UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Activation of carbon monoxide, hydrogen and oxygen by rhodium halide complexes in solution Rosenberg, George Nathan 1974

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

Item Metadata

Download

Media
831-UBC_1974_A1 R68.pdf [ 9.41MB ]
Metadata
JSON: 831-1.0061063.json
JSON-LD: 831-1.0061063-ld.json
RDF/XML (Pretty): 831-1.0061063-rdf.xml
RDF/JSON: 831-1.0061063-rdf.json
Turtle: 831-1.0061063-turtle.txt
N-Triples: 831-1.0061063-rdf-ntriples.txt
Original Record: 831-1.0061063-source.json
Full Text
831-1.0061063-fulltext.txt
Citation
831-1.0061063.ris

Full Text

ACTIVATION OF CARBON MONOXIDE, HYDROGEN, AND OXYGEN BY RHODIUM HAL IDE COMPLEXES IN SOLUTION BY GEORGE NATHAN ROSENBERG B.Sc. (Hons.) University of British Columbia, 1969 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of Chemistry We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA July, 1974 In presenting th is thes is in pa r t i a l fu l f i lment of the requirements for an advanced degree at the Un ivers 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 l ab le for reference and study. I fur ther agree that permission for extensive copying of th is thes is for scho la r ly purposes may be granted by the Head of my Department or by his representat ives . It is understood that copying or pub l ica t ion of th is thes is fo r f inanc ia l gain sha l l not be allowed without my wri t ten permission. Department of CHEMISTRY The Univers i ty of B r i t i s h Columbia Vancouver 8, Canada Date S e p t e m b e r 2 3 , 1974 i i ABSTRACT Kinetic studies are described for the activation of carbon monoxide, molecular hydrogen, and molecular oxygen by rhodium halide complexes in solution. Carbon monoxide reacts with bromorhodate(III) and bromocarbon-ylrhodate(III) complexes to produce the anionic species, [Rh(C0) 2Br 2]" # The reaction of [RhBr^ Ch^ O),,]'•" in acid bromide solutions involves i n i t i a l formation of a Rh^*(C0) substitution product, which then under-goes reductive carbonylation: CO H2° Rh 1 1 1 ^Rh H I(C0) ^[RhCOOH]—*RhT + C0o + 2H+ (1) K l I CO , Rh i(C0) 2 (2) Further production of Rh* is autocatalytic due to a more effective reduction via a CO independent path that involves a [Rh^'-'Br'-'Rh^CO^] bridged intermediate. Reductive carbonylation of the pentabromocarbonyl-rhodate(III) complex in aqueous HBr solution proceeds in a similar manner. Decomposition of Rh^^CO) by water to produce Rh1 species (equations 1 and 2) results in the autocatalytic formation of [Rh(C0)2Br2] ~ through a step that again involves a bromide-bridged intermediate, [(CO)Rh***••*Br •••RhI(C0)2] . Carbon monoxide can be catalytically activated by rhodium(III) for the reduction of substrates such as ferric ion. Kinetic studies sug-gest a mechanism involving reduction of [Rh(C0)Br5] by water. Data are 111 consistent with the following scheme: [Rh n i(C0)Br 5] 2" K l I + YQ »• Rh + C02 + 2H k„\^C0,Br" 2Fe III R h H I + ZFe 1 1 (3) Acid solutions of RhBrg*2H20 activate for the reduction of .III Fe . Formation of a metal hydride in the rate-determining step appears to involve substitution by an associative mechanism: Rh I HBr~ + H, v l -> Rh I HH" + H+ + Br" (4) Activation parameters for the reaction of [RhBr 4(H 20) 2]" with H2 are similar to those found for the corresponding reaction with CO (equation 1) The reaction of [Rh(C0) 2Cl^ ~ with 0 2 in 3M HC1 i n i t i a l l y 111 2 forms [Rh (C0)C151 by a path thought to involve formation of an oxygen complex: Rh 1(C0) 2 + 02- [Rh(C0)2(02)] -> [Rh I H(C0)(C0 3)] 2H' (5) III, Rh 1 A 1(C0) + C02 + H20 III, The observed kinetics indicate that further Rh (CO) is autocatalytically produced according to the following sequence: iv RhIII(CO)+RhI(CO)2 [(C0)Rh I I I---Cl---Rh I(C0) 2]—»Rh 1 (CO)+Rh H I(C0) 2 Rh I H(CO) 2- ^ ^ R h I ( C Q ) + C 0 2 + 2 H + ( 7) 2Rh*(C0) + 0 2 — ^ ?*2Rh***(C0) + 2H20 (8) Subsequent slower reaction to give chlororhodate(III) species 2-occurs via slow hydration of [Rh(C0)Clg] k3 Rh n I(C0)—q-Q *• Rh1 + C02 + 2H+ (9) RhI + l / 2 ' 0 2 — — > Rh 1 1 1 + H20 (10) Studies on the oxidation of [Rh(CO)2C12]~ in LiCl/DMA by molecular 0 2 suggest the involvement of hydroperoxide free radicals for the production of Rh**. Evidence for a carbonato intermediate formed during the oxidation process has been obtained. The Rh** solutions were found to contain active, though inefficient, catalysts for the autoxida-tion of DMA. The presence of Rh** - 0 2 has been detected by esr, and the superoxide species may be responsible for the autoxidation reaction. A paramagnetic chloride-bridged rhodium(II) anion [Rh0Clc(DMA)0] was L 0 C isolated from the oxygenated DMA solutions of [Rh(C0) 2Cl 2]" containing excess chloride. V TABLE OF CONTENTS Page ABSTRACT , , i i TABLE OF CONTENTS v LIST OF TABLES x i LIST OF FIGURES xiv ABBREVIATIONS x i x ACKNOWLEDGEMENTS x x i i Chapter I. INTRODUCTION 1 1.1. Aim of Work 1 1.2. Carbonyl Halides of Rhodium 4 1.3. Activation of Molecular Hydrogen 6 1.4. Reduction of Inorganic Substrates 7 1.5. Bonding in Molecular Oxygen Complexes 8 1.6. Metal Catalyzed Oxidation Reactions Involving Molecular Oxygen 9 1.7. Rhodium(II) Complexes H Chapter I I . APPARATUS AND EXPERIMENTAL PROCEDURE 14 2.1. Materials 14 2.1.1. Rhodium Compounds 14 2.1.2. Gases 14 2.1.3. Solvents 15 2.1.4. Other Materials 15 2.2. Apparatus for Constant Pressure Gas-Uptake and Evoluti.'OmMeasurements 15 vi Pacie 2.3 Procedure for a Typical Gas-Uptake or Evolution Experiment I 7 2.4. Gas Solubility Measurements , 19 2.5. Spectrophotometry Kinetic Measurements 19 2.6. Reaction Product Analysis 2 0 2.6.1. Gaseous Products 20 2.6.2. Liquid Products 20 2.6.3. Solid Inorganic Products 21 2.7. Instrumentation 21 Chapter II I . CARB0NYLATI0N OF RHODIUM(III) BROMIDE COMPLEXES IN AQUEOUS HYDROBROMIC ACID SOLUTION 22 3.1. Introduction 22 3.2. The Reaction of RhBr3i2H20 in HBr with CO 22 3.3. Kinetics of the Reaction in HBr Media 25 3.3.1. The I n i t i a l Reaction 28 3.3.2. The Reaction Involving the Autocatalytic Species 29 3.4. Effect of Bromide and Acid 43 3.5. Discussion 51 3.5.1. Production of the Autocatalytic Species 51 3.5.2. Autocatalytic Reaction 53 Chapter IV. CARBONYLATION OF A RHODIUM(III) BROMOCARBONYL COMPLEX IN AQUEOUS HYDROBROMIC ACID SOLUTION 59 4.1. Introduction 59 4.2. The Reaction of [Rh(C0)Br 5] 2" in HBr Solution with CO 59 v i i Page 4.3. Kinetics of the Reaction in HBr Media ,.... 60 4.4. The In i t i a l Reaction 6 6 4.5. Kinetics of the I n i t i a l Reaction in the Presence of Fe 1 1 1- 68 4.6. Carbonylation of Bromorhodate(III) Complexes after Reduction of Fe 1 1 1 74 4.7. Discussion 77 4.7.1. Reduction of the [Rh(C0)Br 5] 2" by Water 77 4.7.2. The Autocatalytic Reaction ((Rh(C0)Br5]2"/C0).. 80 4.7.3. Carbonylation of Bromorhodate(III) Complexes After Reduction of F e 1 1 1 82 4.8. Reproducibility of Kinetic Results 82 Chapter V. CATALYTIC ACTIVATION OF MOLECULAR HYDROGEN BY BROMORHODATE(III) COMPLEXES IN AQUEOUS HYDROBROMIC ACID SOLUTION 85 5.1. Introduction 85 5.2. Catalytic Wn Reduction of F e 1 1 1 by [RhBr 4(H 20) 2]" .... 85 5.3. Discussion 87 5.3.1. Catalytic Reduction of F e 1 1 1 by [RhBr 4(H 20) 2]' 87 Chapter VI. THE REACTION OF MOLECULAR OXYGEN WITH DICHLORODI-CARBONYLRHODATE(I) IN AQUEOUS ACID SOLUTION 96 6.1. Introduction 96 6.2. The Reaction of [Rh(C0) 2Cl 2]" in Aqueous HC1 Solution with 0 o 96 v i i i Pac[e 6.3. Kinetics of the Formation of Rh I H(C0) According to Reaction (6.1) , 105 6.4. The Autocatalytic Reaction 108 6.5. Decomposition of [Rh(C0)Clg]2" in 3M HCl 119 6.6. Discussion 127 6.6.1. Formation of Rh H I(C0) According to Reaction (6.1) 127 6.6.2. The Autocatalytic Reaction 135 6.6.3. Decomposition of [Rh H I(C0)Cl5] 2" 140 Chapter VII. THE REACTION OF MOLECULAR OXYGEN WITH DICHLORDI-CARBONYLRHODATE(I) IN DMA 144 7.1. Introduction 144 7.2. The Reaction of [Rh(C0) 2Cl] 2 in LiCl/DMA with 0 2 144 7.2.1. Stage I: Induction Period 149 7.2.2. Stage I I : Stoichiometric Absorption of 0 2 ... 156 7.2.3. Stage III: Catalytic Oxidation 158 7.3. Kinetics of the Reaction of [Rh(C0) 2Cl] 2 with 0 2 in LiCl /DMA "159 7.4. Preparation and Characterization of [RhC03(H20)20H]2 . 165 7.5. Discussion 166 7.5.1. Induction Period 168 7.5.2. Stoichiometric Oxidation 170 7.5.3. Catalytic Oxidation 177 ix Page Chapter VIII. PREPARATION AND CHARACTERIZATION OF [Ph 4As] 2[Rh 2Cl 6(DMA) 2] AND [Ph 4As] g[Rh 2Cl g]. 181 8.1. Introduction 181 8.2. Experimental I 8 2 8.2.1. Preparation of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] 1 8 2 8.2.2. Preparation of [Ph 4As] 3[Rh 2Cl g] 1 8 3 8.2.3. Reaction of [Ph 4As] 2[Rh 2Cl g(DMA) 2] with HOAc/NaOAc 1 8 3 8.2.4. Oxidation of [Rh 2Cl 6(DMA) 2] 2~ 184 8.2.5. Attempts at Growing Crystals of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] 184 8.2.6. Esr of Rh 1 1 in a [Ph 4As] 3[Rh 2Cl g] Host Lattice . 184 8.3. Discussion 185 8.3.1. Preparation of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] 185 8.3.2. Preparation of [Ph 4As] 3[Rh 2Cl g] 186 8.3.3. Infrared Spectra of [Ph 4As] 3[Rh 2Cl g] 187 8.3.4. Infrared Spectra of i[.P.h4As]2[Rh2Clg(DMA)2] ... 189 8.3.5. Electrolytic Conductances of .-:\..< • [Ph 4As] 2[Rh 2Cl 6(DMA) 2] and [Ph 4As] 3[Rh 2Cl g] in DMA I 9 0 8.3.6. Molecular Weights of [Ph4As]2[Rh2Clg(DMA)2] and tPh 4As] 3[Rh 2Cl g] in Solution 192 8.3.7. Visible Spectra of [Rh 2Cl g] 3' 192 8.3.8. Visible Spectra of [Rh2Clg(DMA)2]2" 196 X Page 8.3.9. Magnetic Properties and Esr of [Ph 4As] 2[Rh 2Cl 6(DMA) 2l 199 8.3.10. Structure of the [Rh 2Cl g] 3" Ion 206 8.3.11. Possible Structure for the [Rh 2Cl 6(DMA) 2] 2~ Ion 211 8.3.12. Reaction of [Rh2Clg(DMA)2]2" with HOAc/NaOAc . 213 8.3.13. Oxidation of [Rh 2Cl g(DMA) 2] 2~ in DMA 213 Chapter IX. GENERAL CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK 214 9.1. 2Formation-of-.RhI.1!1 (CO^ i-.ahd Rh}ll-Wl 2 U 9.2. Sihe-Autocatalytic Reactions;.