Hieber and H. Lagally, Z. Anorg. Chem., 245, 321 (1940). 16. S.H.H. Chaston and F.G.A. Stone, J. Chem. Soc, (A), 500 (1969). 17. L. Malatesta, G. Caglio and M. Angoletta, Chem. Comm., 532 (1970). 18. R.B. King, private communication to B.R. James. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0061133/manifest

Comment

Related Items