•;455, 1 2 ^ 9.3. Decomposition of [Rh 1 1 1(C0)X 51 2" (X=Br,Cl) 221 9.4. Autoxidation of DMA 222 9.5. Complexes of the Type [Rh 2X g] 3" (X=Br,Cl) L. 224 REFERENCES 226 xi LIST OF TABLES Table Page Carbonylation of rhodiumllll'), bromide complexes in aqueous hydrobromic acid solution I Summary of kinetic data in 0.5M HBr at 40° 30 II Temperature dependence of k^  . 33 III Maximum rates in 0.5M HBr at 40° 35 IV Calculation of k^  by use of equation (3.7) 40 V Calculation of k^  by use of equation (3.10) 42 VI Temperature dependence of k^  44 VII Effect of [Br -] on the rate of reaction at 40° 47 VIII Effect of [H+] on the rates of reaction at 40° 48 Carbonylation of a rhodium(IHlI)) bromocarbonyl complex in aqueous hydrobromic acid solution IX Summary of kinetic data in 0.5M HBr at 60° 62 X Summary of kinetic data for the [RhJ 1 1 (•C-OJ'Br*]'2!, ca:tailyzed5reducti6n.of..EeIiI By CO in 0.5M HBr at.60° . 7 3 XI Summary of kinetic data for_car:b6nyl.at'i6h^of^Rh -mat 60 in 0:5M 'HBr-^v. ;.i?:.?f. . .oy.CO.. r .G.So. .?t.>?•:<:..... 75 Catalytic activation of molecular hydrogen by bromorhodate(III) complexes in aqueous hydrobromic acid solution XII Kinetic data at 40° in 0.5M HBr 89 XIII Temperature dependence of kn 91 x i i Table Page The reaction of molecular oxygen with dichlorodicarbonyl-rhodate(I) in aqueous acid solution XIV Summary of kinetic data in 3M HCl 109 XV In i t i a l and maximum rates in 3M HCl 115 XVI Ratio of maximum: i n i t i a l rates in 3M HCl 116 XVII Summary of kinetic data in 3M HCl at 55° 121 XVIII 0eefl.mp9Sjtip.n1-,[IShcI:CQQ)^l;5]§" in.,3M HCl at 70° 124 XIX Effect of [CI"] on the rate of reaction of Rh I H(C0) at 70° 128 XX Effect of [H+] on the rate of reaction of Rh I H(C0) at 70° 128 XXI Temperature dependence of kg in 3M HCl 131 The reaction of molecular oxygen with dichlorodicarbonylrhodate(I) in DMA XXII Summary of kinetic data at 75° 151 XXIII Effect of [Cl~] on the rate at 75° 152 XXIV Efefieetsof."additives on the rate at 75° 155 XXV Temperature dependence of 163 Preparation and characterization of [Ph^As^] [Rh2Clg(DMA)2] and [Ph 4As] 3[Rh 2Cl g] XXVI Infrared spectra of [Ph4As23,[Rh2Cl6(DMA)2] and [Ph 4As] 3[Rh 2Cl g] 188 XXVII Electrical conductances of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] and [Ph 4As] 3[Rh 2Cl g] in DMA at 23° 191 XXVIII Molecular weights of [Ph 4As] 2[Rh 2Cl g(DMA) 2] and [Ph.As]~[Rh9ClQ] in solution 193 x i i i Table Page XXIX Visible spectral data for [RhCl g(DMA) 2] 2" and [Rh 2Cl g] 3" 194 XXX Molar susceptibility and magnetic moment of [Ph^As]^ [Rh2Cl6(DMA)2] 200 XXXI Esr spectra of [Pf^As] 2[Rh 2Cl 6(DMA) 2] 202 XXXIIC Principal bond distances in [Ph4As].3[Rh2Clg] •.. 209 XXXIII Principal bond angles for [![Rh4As]3[Rh2Clg] 210 General conclusions and recommendations for future work XXXIV Summary of kinetic data for the reaction of CO and H2 with some rhodium(III) complexes 215 XXXV Summary of kinetic data for autocatal-yttc- reactions^,.,,, 2">8 XXXVI Summary of kinetic data for the decomposition of Rh***(C0) species at 60° 223 xiv LIST OF FIGURES Figure Page Apparatus and experimental procedure 1. Apparatus for constant pressure gas-uptake or evolution measurements 16 Carbonylation of rhodium(III) bromide complexes in aqueous hydrobromic acid solution 2. Absorption spectrum.'" RhBr3'-2H20 in 0.5M HBr 23 3. Absorption spectrum.[Rh(C0)2Br2]" in 0.5M HBr 24 4. Rate plots for the autocatalytic reduction of Rh*** at 40° 26 5. Typical plots of -d[C0]/dt vs.[Rh 1 1 1] at 40° 27 6. Dependence of i n i t i a l rate on [Rh**1] at 40° 31 7. Effect of CO pressure on the i n i t i a l rate at 40° 32 8. Arrhenius plot for reaction of Rh*** with CO 34 9. Dependence of maximum rate on the product [Rh***]x [Rh*] at 40° 36 10. Effect of CO pressure on maximum rate at 40° 38 11. Plots of log [[Rh*(C0)2Br2]"] - log [Rh***] vs. time various CO pressures at 40° 41 12. Arrhenius plot for reduction of Rh*** by [Rh*(C0) 2Br 2f in the presence of CO 45 13. Inverse acid dependence of k^  at 40° 49 14. Inverse acid dependence of k 9 at 40° 50 XV Figure Page Carbonylation of a rhodium(III) bromocarbonyl complex in aqueous hydrobromic acid solution 15. Rate plots for the reduction of [Rh(C0)Br5] at 60° ..... 61 16. Dependence of maximum rate on [Rh**1]x[Rh*] at 60° 64 17. Effect of CO pressure on maximum rate at 60° ........ . 65 18. Plots of log [[Rh*(C0) 2Br 2]"] - log [Rh***(C0)] vs. time for various i n i t i a l Rh***(C0) concentrations at 60° 67 19. Rate plots for the [Rh(C0)Br 5] 2" catalyzed reduction of Fe*** at 60° 69 20. Dependence of i n i t i a l rate (Fe*** reduction) on [Rh(C0)Br 5] 2' at 60° 71 21. Effect of CO pressure on the rate of Fe*** reduction at 60° 72 22. Effect of CO pressure on the rate of Rh*** carbonylation at 60° 76 Catalytic activation of molecular hydrogen by bromorhodate(III) complexes in aqueous hydrobromic acid solution 23. Rate plots for Rh*** catalyzed reduction of Fe 1 1* by H2 .. 86 24. Dependence of rate of H2 uptake on [Rh***] at 40° 89 25. Effect of H2 pressure on rate of H2 uptake at 40° 90 26. Arrhenius plot for Rh*** catalyzed reduction of Fe*** by H 2 in 0.5M HBr 92 The reaction of molecular oxygen with dichlorodicarbonyl-rhodate(I) in aqueous acid solution 27. Absorption spectrum of [Rh(CO) 0Cl 9]" in 3M HCl 97 xv i Figure Page 28. Absorption spectrum of [Rh(C0)CT5]2" in 3M HC1 .......... 99 29. Uptake plots for the oxidation of [Rh(C0)^Cl 2]" in the presence of a sodalime tube at 70° 100 30. Absorption spectrum of rhodium(III) chloro species in 3M HC1 102 31. Kinetic plot for the loss of Rh H I(C0) at 70° , 104 32. Kinetic plot for the loss of [Rh(C0^Cl 2l" at 70° (0.00348M Rh1) 106 33. Kinetic plot for the loss of [Rh(C0) 2Cl 2] ~ at 70° (0.020M Rh1) 107 34. Arrhenius plot for reaction of [Rh(C0) 2Cl 2]" with 0 2 in 3M HC1 no 35. Kinetic plot for the loss of [Rh^COjClg] 2~ at 393 nm at 70° 112 36. Dependence of [maximum rate] - [ i n i t i a l rate]/2 on [Rh^xfRh 1 1 1] at 70° 114 37. Plot of ln(k^ + k 2 [ R h I H ] ) - i n [Rh1] vs. time at 70° .... 118 38. Kinetic plot for the loss of [Rh1(CO)2C12]" at 326 nm at 55° 120 39. Arrhenius plot for oxidation of [Rh I(C0) 2Cl 2]~ by [R h H I ( C 0 ) C l 5 ] 2 " in the presence of 0 2 in 3M HC1 ......... 122 40. Dependence of the rate of formation of Rh 1 1 1 chloro species from Rh H I(C0) on i n i t i a l [Rh(C0)Cl 5] 2" at 70° ... 125 41. Effect of 0 2 pressure on the rate of formation of Rh 1 1 1 chloro species from [Rh(C0)Cl K] 2" at 70° 126 xvi i Figure Page 42. Effect of [Cl~] on formation of Rh 1 1 1 chloro species from [Rh(C0)C"I5]2~ at 70° ................................ 129 43. Plot of k 3 vs. [HV1 at 70° ............................. 130 44. Arrhenius plot for formation of Rh1** chloro species from Rh***(C0) in 3M HCl 132 45. Plot of log [Rh***(C0)] - log { [Rh*(C0)2Cl2]'"} vs. time at 70° (0.0183M [Rh*] , 0.0217M [Rh***]) 139 Reaction of molecular oxygen with dichlorodicarbonylrhodate(I) in DMA 46. Absorption spectrum.[Rh(C0)2Cl]2 in 0.5M LiCl/DMA and [RhCC0)2CaJ2 i nv.DMA »•. A: A .-. .V. AK 145 47. Reaction of [Rh(C0) 2Cl 2]" with 0 2 in 0.5M LiCl/DMA at 75° followed by i r spectroscopy 147 48. Esr spectra of [Rh(C0) 2Cl 2]" with 0 2 in DMA measured at 77° K 148 49. Rate plots for oxygenation of [Rh(C0) 2Cl 2]" at 75° 150 50. Rate plots for oxygenation of [Rh(C0)2C12]~ in the presence of DPPH at 75° 154 51. Absorption spectrum after oxygenation 157 52. Dependence ofi rate of oxygen uptake on i n i t i a l [Rh] at 75° in DMA 160 53. Dependence of k ] on [Cl~] in DMA 162 54. Arrhenius plot for reaction of [Rh(C0)2C12]" with 0 2 in LiCl/DMA 164 x v i i i Figure Page Preparation and characterization of [Ph^As]n [Rh2Clg(DMA)2] and [Ph 4As] 3[Rh 2Cl g] 55. Absorption spectrum of [Ph 4As] 3[Rh 2Cl g] in DMA 195 56. Absorption spectrum of [Ph4As]2[Rh2Clg(DMA)2] with air in DMA 197 57. Esr spectrum of [Ph 4As] 2[Rh 2Cl g(DMA) 2] in DMA at 77°K .... 203 58. Esr spectrum of [Ph 4As] 2[Rh 2Cl g(DMA) 2] in DMA at R.T 204 59. Esr spectrum of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] with air in DMA at 77°K 205 60. Structure of the [Rh 2Cl g] 3" ion 207 61. Possible structure of the [Rh 2Gl 6(DMA) 2] 2" ion 212 xix ABBREVIATIONS The following l i s t of abbreviations, most of which are commonly adopted in chemical research literature,will be employed in this thesis. A l l temperatures are in °C unless specifically denoted °K. acac Bu tBu or Bu* (Me, Et, and Pr used similarly for simple alkyl groups) Calcd. DBPC DMA DMF DM6 DPPH en esr I i r L log In 0 0. 1 II acetylacetonate, CH3C=CHCCH3 butyl, -C4Hg t-butyl, -C(CH3)3 calculated 2,6-di-tert-butyl-p-cresol, (CH3)3C C(CH 3) 3 N,N -dimethylacetamide, CH3C0N(CH3)2 dimethylformamide, HC0N(CH3)2 monoarii;on of dimethylglyoxime, H0N=C(CH3)C(CH3)=N0 2,2-diphenylpicrylhydrazyl, (C6H5)2NNCgH2(N02)3 ethylenediamine, NH^H^NH,, electron spin resonance ionic strength infrared 1igand common logarithm natural logarithm XX M molar, mole l i t e r " 1 M n + metal atom ^ MNT dianion of malenitriledithiolate, || - S / C \ N MW molecular weight min minutes m.p. melting point nm nanometer OAc acetate, CH3C00C O.D. optical density, absorbance o-phen 1,10-6-phenanthroline, r~\ Ph phenyl, -CgHg PPh3 triphenylphosphine, (CgHg)3P Ph^As+ tetraphenylarsonium cation, (CgHg)4As+ p" partial pressure py pyridine, (f J N R alkyl or aryl group rmsd root mean square deviation R.T. room temperature sec seconds siphos tris(trimethylsilylmethyl)phosphine, {(CH^SiTrt,}^ xx i TPP uv ultraviolet vpc vapor phase chromatography X halogen effective magnetic moment, BM e molar extinction coefficient molar susceptibility, cgsu -1 2 -1 A molar conductance, ohm cm mole m v frequency, cm"* [ ]Total! o r f ^ T denotes total concentration of a species present in solution xx i i ACKNOWLEDGEMENTS I wish to thank Dr. B.R. James for his guidance and encour-agement throughout the course of this work. I am also indebted to Dr. C. Bear for the preliminary data concerning the crystal structure of [Ph^AslgfRh^Clg] . I am grateful to Mr. P.W. liau for his assistance in obtaining and interpreting the single-crystal esr data, and to Dr. E i . Ochiai for many helpful discussions. I also wish to thank the National Research Council of Canada for the award of a Postgraduate Scholarship. 1 Chapter I INTRODUCTION 1.1. Aim of Work Rhodium complexes in solution are known to homogeneously activate small covalent gas molecules such as CO, H2, and 02. The aim of this work was to study, in particular, the interaction of these gas molecules with simple rhodium halide species. An understanding of these processes seemed important in view of their commercial potential as catalysts. For example, since the studies with CO were initiated originally by James and Rempel 1 2 in this laboratory, ' Monsanto have announced their process for carbonyla-29 tion of methanol to give acetic acid. The catalyst used in the process is [ R h ( C O ) 2 ] • Results obtained previously in this laboratory had shown that this dicarbonyl anion, [Rh(C0) 2Cl 2]~, was readily formed through CO reduction of RhCl^'3 H^ O in aqueous acid solution and the reaction proceeded 1 2 via an autocatalytic path. ' In DMA, carbonylation of RhCl3"3 h^ O to give [Rh(C0) 2Cl 2]" involved decomposition by water (or hydroxide) of an intermed-iate rhodium(III) carbonyl species. Recently, some halocarbonylrhodate(III) salts^'^ have been prepared, and these types of compounds are suitable starting materials for a detailed investigation of the reductive step (ie. decomposition by water of Rh***(C0)) in the carbonylation reaction. In this work, the reduction of RhBr3'2 H20 by CO was studied in aqueous HBr. The kinetics and mechanism for the carbon-ylation of [Rh(C0)Brg] in acid solution which gave rhddium(I) carbonyls were also determined. 2 Anionic chlororhodate(III) species have been shown previously to catalyze the reduction of ferric ion in aqueous acid solutions^' 7 under mild conditions. The present studies of the catalytic behaviour of bromo-rhodate(III) complexes in aqueous HBr indicated that these complexes were more effective catalysts than the corresponding chloro species under similarly mild conditions. 8-10 James and coworkers ~ have studied the mechanism of oxidation reactions using molecular catalyzed by [Rh(C. H^^Cl in Li Cl /DMA. g Results indicated the formation of a 1:1 molecular adduct as well as II 9 the presence of some Rh intermediates. Evidence was obtained for involve-ment of rhodium hydroperoxides in such oxidation processes.*^ The catalytic oxidation by [RhtCgH-j^Cl]^ in DMA solution of CO/C^  mixtures to CC^  was also investigated. In the present work, oxidation of [Rh(CO)2C1]2 in aqueous acid and in LiCl/DMA was studied in detail. The kinetics show that quite different mechanisms are operative in each solvent system. In aqueous HC1, oxidation proceeds i n i t i a l l y by an autocatalytic path which is subse-quently followed by water (or hydroxide) decomposition of an intermediate rhodium(III) carbonyl anion. In Li Cl/DMA, the mechanism is free radical with lik e l y formation of hydroperoxides and Rh** species. The chemistry of rhodium(I) and rhodium(III) are often closely inter-related. For example, catalyzed substitution reactions of some inert rhodium(III) complexes have been reported by a number of workers,**"*7 and these were thought to involve traces of rhodium(I) or rhodium(III) hydride species. Such reactions were found to occur for rhodium(III) in the presence 11 12 of hydridic reducing agents (eg. ethanol and hydrazine), ' and intermed-iate Rh*** hydrides were suggested as the active species. 3 Basolo and coworkers f i r s t proposed that these catalyzed sub-18 stitution reactions might involve rhodium(I) complexes, and Rund found that rhodium(I) compounds were more active catalysts than hydride producing reagents, and that some rhodium(III) hydrides were inactive. For the nuc-leophilic replacement of halide by pyridine (eg. (1.1)), the reaction is thought to proceed via RhClg 3 - + 4 py > Rhpy 4Cl 2 + + 4CT (1,1) a chloro-bridged mixed-valence intermediate. py ^py CI /CI • ' 1 I ' 1 i l l f 90 —Rh — C I — Rh — O H A / I / I 2 L py py ci ci J After two electron transfer , the bond breaks at the new labile rhodium(I) center to give the product and regenerated catalyst. Chloride bridged dimers containing Rh1 and Rh 1 1 1 have been obtained from reaction between 19 20 21 azobenzene ' or benzylideneanil ine with ^(CO^Cl^. Non-hydridic reducing agents, CO and ethylene.are also effective 2 22 24 I III catalysts ' ' for substitution, and Rh Rh bridged intermediates have again been postulated. Interestingly, some catalysis even arises from deoxygenation as a result of bubbling gases through solutions, since nitro-23-25 gen, argon, and cyclopropane, as well as evacuation are observed to be effective; trace amounts of rhodium(I) catalysts are thought to be involved. The possibility that these catalyzed substitution reactions involve Rh1* intermediates was not completely ruled out. 4 1.2. Carbonyl Hal ides of Rhodium 26 27 Numerous carbonyl complexes of rhodium have been reported. ' Some of these compounds can function as catalysts, ie. for carbonylation, 28 29 30 of methanol to acetic acid, ' in olefin hydroformylation, and in quadricyclene or cubane valence isomerization. ' The general topic of rhodium carbonyl complexes is too broad in scope to be adequately considered here, hence this discussion w i l l be limited to the carbonyl halides of rhodium, and in particular, to their aqueous chemistry, much of which has been reported since the studies for this thesis were initiated. The dimeric carbonyl hal ides , 3 3 [Rh-fGO) 2_X] 2 (X = C l , Br, I ) , are well known and have been extensively studied. [Rh(C0) 2Cl] 2 can be prepared by passing carbon monoxide over solid hydrated rhodium(III) trichloride at 36 100°. The x-ray crystal structure shows that the solid contains planar o Rh(C0)2Cl groups joined by chloride bridges. The Rh-Rh distance is 3.31 A. In the solid state, the dimers are held together in an in f i n i t e chain by weak metal-metal bonds, which are broken in solution. A stronger metal-metal bond within the dimer arises through bent bonding, ie. the overlap of 2 3 d sp hybridized o metal orbitals which make an angle of ca_ 56° with each other. The chlorocarbonylrhodate dimer, [Rh(C0)2Cl.]2 ,1's a useful starting 37 material for the synthesis of other Rh(I) compounds. It reacts with the aqueous halogen acids (HC1, HBr, and HI) to produce the series of anions 38 [Rh(C0)2X2] (X = Cl, Br, I). These anions were isolated as the tetrabutyl-ammonium and tetraphenylarsonium salts, and their structures were deduced from magnetic, conductivity, and absorption measurements. The x-ray crystal structure of [(n-C^Hg^N] [Rh(C0) 2Cl 2] indicates that the overall co-ordination 5 geometry about the rhodium is square planar with the two carbonyl ligands 39 arranged in a cis configuration. Halide exchange equilibria for [Rh(C0) 2I 2]"7Cl~ show that Cl" is preferred in 1,2-dichloroethane, and CHgCN; but in aqueous media, CH3CN/H20,I~ 40 displays a greater binding preference. CH^I oxidatively adds to [Rh(C0)2X2] (X = Cl, Br, I) with concurrent halide exchange, and the product rapidly isomenzes to acetyliodorhodium(III) species. 4* 2 Rhodium(III) halocarbonyl anions of the type [Rh(C0)Xg] ~ and [Rh(C0)X4]~ have been obtained by the addition of halogens (Cl,,, Br 2) to the dihalocarbonylrhodate ions [Rh(C0) 2X 2]~ (X = Cl, Br) in alcohols. 4 These reactions probably proceed via an oxidative addition of the halogen to [Rh(C0) 2X 2]" to give [Rh(C0) 2X 4]" followed by CO evolution. [Rh(C0)I 4]" and [Rh(C0)I 5] 2" have also been reported. 4' 3 8' 4 2 4 42 ?-Forster ' suggests that the pentahalo [Rh(C0)X5] and tetrahalo [Rh(C0)X4] species exist in equilibrium in solution, since the solid product obtained from solution depends on the cation used. Small cations (CH 3) 4N + or (C 2 H 5) 4N + give salts of the [Rh(C0)X 5] 2 - ions (X = Cl , Br, I ) , whereas (n-C 4H g) 4N + and (Ph 4As) + give rise to salts formulated as [Ph 4As][Rh(C0)I 4] or [Ph4As] [Rh(C0)X4]-xCH30H (X = Cl, B r ) . 4 ' 4 2 Methanol can be removed from these salts yielding compounds that contain the dimeric anions 2- 43 [Rh(C0)X 4] 2 (X = Cl , Br, I). Raman and i r spectra of these complexes have 43 been studied. Complexes such as [Cs]2[Rh(C0)X5] (X = Cl, Br) can be prepared by treatment of the hydrated hexahalorhodate(III) salt with a mixture of formic 5 acid and the halogen acid. More prolonged reaction in the case of the chloro-and bromo-species gives cis-[Rh(C0) 2X 2]" (but the iodo reaction stops at the rhodium(III) stage to give [ R h ( C 0 ) I 5 ] 2 - ) . 4 3 ' 4 4 ' 4 5 The anionic rhodium(III) 6 monocarbonyl species, [Rh(C0)X4]~-xCH30H, can also be reductively carbony-lated in CHC1^  to give the cis-rhodium(I) dicarbonyl species. 4 Qualitative equilibrium data in CH3N02 and CH3CN for the systems [Rh(CO)I 5] 2"/Cl" and [Rh(C0)I 5] 2"/Br~ indicate a preference for the heavier halide. 4 0 1.3. Activation of Molecular Hydrogen 46 A recent text on homogeneous hydrogenation reviews the l i t e r a -ture to 1972 and provides a detailed examination of the activity of metal ions for the reduction of both organic and inorganic substrates. The mech-anism of hydrogenation involves at some stage activation of the hydrogen 47 molecule through the formation of a reactive metal hydride species. 47-50 Hydrogen activation has been recognized to occur in three ways. 51 Heterolytic s p l i t t i n g , demonstrated by reaction (1.2), Ru H IClg" + H2 . R u I H C l 5 H 3 " + H+ + CI" (1.2) involves substitution of a hydride for another ligand without change in the oxidation state of the metal. Reactivity toward H2 in such cases is governed by the stab i l i t y of the hydrido complex, the case of displacement of the original ligand, and the presence of a suitable base to stabilize the released 47 52 proton. ' , 0 d Hydrogen activation by homolytic s p l i t t i n g 4 7 ' 5 3 (equation (1.3)) 2[Co(CN) 5] 3" + H2 v 2[HCo(CN)5]3" (1.3) 54 and by oxidative addition (dihydride formation) (reaction 1.4)) 7 IrCl(CO)(PPh 3) 2 + H2 v * H 2IrC1(C0)(PPh30 2 (1.4) is dependent on the susceptibility of the metal to oxidation and the a b i l i t y of the metal ion to expand i t s co-ordination shell. 1.4. Reduction of Inorganic Substrates The reduction of inorganic substrates commonly proceeds by homolytic or heterolytic splitting of molecular hydrogen by the catalytic species. Examples of each w i l l be given. I 55 Cu acetate in quinoline solution catalyzes the homogeneous reduction of Cu**. The rate-determining step (1.5) is the 2Cu* + H9 v k* * 2[CuI*H]+ (1.5) k - l [Cu**H]+ + Cu** f a s t > 2Cu* + H+ (1.6) formation of a metal-hydride complex in a termolecular reaction via homolytic splitting of H2; this is followed by a fast step (1.6) involving reduction of the substrate. 6 III Harrod and Hal pern studied the H2 reduction of Fe catalyzed by RhCl3'2H20 in aqueous 3M HC1. The R h I H + H2 ^ k* » Rh***H- + H+ (1.7) k - l Rh***H" + 2Fe*** - ^ ^ M Rh*** + 2Fe** + H+ (1.8) 8 mechanism involved heterolytic splitting of H2 followed by rapid reduction of the intermediate hydride with f e r r i c . Direct evidence for the equilibrium in the i n i t i a l step has been obtained in the corresponding Ru*** system by 56 57 isotopic exchange studies with deuterium. ' Complexes of copper(II), s i l v e r ( I ) , mercury(II), palladium(II), ruthenium(II), and cobalt(II) have also been found to be active catalyts for the reduction of inorganic sub-48 strates through such homolytic or heterolytic mechanisms. CO Less wel1-documented observations suggest that inorganic sub-strates may be catalytically reduced by complexes that activate H2 through dihydride formation, but the mechanisms of these reactions have not been elucidated. 1.5. Bonding in Molecular Oxygen Complexes Complexes of the second and third row group VIII metals generally involving triphenylphosphine as ligands have been found to add molecular 59 60 oxygen. ' Attempts have been made to distinguish between oxygen bound as 2-co-ordinated oxygen, the related peroxo (0 2 ), and superoxo (0 2 ) ligands. The addition of oxygen to square planar d systems was discussed by Ibers and is based on the -rr-bonding scheme used to explain the co-ordination of the 62 63 ethylene molecule to platinum in Zeisse's salt. ' 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 occurs to offset the resultant dipole from a f i l l e d metal orbital to a IT* antibonding orbital on the ligand mole-cule. If electron transfer occurs from the metalj the oxygen molecule as-2-sumes the character of the superoxide ion, Qn , or the peroxide ion, 0 2 , depending on whether one or two electrons are transferred. 9 1.6. Metal Catalyzed Oxidation Reactions Involving Molecular Oxygen The oxygen adducts are reactive and various oxidations have been achieved under exceptionally mild conditions. Examples of ligand oxidations 64-66 67 include phosphine to phosphine oxide, ~ isocyanide to isocyanate, carbon monoxide to chelated carbonate ' or carbon dioxide, n i t r i c oxide to n i t r i t e 7 * or nitrate,^ 8 and sulphur dioxide 72-74 t Q c n e i a t e < j sulphate. Some of these reactions are stoichiometric with the oxidized ligand being retained in the co-ordination sphere of the metal, but the process may become catalytic i f a further ligand can replace the oxidized product. Several homogeneous autoxidation reactions of hydrocarbons catalyzed by transition metal complexes have been reported. " The catalytic role of the metal 78-80 when direct oxygen activation is involved is uncertain. . Mechanisms involving dissociative "oxygen insertion" or free radical intermediates have been proposed for the various catalytic oxygenation reactions. 75-78 81 82 Hal pern ' has studied the oxidation of triphenylphosphine using Pt(PPh 3) 20 2 as catalyst. The results are consistent with the following mech-anism. Pt(PPh 3) 3 + 0 2 slow -> Pt(PPh 3) 20 2 + PPh3 Pt(Ph 3P) 20 2 + Ph3P s 1 o w > 2Ph3P Pt(Ph 3P) 3 + 2Ph3P0 * 1 - r ,PPh.3 / ;>o PfuP - Pt*:V • \ > o Wh., Ph3P - Pt ^ 0 P P h 3 OPPh, (1.9) (1.10) (1.11) 10 This type of reaction may be considered an example of a metal.ion promoted 71 83 atom transfer redox reaction, ' in which both the reductant (triphenyl-phosphine) and oxygen molecule co-ordinate to the metal. A rhodium(I) catalyzed oxidation of terminal olefins to methyl ketones is believed to on occur via a similar atom transfer mechanism. The rhodium(I) catalyzed autoxidations of cyclohexene, 7 6 ' 7 8 ethyl-benzene,76 and diphenylmethane77 have been shown to proceed by a free radical chain mechanism. The probable role of the metal complex is to cause the breakdown of hydroperoxide intermediates. The chain nature of the reaction was demonstrated by the complete inhibition of the oxidation by radical scavengers. Rhodium, by analogy with cobalt, may catalyze chain i n i t i a t i o n 85 by a Haber-Weiss type mechanism. R00H + Rh 1 1 1 > R00- + H+ + Rh 1 1 (1.12) R00H + Rh 1 1 >R0' + OH" + Rh 1 1 1 (1.13) An alternative mechanism involving a two electron transfer was also 76 suggested. This mechanism requires termolecular 2R00H + R h I H > 2R00- + 2H+ + Rh1 (1.14) 2R00H + Rh1 > 2R0' + 20H" + Rh 1 1 1 (1.15) reaction and seems less li k e l y to occur. For the case of c o b a l t 8 6 ' 8 7 and other metal 7 5 catalyzed oxidations, complex formation between the hydroperoxide and catalyst prior to electron 11 transfer has been postulated. In the case of low valent phosphine complexes, the hydroperoxide-catalyst complex could correspond to a two electron oxi-dative addition of the hydroperoxide by the following two possible routes: 78 Mn + + ROOH (a) (b) ROO H RO HO (n + 2) + (n + 2) + -> radicals (1.16) radicals 1.7. Rhodium(II) Complexes Most of the reported compounds of divalent rhodium are diamagnetic QQ involving some degree of metal-metal bonding. Rhodium(II) acetate mono-91 hydrate is isostructural with binuclear copper(II) acetate monohydrate and 89 90 ° the Rh-Rh bond length ' is * 2.39 A. Many adducts of the type R h ^ C H g ^ ^ - l ^ 2 - ^ have been prepared in which the rhodium atoms are bridged by an acetate cage, and the addend L is terminally co-ordinated to the metal 92-96 atoms. This type of structure is consistent with the analytical data, 97 98 diamagnetism, and electronic spectra for these compounds. Similar 99-101 96 rhodium(II) formates and propionates have been obtained. Diamagnetic compounds such as [Rh(CH3C00) (6-diketonato)L] 2 (where g-diketonato is the anion of the enol form of acetyl acetone or fluorine sub-12 stituted acetylacetone) have been made. Two dinuclear Rh** complexes Rh 2(DMG) 4(PPh 3) 2 1 0 3 and Rh 2(DMG) 2(CH 3C0 2) 2(PPh 3) 2 1 0 4 (DMG = monanion of dimethylglyoxime) have been prepared; the rhodium-rhodium distances* 0 4~*^ were measured in order to relate bond length to bond multiplicity. The diamagnetic Rh** complexes of empirical formulae [Rh(CNBu t) 3(PPh 3)I] n[PFg] n and [Rh{CN(p-ClC 6H 4)} 2(PPh 3) 2I] n[PF 6] n (where n is presumably 2) have been obtained. 1 0 7 Binuclear rhodium(II) complexes [Rh(CNPh) 4I] 2X 2 (X = I, ClO^) have also been reported.* 0 8 1 Q Q Wilkinson and coworkers have protonated rhodium(II) acetate in methanol-fluoroboric acid media to displace the acetate yielding a green air-stable diamagnetic ion Rh 2 +, and the binuclear ion [Rh2(H20)-|Q]4+ has been characterized by Taube.**0 Preliminary magnetic measurements on the aquo ion suggest that i t is 6-10% dissociated into the paramagnetic monomeric form. Paramagnetic rhodium(II) complexes of the type trans-[RhCl 2(PBu 2R) 2] 111,112 ^ R _ ^ ^ or nPr) are formed by treating ethanolic rhodium(III) chloride trihydrate with bulky tertiary phosphines at 25°C. The complexes have low magnetic moments in the solid state (< 1.4 B.M.) but in solution trans-[RhCl 2(PBu 2Me) 2] has the expected moment of ca. 2.12 B.M. Similar reactions using tri-o-tolylphosphine, (o-CH3CgH4)3P> (at 25° or below 0°) yield the blue-green trans complex, [RhCl 2((o-tol) 3P) 2] ( v e f f = 2.27 BM), or 113 its purple modification ( v e f f = 2.0 BM). Phenyldi-o-tolylphosphine, (CgH5) (o-CH3CgH4)2P; gives a red divalent rhodium complex ( y e f f ^ 1.0 BM) which may contain Rh-Rh bonds.**3 Wilkinson** 4 has recently prepared tris(trimethyl-silylmethyl)phosphine (siphos) complexes of rhodium(II). [RhCl 2(siphos) 2] 2 is dimeric in benzene while trans-[RhBr2(siphos)2] is a monomeric species. 13 Both complexes have a magnetic moment of ca_. 1.5 BM. For the monomeric rhodium(II) compounds, "pseudooctrahedral" structures containing trans 112-114 phosphine ligands have been proposed. The st a b i l i t y of such complexes has been related to steric effects. 115 Gray and coworkers prepared the paramagnetic square planar Rh 1 1 complex [(n-C4Hg)4N][Rh(MNT)2] ( y f f = 1.91 BM) (MNT = maleonitriledi-thiolate) in which each Rh ion is co-ordinated to four sulfur atoms. Single crystal esr spectra have been obtained.** 6 Other genuine paramagnetic Rh** species are the hexamethylbenzene complex,**7 [RFHCgtCH^g}]2* (u eff = 1-32 118 BM), the bis-iT-cyclopentadienyl complex, Rh(CgHg)2 and the tetraphenyl-porphyrin complex, 1 1 9 Rh(TPP) ( y f f <v 1.2 BM). G i l l a r d 1 2 0 suggested that a dimeric Rh** complex [Rh(en) 2(H 20)] 2 + (en = ethylenediamine) was formed during the controlled potential reduction + 121 of trans-[Rh(en) 2Cl 2] at a mercury cathode. Later workers, however, identified the rhodium(II) dimer as a mercury (Il)-bridged complex of rhodium .4+ (I), [(en)2Rh]2Hg 14 Chapter II APPARATUS AND EXPERIMENTAL PROCEDURE 2.1. Materials 2.1.1. Rhodium Compounds Rhodium(III) tribromide was obtained as the dihydrate, RhBr^* 2 h^ O, from Platinum Chemicals; rhodium(III) trichloride was obtained as the trihydrate, RhCl3*3 H20, from Johnson Matthey Co. [Cs] 2[Rh(C0)Br 5] and [Cs]2[Rh(C0)Clg] were prepared according to the method of Cleare and Grif f i t h from the corresponding Rh(III) trihalides. The precipitated salts decomposed readily when washed with d i s t i l l e d H20, being reduced to 122 metal in the absence of excess halide; when treated with the hydrohalic acid, orange powders were obtained and no decomposition occurred. [Rh(C0) 2Cl] 2 was synthesized according to the method of McCleverty 35 and Wilkinson by passing a CO stream over solid RhCl^'3 H20 at 100°. The product was purified by sublimation at ca_. 80° and at a pressure of ca_. 0.1 mm. yielding dark red crystals. Stock solutions of [Rh(C0) 2Cl 2]~ were obtained by dissolving the solid in 3M HC1 under N2- These solutions were then stored under CO in flasks fitted with a gas bubbling device. Samples of stock solution were always removed under CO and degassed prior to use. 2.1.2. Gases Prepurified grade hydrogen obtained from Matheson Co. was passed through a Deoxo catalytic purifier before use to remove traces of oxygen. Carbon monoxide was obtained as CP. grade from Matheson Co. Nitrogen (L. 15 grade) and oxygen were from Canadian Liquid Air Co. 2.1.3. Sol vents Concentrated HBr, HCl, and HCIO^ were diluted to give stock acid solutions^which were standardized against alkali solutions prepared from B.D.H. standard NaOH ampuls. Certified N,N-dimethylacetamide, obtained from Fisher Scientific Co., was purified by d i s t i l l a t i o n from calcium hydride under nitrogen and stored over Linde 4A molecular sieves under a nitrogen atmosphere. 2.1.4. Other Materials Sodium bromide was CP. grade obtained from B.D.H. Anhydrous l i t h -ium chloride was certified grade from Fisher Scientific Co. Fe(C104)3*6 H20 was reagent grade product obtained from Alpha Inorganics Inc. Anhydrous Li Cl0^ was reagent grade from G. Frederick Smith Chemical Co. Ph^AsCl was obtained as Baker grade from Baker Chemical Co. 2,2-Diphenylpicrylhydrazyl (reagent grade) and 2,6-di-tert-butyl-p-cresol (practical grade) were obtained from Eastman Organic Chemicals. All other chemicals were reagent grade and d i s t i l l e d water was used in a l l experiments. 2.2. Apparatus for Constant Pressure Gas-Uptake and Evolution Measurements Gas-uptake measurements were performed using the constant pressure apparatus shown in Figure 1. A flexible glass spiral tube connected a capil-lary manometer D at tap C to the pyrex reaction flask A; two-necked reaction Apparatus for constant pressure gas-uptake or evolution measurements. 17 flasks could also be used, allowing attachment of a sampling tube for col-lection of gaseous reaction products. The flask was immersed in a thermo-stated bath B containing silicone o i l (Dow Corning 550), and clipped to a piston-rod and wheel driven by a Welch variable speed electric motor for shaking purposes. The capillary manometer D containing n-butyl phthalate (a liquid of negligible vapor pressure) was connected to the gas-measuring burette consisting of a mercury reservoir E and a precision bored tube N of known diameter. The gas burette was in turn connected, via an Edwards high vacuum needle valve M, to the gas-handling part of the apparatus; this consisted of a mercury manometer F, gas inlet Y, and connections to the Welch Duo-Seal rotary vacuum pump G. The capillary manometer and gas burette were contained in a transparent perspex water tank thermostated at 25°. The silicone o i l bath consisted of a four-litre glass beaker insulated by polystyrene foam on a l l sides and enclosed by a wooden box, with the top also covered by polystyrene. Both thermostat units were controlled by Jumo thermo-regulators and Merc-to-Merc relay control c i r c u i t s , with heating provided by a 40 watt elongated light bulb. With mechanical stirring and good insu-lation, the temperature could be maintained to within ±0.05°C. A vertically mounted cathetometer was used to follow the gas uptake, and time was recorded during kinetic experiments using a Lab-Chron 1400 timer. 2.3. Procedure for a Typical Gas-Uptake or Evolution Experiment For each experiment, known amounts of the reactants (eg. rhodium complex) were added to the solvent in the reaction flask A. Solid materials were added in glass buckets, and predissolved solid was transferred via pipette. 18 The solvent was always degassed before the addition of catalyst, either by repeated freezing followed by warming under vacuum, or simply by shaking under vacuum at room temperature and pumping off the evolved gas (the la t -ter method being used for a solvent of very low vapor pressure, such as DMA). The reaction flask and spiral were f i l l e d with the reactant gas, at a pressure somewhat less than that required for the experiment, by connec-tion to the gas-handling part of the apparatus at 0. The taps C and P were then closed and the reaction flask complete with spiral was disconnected from 0 and attached to the motor driven shaker I. The whole system up to tap C was then pumped down with taps H, K, L, J, and M open. Reactant gas was admitted to the rest of the apparatus beyond C to a pressure slightly less than that desired for reaction. Tap C was then opened and pressure adjusted to the desired reaction pressure by introduction of gas through Y. A run was started by closing taps K and L. and simultaneously starting the shaker and timer. Gas uptake was indicated by the difference in the o i l levels of the manometer D. The manometer was balanced by allowing gas to enter the burette through the needle valve, and a constant pressure was thus maintained in the reaction flask. The resulting rise in the mercury level in N was measured at appropriate intervals of time. Since the diameter of the tube N was known, the volume of gas consumed could be calculated and expressed as moles of uptake per l i t r e of solution. For gas evolution experiments tap L was closed, prior to the final pressure adjustment, which resulted in a rise in the mercury level in N. Gas evolution was followed by withdrawing gas across the needle valve M and observing the f a l l in the mercury level. A rapid shaking rate together with the use of small volumes of solution (3-5 ml) in a relatively large indented 19 reaction flask (25 ml) ensured absence of diffusion control due to slow dissolution (or evolution) of the reactant (or product) gas. 2.4. Gas Solubility Measurements The solubility of a gas in a solvent under specific temperature and pressure conditions could be determined using the gas-uptake apparatus and a reaction flask containing a stopcock in i t s neck. The entire system in-cluding the reaction flask containing a measured amount of solvent was evacuated at room temperature. The tap on the flask was then closed, and the flask was placed in the o i l bath at the desired temperature. The system was then evacuated to the flask tap and f i l l e d with gas to the approx-imate pressure desired. The flask tap was then opened and the pressure ad-justed immediately to that required. Taps K and L were closed, the shaker started and the immediate uptake measured as described in the previous sec-tion, allowing calculation of the gas solubility. 2.5. Spectrophotometric Kinetic Measurements The kinetics of some reactions were studied using spectrophotometric techniques in the visible range. A kinetic experiment, involving gas uptake, could be carried out directly in an optical s i l i c a c e l l . The cell was fitted with a serum cap containing teflon capillary tubes for the introduction and exhaust of a gas-eous stream through the solution. The rapid gas bubbling rate through the solution ensured the absence of physical control. The reaction was followed by the decrease or increase in optical density of the absorption maxima of 20 the reacting species as a function of time, in a spectrophotometer fitted with a thermostated cell compartment. The gaseous stream was briefly stopped during optical density measurements. 2.6. Reaction Product Analysis 2.6.1. Gaseous Products For collection of gaseous reaction products a double-necked reac-tion flask was employed, one neck being connected to an evacuated sample bulb having a stopcock. At the completion of the reaction, the stopcock on the bulb was opened momentarily and a gas sample collected. The sample was then subjected to either gas chromatographic or mass spectrograph^ analysis. Carbon dioxide was also identified by absorption on sodalime which would be loosely packed between glass-wool in one neck of a double-necked reaction flask. 2.6.2. Liquid Products Di s t i l l a t i o n under vacuum (ie. pumping through a liquid nitrogen cold trap) was used to separate the solvent plus liquid organic products from the inorganic complex and non-volatile products. The d i s t i l l a t e could then be investigated by gas chromatography. As an alternative method, aliquots of the reaction mixture were directly applied to the chromatographic column without prior separation of the inorganic product. 21 2.6.3. Solid Inorganic Products Inorganic products were isolated by reducing the final reaction solutions to a small volume, followed by addition of a complexing ligand or counter ion (eg. particularly Ph^As*). Microanalyses were performed by P. Borda of this department and Galbraith Laboratories of Tennessee. Molecular-weight determinations were also done by Galbraith. 2.7. Instrumentation Visible and ultraviolet absorption spectra were recorded using a Perkin-Elmer 202 spectrophotometer, fi t t e d when necessary with a thermostated cell compartment. Matched s i l i c a cells of 1 mm or 10 mm path length were used. Infrared spectra were recorded on a Perkin-Elmer 700 or a Perkin-Elmer 457 spectrophotometer. Solution spectra were obtained using 0.1 mm NaCl cells and solid spectra as nujol mulls on Csl or KBr plates or as KBr discs. Esr spectra were recorded using a Varian E3 instrument, and mass spectra on an Associated Electrical Industries MS9 mass spectrometer. A Beckman GC-2A unit with thermal conductivity type detector was used for gas chromatographic separations with dinonylphthalate or s i l i c a gel columns. Conductivity measurements were carried out using a Thomas Serfass conductivity bridge and dip-type conductivity c e l l . Melting points were recorded using a Superior Electric apparatus. Magnetic susceptibility measurements were made on a Gouy balance by Dr. CT. Wong. 22 Chapter III CARBONYLATION OF RHODIUM(III) BROMIDE COMPLEXES IN AQUEOUS HYDROBROMIC ACID SOLUTION 3.1. Introduction 1 2 James and Rempel ' have reported on the kinetics and mechanism of the reduction by carbon monoxide of RhCl^*3 H20 in aqueous hydrochloric acid solutions; the reaction produces the dichlorodicarbonylrhodate(I) species, [Rh(C0) 2Cl 2]~ in an autocatalytic process. An attempt was made in the present work to extend this study to other Rh-ligand species, particular-ly the other halides (Br, I) in order to systematically investigate the redox properties of such systems. Useful information can then be obtained concerning the role of the ligand (electronic and steric effects) and the metal centre (susceptibility to oxidation and reduction) in the carbonyla-tion reactions. This chapter discusses the beginning of such efforts. Since only a bromorhodate(III) complex could be prepared and characterized, the study was limited to an investigation with this system. 3.2. The Reaction of RhBr3*2 H20 in HBr with CO Solutions of ca_. 0.02M RhBr3"2 H20 in 0.5M HBr, thermally equi l i -brated for a minimum of 20 hours at 80°, exist essentially as the 4:1 bromo complex, [RhBr 4(H 20) 2]". The red solutions have an absorption maximum at 538 mm, e = 173, Figure 2, which is very similar to that reported by Cozzi 123 and Pantani for the tetrabromodiaquo species (540 nm; e ^  180). The solu-tions absorbed CO at measurable rates between 40-55° and gave autocatalytic 23 450 500 550 600 650 700 Wavelength, nm Figure 2. Absorption spectrum. RhBr3*2 h^ Q in 0.5M HBr (equilibrated at 80° for 20 hours). 24 280 300 320 340 360 380 Wavelength, nm Figure 3. Absorption spectrum. [Rh(C0) ?Br ?]" in 0.5M HBr. 25 uptake plots of the type shown in Figure 4. The net measurable gas uptake corresponded to a 2:1 mole ratio of gas:Rh, and experiments in the presence of a sodalime tube resulted in a total measurable gas uptake corresponding to a 3:1 mole ratio of C0:Rh. These observations indicate an overall stoi-chiometry given by equation (3.1). H 0 Rh 1 1 1 + 3 CO > Rh I(C0) 2 + C02 + 2H+ (3.1) (bromide and water ligands are omitted) A mass spectral analysis of the gas mixture at the end of the reaction confirmed the C02 production. The original red solution becomes yellow indicating l i k e l y reduction to a Rh1 s t a t e 4 0 ( A „ = 326 nm, e = 3700) (Figure 3). A rhodium carbonyl reaction product could be precipitated by addition of an aqueous solution of tetrabutylammonium chloride to a concen-trated volume of the yellow solution. The properties of the isolated solid V C Q = 2058 cm-1, 1983 cm-1; mp = 83°, corresponded to those reported by 3 R Vallarino for [(n-C 4H g) 4N][Rh(C0) 2Br 2]. Thus, i t is probable that the carbonylation product exists in solution as [Rh(C0) 2Br 2]". 3.3. Kinetics of the Reaction in HBr Media The autocatalytic plots exemplified in Figure 4 yield a maximum slope at a point representing ca. 50% of the total gas uptake regardless of the i n i t i a l [Rh 1 1 1] used. Differentiating the uptake plots graphically pro-duces curves of the type shown in Figure 5 which are nearly symmetrical about the 50% reaction point. It is readily shown that a rate proportional to the 0 1.6 3,2 4.8 6,4 8.0 9,6 11.2 12,8 Time x 10~3, sec Figure 4. Rate plots for the autocatalytic reduction of Rh 1 1 1 at 40°, 759 mm. CO, [Rh 1 1 1]: (0) 0.020M; (A) 0.015M, 0.5M HBr. 0 0.4 0.8 1.2 1.6 2.0 0 0.4 0.8 1.2 1.6 [Rh 1 1 1] x 102,M [RhlH] x 102,M Figure 5. Typical plots of -d[C0]/dt vs. [Rh 1 1 1], 40°, 759 mm CO, [Rh 1 1!]. ... ,: (0) 0.020M; ( A ) 0.015M, 0.5M HBr. 28 product [Rh I I I][Rh I] becomes a maximum when [Rh 1 1 1] = [Rh1] = l/2[Rh] T t and thus the dominant term in the rate expression has a rhodium dependence expressed by equation (3.2). -d[C0]/dt = k^[Rh I I I][Rh I] (3.2) where is a pseudo-second order rate constant containing a l l other chemi-cal and physical variables except the dependence on rhodium. This type of behaviour has been found earlier in the rhodium chloro system, as well as 124 12 for the reductive carbonylation of cupric salts in aqueous solutions. ' It is important to note that the curves in Figure 5 show a non-zero rate at the highest (ie. i n i t i a l ) [Rh 1 1 1] or, in other words, a s i g n i f i -cant uptake rate at the start of the reaction at zero[Rh*]. This contribu-tion to the rate may be taken into account by adding an additional term to the rate law in equation (3.3). -d[C0]/dt = k ^ [ R h n i ] n + k 2[Rh I I I][Rh I] (3.3) where k-J is a pseudo-first order constant. 3.3.1. The I n i t i a l Reaction Since no Rh* is present at the start of the reaction,and the amount of Rh* produced in the early stage of reaction is very small relative to the Rh*** present, the i n i t i a l slow uptake of CO proceeds by a path which contributes to the f i r s t term of the rate equation (3.3). The kinetics of the i n i t i a l reaction were determined from i n i t i a l slopes which could be 29 measured with reasonable accuracy. Table I summarizes these data at 40° which show a good f i r s t order dependence on Rh*** up to 0.02M (Figure 6) -4 / x and on CO up to 7.91 x 10 M (Figure 7). The rate law becomes -d[C0]/dt = k 1[Rh I I I][C0] + k2[Rh***][RhI] (3.4) and calculated values of k-j are given in Table I. The values of k-j at 40° calculated from the slopes of Figure 6 and Figure 7 are 0.14M-* sec * and 0.16M * sec * respectively. Measurements over the temperature range 40-55° (Table II) yielded a good Arrhenius plot (Figure 8), and the activa-tion parameters AH-^ = 26.9 ± 2.0 kcal/mole and A S ^ = 23.5 ± 5.9 eu. 3.3.2. The Reaction Involving the Autocatalytic Species The analysis of the k^  term has been carried out in a number of ways. (a) Maximum rate measurements were made in a series of experiments at 40° using different rhodium concentrations at a constant CO pressure (Table III). A plot of these rates against [Rh***] x [Rh*] at the maximum rate showed a linear dependence (Figure 9), which supports the dominance of the second term in the rate law (3.4) for this region of the reaction. The possible non-zero intercept of the plot indicates that the f i r s t term of the rate equation (3.4) may not be completely insignificant at this stage. A value of k^  = 0.058M-* sec -* was calculated from the slope of Figure 9. (b) A series of maximum rate measurements were made for varying CO 30 Table I Carbonylation of [RhBr 4(H 20) 2] Summary of kinetic data in 0.5M HBr at 40° [Rh I n]X CO [CO] In i t i a l rate 102,M mm 104,M* of CO uptake M~lsec~l xl06,Msec-l 2.03 759 7.91 2.26 0.14 2.00 759 7.91 2.48 0.16 2.00 570 5.94 1.87 0.16 2.00 380 3.96 1.31 0.16 2.00 190 1 .98 0.67 0.17 1.52 759 7.91 1.65 0.14 1 .02 759 7.91 1.10 0.14 *[C0] computed from solubility data of Seidell assuming the solubility to be the same as in pure water. 31 32 Figure 7. Effect of CO pressure on the i n i t i a l rate, 40°, 0.020 M Rh 1 1 1, 0.5M HBr. 33 Table II Carbonylation of [RhBr^h^O^ -Temperature dependence of k-| ( [ R h H I ] = 0.020M; 0.5M HBr) Temp. CO [C0]X °C mm 104,M* M"lsec~l 40 759 7.91 0.14 45 743 7.35 0.39 50 723 6.85 0.57 55 698 6.34 1.23 *[C0] computed from solubility data of 1 2g Seidell assuming the solubility to be the same as in pure water. Figure 8. Arrhenius plot for reaction of Rh with CO, 0.020M, Rh 0.5M HBr. 35 Table Carbonylation of Maximum rates in III [RhBr 4(H 20) 2]" 0.5M HBr at 40° [ R h m ] X CO [C0]X Maximum rate X 1 02,M mm 104,M* lOS.Msec1 2.03 759 7.91 6.25 2.00 759 7.91 6.04 2.00 570 5.94 5.55 2.00 380 3.96 5.32 2.00 190 1 .98 3.92 1 .52 759 7.91 3.65 1 .02 759 7.91 1.72 *[C0] computed from solubility data of Seidell assuming the solubility to be the same as in pure water. Figure 9. Dependence of maximum rate on the product [Rh ] x [Rh ], 40°, 759 mm CO, 0.5M HBr. 37 pressures at 40° and a constant [Rh 1 1 1]. A plot (Figure 10) of these rates versus the CO pressure (p'CO) showed no simple order dependence in CO. At higher pressure, the rate only increases slightly with pressure and an examination of the rate law (3.4) suggests that this increase arises from the CO pressure dependence of the f i r s t term. The slope of maximum rate versus p'CO in Figure 10 should therefore provide a value for k-|. The value obtained for the line drawn is 0.16M"* sec - 1 which is in good agreement with values obtained from i n i t i a l slopes (Table I). Equation (3.4) suggests that the second term should-be identified with the intercept of Figure 10, the value of kl, thus obtained is 0.047M-* sec"* which agrees reasonably well with the value obtained in a). Hence, i t would appear that the reaction involving the autocatalytic species is virtually independent of CO at higher pressure. Thus k'2 becomes a true second order rate constant, k9, and the rate law fi n a l l y becomes: Figure 10 indicates that a CO dependence effect would be f e l t on the second term of the rate law at low CO pressures; this is attributed to a different step becoming rate determining namely the formation of [Rh(C0) 2Br 2]" from Rh1 and CO (see discussion sec. 3.5.2.). (c) It may be readily shown that the ratio (R) of maximum to i n i t i a l reaction rate is given by: -d[C0]/dt = k 1[Rh I I I][C0] + k 2[Rh I I I][Rh I] (3.5) R = (-d[C0]/dt) max = (k 1[RhI I I][C0]+k 2[Rh I I IjRh 1]) max 38 0 100 200 300 400 500 600 700 800 p'CO, mm .' • :.... Figure 10. Effect of CO pressure on the maximum rate, 40°, 0.020.M R h I H , 0.5M HBr. 39 Since the [Rh***] at the maximum reaction rate wi l l be 1/2 the [Rh***] i n i t i a l l y present, equation (3.6) can be formulated as: R = k 2 f R h \ a x + 1 / 2 (3.7) 2k][C0] where [Rh*] m a x is the concentration of the autocatalytic species at the point of the maximum rate. Values of k 2 may be obtained from equation (3.7) using the data of Tables I and III and are included in Table IV; they agree well with values from methods a) and b). (d) Since -d[C0]/dt = -2d[Rh***]/dt = 2d[Rh*]/dt (3.8) rate-law (3.5) can be integrated directly to give the expression log(k 1 l + k2[Rh*])-log[Rh***] = (k^ + k 2 [ R h ] T o t a l ) t /4.6 + log Y - l o g L R h ] ^ ( 3" 9 ) where k-j1 = ^  [CO] and t is the reaction time. Once [Rh*] has built up to an appreciable concentration and [Rh***] has correspondingly decreased, the second term of the rate equation w i l l dominate the consumption of CO. Thus, in equation (3.9), the log term containing k-j 1 can be simplified, since k^  1 w i l l be considerably less than both k2[Rh*] and ^ [^n^Tota!' Equation (3.9) reduces to log[Rh*] - log[Rh***] = k 2 L R h ] T o t a 1 t /4.6 + constant (3.10) which neglects the f i r s t part of the reaction. Expression (3.10) suggests that a plot of log[Rh*] - log[Rh***] against time should give a straight line of slope k2^ R n-'Total 7 4 , 6 (Figure 11). k 2 values calculated by this method are summarized in Table V. As expected, the plots deviate considerably from linearity during the f i r s t portion of the reaction. An estimate of the im-40 Table IV Carbonylation of [RhBr 4(H 20) 2]~ Calculation of k 2 by use of equation (3.7) (40°, 0.5M HBr) [Rh I n]X 102,M [C0]X 104,M* ["Max. rate ] LInit. rateJ kl M -lsec -! M" k 2 -lsec-1 2.03 7.91 2.77 0.14 0, .050 2.00 7.91 2.44 0.16 0. .048 2.00 5.94 2.97 0.16 0. .046 2.00 3.96 4.07 0.16 0. .045 2.00 1.98 5.90 0.17 0. .036 1.52 7.91 2.21 0.14 0. ,049 1.01 7.91 1.57 0.14 0. 047 *[C0] computed from solubility data of Seidell assuming the solubility to be the same as in pure water. 41 Figure 11. Plots of log [[RhkcO^B^]'] - log [Rh 1 1 1] vs, time for various CO pressures (40°, 0.020M Rh, p'CO: (0) 759 mm; (A) 380 mm; (•) 190 mm, 0.5M HBr). 42 Table V Carbonylation of [RhBr 4(H 20) 2]~ Calculation of k 2 by use of equation (3.10) (40°, 0.5M HBr) [Rh I H]X 102,M [C0]X 104,M* k 2 M"lsec _l 2.03 7.91 0.060 2.00 7.91 0.060 2.00 5.94 0.055 2.00 3.96 0.053 2.00 1.98 0.048 1.52 7.91 0.067 1 .01 7.91 0.074 *[C0] computed from solubility of S e i d e l l 1 2 6 assuming the solubility to be the same as in pure water. 43 portance of the f i r s t term of the rate law can be made at the maximum rate for a typical reaction. At 40° when [ R h ] T o t a 1 = 0.020M and [CO] = 7.91 x -4 10 M, the contribution of this term to the overall rate is approximately 20%, therefore, the value of kg calculated by this method would be larger than expected. As the [R h ] j 0 ^ a i 1 S decreased or [CO] is increased, the effect of the f i r s t term becomes more significant. Consequently at higher [Rh], the second term becomes dominant, and a more accurate estimate of kg can be made. Kinetic measurements over the temperature range 40-55° with analy-sis by method (c) yielded a good Arrhenius plot (Table VI, Figure 12),and the activation parameters hW^ = 15.1 ± 1.0 kcal/mole and ASg^ = -16.5 ± 3.1 eu. 3.4. Effect of Bromide and Acid By decreasing the [HBr] and adding sufficient HCIO^ to provide a 0.5M acid media, experiments were carried out in which the [Br~] was varied while maintaining constant acidity. The kinetic data obtained at 40° are summarized in Table VII. Over the [Br -] range 0.20M - 0.50M the major Rh 1 1 1 species i s the 4:1 bromo complex U M A X = 540 nm, e ^ '200). However, at 0.10M [Br~] a somewhat different spectrum is observed (A = 514 nm, e = 193) and L max is most li k e l y due to the presence of an increased amount of the 3:1 complex [RhBro(H 00),] for which the absorption spectrum has been measured ( A M 3 V = o c o max 123 510, e % 230). The results perhaps indicate a slight dependence on bro-mide for the i n i t i a l rate. An. increase in rate by a factor of 2, was observed when the [Br~] was varied from 0.50M to 0.40M. On further decreasing [Br~] to 0.25M, the 44 Table VI Carbonylation of [RhBr 4(H 20) 2] Temperature dependence of k 2 ([Rh] = 0.020M; 0.5M HBr) Temp. CO [C0]X k 2 °C mm 104,M* M _lsec"l 40 759 7.91 0.050 45 743 7.35 0.065 50 723 6.85 0.109 55 698 6.34 0.157 *[C0] computed from solubility data of S e i d e l l 1 2 6 assuming the solubility to be the same as in pure water. 45 46 i n i t i a l rate increased by only a further 10%, compared to the previous value at 0.40M. This inverse dependence results at least in part,from a varia-tion in CO solubility on altering the reaction media. Both H2 and CO show increased solubility with increasing HCl0^ in solutions with a constant 2 7 total concentration of hydrochloric and perchloric acids. ' In addition, the method of equilibration for the rhodium bromide solutions in 0.5M HBr (see sec. 3.2.) differed from the procedure employed during the [Br -] variation studied (0.40-0.10M). Stock solutions of 0.100M RhBr3-2 H20 were made up in 0.5M HBr. The solutions were diluted to 0.020M rhodium and adjusted to the desired [Br~] concentration by adding appropriate amounts of HCl0 4; the diluted solutions were then equilibrated at 80° for 20 hours prior to reaction. The fact that a different stock solution was used for the experiment at 0.50M may possibily explain the anomaly in the results. Since the variation in rate at the other [Br~] is small, the effects can be mainly attributed to increased CO solubil ity, al -though i t must not be overlooked that the 0.10M [Br~] solution contains increased amounts of [RhBr 3(H 20) 3]. An inverse dependence on [Br -] solution is observed for the maximum rate which is essentially independent of CO (Table VII) over the concentration range 0.25-0.50M. The rate at 0.10M [Br~] i s , how-ever, lower than at 0.25M. The effect of acid on the reaction at 40° was studied by varying the composition of HBr:NaBr in the media while maintaining a total bromide concentration of 0.50M. The data are summarized in Table VIII, method (c) being used to compute the values of k2- A slight inverse acid dependence is observed for both the i n i t i a l reaction and the autocatalytic one. Figures 13 and 14 show plots of k1 and k ? respectively versus [H +]~*. 47 Table VII Carbonylation of [RhBr^HgO)] Effect of [Br -] on the rates of reaction at 40° ( [ R h I H ] = 0.020M, [H+] * 0.5M) [Br -] [C0]X I n i t i a l rate X Maximum rate X k 2 104,M* 106,Msec-l 106,Msec-l M-l sec 0.50 7.91 2.26 6.04 0. 050 0.40 (15.9) 4.46 9.16 0. 070 0.25 (17.2) 4.82 11.33 0. 090 0.10 (17.8) 5.04 8.77 0. 062 Values in brackets are calculated using the i n i t i a l reaction rates and the known solubility of 7.91 x IO - 4 M in 0.5M HBr and assuming that the rate variation is due entirely to CO solubility effects. 48 Table VIII Carbonylation of [RhBr 4(H 20) 2]" Effect of [H+] on the rates of reaction at 40° ([Rh] = 0.020M; [CO] assumed = 7.91 x 10~4M throughout; [Br -] = 0.5M) [H+] In i t i a l rate lOMsec-"1 X k-| M"1 sec" Maximum rate X 1 lOMsec" 1 k2* -1 M' sec 0.50 2.48 0.16 6.04 0.048 0.40 2.25 0.14 5.78 0.051 0.25 2.49 0.16 6.69 0.054 0.10 3.86 0.25 7.71 0.059 Calculated using method (c). 49 Figure 13. Inverse acid dependence of lq , 40°, 0.020M Rh 759 mm CO, 0.5M [Br"]. 50 Figure 14. Inverse acid dependence of k2, 40°, 0.020M Rh 759 mm CO, 0.5M [Br -]. 51 3.5. Discussion 3.5.1. Production of the Autocatalytic Species CO reacts with aqueous solutions of rhodium bromide complexes under mild conditions to form the [Rh(C0) 2Br 2]~ species. The f i r s t term of the rate law (3.5) is concerned with the production of the autocatalytic species, [Rh(C0) 2Br 2]~, and is f i r s t order in both Rh*11 and CO. The reduction of metal complexes by CO is thought to involve at some stage an "insertion reaction" of the CO molecule between the metal and co-ordinated 4 12 7 water or hydroxide, ' ' CO insertion reactions have been studied exten-128-133 sively and i t appears that before the insertion reaction can occur, the CO must co-ordinate to the metal; furthermore, i t is lik e l y that the 134 135 net insertion then involves a group i n i t i a l l y bonded to the metal ' (eg. h^ O or OH ) migrating to the co-ordinated carbonyl group. The kinetic data in the present study provide information only about the step involving co-ordination of CO, but based on the knowledge of the insertion reaction, the following scheme seems reasonable for the formation of the [Rh(C0) 2Br 2]" species: k l [RhBr„(H90)9] + CO ! » [RhBr,(C0)(H90),] + Br" 4 L L l t d . ( 3 J 1 j or rRhBr4(C0)H20]" + H20 I I [RhBr3(C0)(H20)0H]" + H+ [-Rh-C0 2H]Rh 1 + C02 + H+ (3.12) or [RhBr4(C0)0H]2" III II Rh1 + 2C0 — ^ [Rh(C0) 2Br 2] (3.13) 52 The relatively small dependence on added acid suggests that the major path in the rate determining step involves reaction with the diaquo species, [RhBr 4(H 20) 2]~; the intercept on the k-| axis of Figure 13 gives the contribution of this acid independent path (equation 3.11). Ionization of co-ordinated water followed by hydroxide migration then gives the car-boxylate complex (III) which decomposes to give Rh*, C02 and H+ (equation 3.12). The reaction of Rh* with CO (equation 3.13) must be rapid since unco-ordinated Rh* in aqueous acid solution rapidly gives metal. 7 This mechanism yields for the i n i t i a l reaction: rate = k[RhBr 4(H 20) 2][C0] (3.14) The slope of Figure 13 indicates that there is also a contribution to the carbonylation by an acid dependent path; the direct inverse depend-ence for this path i s consistent with a pre-equilibrium involving acid dissociation of [RhBr 4(H 20) 2]~ followed by a rate determining S^ 2 reaction of CO with the hydroxy species: [(H 20)Br 4Rh-(0H 2)]" , ^ > [(H20)Br4Rh-0H]2" + H+ (3.15) IV k' IV + CO > [(H20)Br3Rh(C0)0H]~ + Br" (3.16) or [Br4Rh(C0)0H]2" + H20 Hydroxide migration again produces the unstable carboxylate complex (III). The overall, i n i t i a l rate w i l l be represented by equation (3.17). rate = {k' [RhBr4(H20)(OH)] + k[RhBr 4(H 20) 2]}[C0] (3.17) 53 where [RhBr 4(H 20)0H] 2" is K a/[H +] [RhBr 4(H 20) 2]. Since Ka is very small (eg. Ka for [Rh(H 20) 6] 3 + ca. 10 - 3; Kg for [ I r C l 5 ( H 2 0 ) ] 2 _ is ca. 10" 1 0) , 1 3 6' 1 3 essentially a l l the Rh is present as [RhBr 4(H 20) 2]~ and equation (3.17) becomes: rate = (K f lk7[H +]+k)[Rh] T o t a l[CO] (3.18) The tabulated values of k, (Table I) can be identified with K.k'/[H+]+k. • a Figure 13 gives k ca. 0.11M"1 sec"1 Kk ' ca. 0.014 sec - 1. a Comparison of k, 0.11M-1 sec" 1, with the combined rate constant k-|, 0.14M 1 sec - 1, in 0.50M acid shows that the acid independent path represents the major protion of the reaction at 40° at this acidity; at 0.10M acid, the acid dependent path contributes about 50% of the total reac-III 2 tion. The behaviour contrasts with that of the' Rh chloride complexes for which essentially a l l of the reaction proceeds via an acid dependent path over the acid concentration range of 0.10-0.50M at 80°. To some extent this difference may reflect a lower acidity of [RhBr 4(H 20) 2]" relative 2-to [RhClg(H20)] , although comparable figures for the pKa's of these com-plexes have not been reported. The activation parameters wi l l be discussed in comparison with some related species in Chapter IX. 3.5.2. Autocatalytic Reaction The overall autocatalytic stage of the reaction may be represented by equation (3.19). Rh I(C0) 2 + Rh m(H 20) — » 2RhI(C0)2 + C02+ 2H+ (3.19) 54 The absence.of a CO dependence at the higher pressures and a f i r s t order dependence in both Rh1 and Rh 1 1 1 indicates that the rate-determining step involves a bimolecular reaction between [Rh(C0) 2Br 2]~ and a Rh 1 1 1 species. The path i s thought to involve a mixed valence bridged transition state [Rh 1••-Cl••-Rh 1 1 1] analogous to that originally described 1 o o for the Pt(II) catalyzed substitution reactions of Pt(lV). More recently, Rh1 species have been similarly postulated as the active catalyst for the nucleophilic substitution of Rh(III) complexes through a bridged mechanism. 2,12,13,15-18 A li k e l y mechanism for the autocatalytic reaction can be written ^9 T T T T 0 _ H20 +- [Rh(C0) 2Br 2]" + [RhBr 4(H 20) 2]"—^4 [(-H20)Br2('C0)2Rh -Br-Rh i l iBr 3(H 20) 2] £ S l 0 W V (3.20) V > [(H 20)Br 2(C0) 2Rh I I I - B r4Rh IBr 3(H 20) 2] 2" (3.21) VI VI >[Rh H I(C0) 2Br 3(H 20)] + Rh1 (3.22) "VII fast 3" Rh1 + CO . B/~ > [Rh^COJBr.,]2- (3.23) (3.24) VII >, [Rh :(C0)Br 3] 2" + C02 + 2H+ VIII [Rh I(C0)Br 3] 2" + CO , K » [Rh 1(C0) 2Br 2]" + Br" (3.25) The planar [Rh(C0) 2Br 2]" complex forms the bridged intermediate via a bromide ligand of the Rh 1 1 1 complex, and also co-ordinates a sixth ligand 55 (probably water) giving (V). A two electron transfer then occurs with subsequent breaking of the labile Rh1 ligand bond to give the Rh 1 1 1 pro-duct (VIJ and regeneration of the Rh1 catalyst. Compound (VII) contains both co-ordinated carbonyl groups and water and would undergo decomposition to the rhodium(I) product (VIII); this then probably reacts with CO to regenerate catalyst (product). The observed CO pressure dependence at lower [CO] (Figure 10) could arise from a pre-equilibrium involving forma-tion of the [Rh(C0) 2Br 2]~ complex (eq. 3.25) from a carbonyl rhodium(I) species such as VIII. At least one of the carbonyl groups must co-ordinate rapidly to stabilize the Rh1 against reduction to the metal. The rate law at the maximum rate becomes: -d[C0]/dt = k l ^ R h ^ T ^ C 0 ^ + k / [ R h ] T 2 [ C 0 ] 2 ( 3 > 2 6 ) (2K[C0] + [Br]) (2K[C0]+[Br},)2 At high pressures of CO, when 2K[C0] > [Br], the rate expression reduces to equation (3.27): -d[C0]/dt = k ] [ R h ] T [ C 0 ] + (3.27) This expression (3.27) is in agreement with the experimentally observed rate law, equation (3.5) in which k-j represents k-j/2 and k 2 represents k2/4. At lower CO pressures, when 2K[C0] - [Br], the rate law is given by equation (3.26) and accounts for the observed decrease in rate. The possibility that the reaction of Rh^CO) with CO to give Rh I(C0) 2 becomes rate determining cannot be ruled out. 2_ There is no kinetic evidence to suggest that [Rh(C0)Br3] might 56 i t s e l f react with [RhBr 4(H 2 0) 2 ] " to give a bridged species (cf. equation 3.20). Such a reaction involving a doubly negative anion is like l y to be less favourable than reaction (3.20). The autocatalytic stage could alternatively take place through a carbonyl bridged intermediate [Rh I - - -C0---Rh 1 1 1], f o l i o quent decomposition of an 1,• CO-•-Rh 11  followed as before by a two electron tranfer and subse-SloW T TTT 2 [Rh(C0) 2Br 2]" •+ [RhBr 4(H 20) 2]~ R Q g >- [(H 2 0)Br 3 (C0)Rh -C0-Rh 1 1 1Br 3(H 2 0 ) 2 ] "+ Br" IX (3.28) IX » [ ( H 2 0 ) B r 3 ( C 0)Rh I I I - C 0{Rh I B r 3 ( H 2 0 ) 2 ] 2 " (3.29) X X > [-Rh H I ( C 0 y H 2 0 ) - ] + Rh1 (3.30) XI XI > Rh !(C0) + C0 2 + 2H + (3.31) aquodicarbonylrhodium(III) species XI to Rh*(C0). However, the carbonyl bridge is thought to be much less l i k e l y than the bromide bridge, since the former requires co-ordination into the relatively substitution inert Rh*** complex. Substitution of an unco-ordinated CO molecule to Rh*** has been shown a l -ready to be slow in the i n i t i a l , reaction (see sec. 3 . 3 .1 . ) . The plot of k 2 versus [H +]"* shown in Figure 14 possibly indicates a very slight acid dependence and shows that a reaction path via a hydroxy rhodium(III) species is relatively unimportant. The figure shows that the aGid independent path through the [RhBr 4(H 2 0) 2 ]~ complex predominates for the autocatalytic reaction with a rate constant 0.047M -* sec"*. The slope 57 of Figure 14 gives K k" ca_. 0.001 sec"1 where k" is the rate constant for a _3 the route involving the hydroxy species. K is certainly <10 and thus implies that a Rh*1* hydroxy complex is much more efficient for formation of the Rh*-Rh*** intermediate. A similar conclusion was drawn for the chloride systems, although in the chloride case the acid dependent path was more definitely demonstrated. Minor ionic effects (varying [H +]/[Na +]) might account for the very slight rate variation with added [H +] in the bromide systems, as there is no strong evidence for involvement of a hydroxy species. On electrostatic arguments an aquo complex would be preferable to a hydroxy complex for a reaction such as (3.20), but l a b i l i z i n g effects 139 of OH are well-documented and could be important. There appears to be a genuine inverse bromide dependence over the concentration range 0.25-0.50M. This may result from the effect of bromide inhibiting formation of the bridged intermediate. Rhodium(I) chloride 34 38 doubly bridged dimer complexes are known to be cleaved by donor ligands. ' [Rh(C0) 2Cl] 2 + 2L > 2Rh*(C0)2ClL (3:32) Since the postulated bridged intermediate involves only one bridging ligand, cleavage by Br" would presumably occur more readily in this case than in the doubly bridged dimer. The efficiency of the reduction of the newly formed Rh*** species requires the presence of a co-ordinated H20,which originates as a sixth ligand co-ordinating axially on the [Rh(C0) 2Br 2]~ species. Occupancy of the sixth ligand position of the Rh* complex by bro-mide would clearly also inhibit the reduction mechanism. The magnitude of the inverse bromide pentachloro system. the inverse  effect is similar to that observed for chloride in the 2 58 It is clear that the good linear Arrhenius plots obtained verify the dominance of the acid independent paths for the reaction and present some evidence against a composite rate constant for k-j and k^ respectively. The activation parameters suggest that reduction through the bridged intermediate is more favourable, owing to a much lower activation energy. The parameters obtained for k^ are similar to those reported for redox reactions involving bridged intermediates for some Rh^-Rh 1, C o 1 1 1 - ^ 1 1 , C r I I ! - C r n , and P t I V - P t ! I systems. 1 6' 1 4 0' 1 4 1 A negative AS^ seems reason-able for formation of a bridged intermediate especially for interaction 142 between two negatively charged species. 59 Chapter IV CARBONYLATION OF A RHODIUM(III) BROMOCARBONYL COMPLEX IN AQUEOUS HYDROBROMIC ACID SOLUTION 4.1. Introduction The importance of the role of H20 has already been suggested for the reduction of Rh1** carbonyl anions (Chapter III). In this chapter, the general reduction of such species is demonstrated for the pentabromocarbonyl-rhodate(III) complex i t s e l f (which involves an autocatalytic process), and for a catalytic reduction of iron (III) using the [Rh(C0)Br5] /CO system. 4.2. The Reaction of [Rh(C0)Br 5] 2" in HBr Solution with CO Several preliminary kinetic CO gas-uptake experiments were conduc-2- 5 ted with solutions containing [Rh(C0)Brg] prepared from the corresponding Cs salt, [Cs],[Rh(C0)BrK] (reflectance spectrum, \ v = 440 nm; x „ = 550 nm). The i n i t i a l absorption spectrum for the compound in 0.5 M HBr at room temp-erature showed maxima at 435 nm, e = 1050, and 550 nm, e = 119. Considerable di f f i c u l t y was encountered in achieving reproducible kinetic results. Differ-ent preparations of stock samples of [Cs]2[Rh(C0)Brg] produced a variation in rate constants, yet results within each sample lot were consistent in them-selves (see sec. 4.8.). Solutions of [Rh(C0)Br 5] 2" in 0.5 M HBr absorbed CO readily at 60°. The total net measurable gas uptake corresponded to an apparent 1:1 mole ratio of gas:Rh; experiments in the presence of a sodalime tube showed, however, that the true CO uptake corresponded to a 2'.1 mole ratio of C0:Rh, with 1 mole of C02 being formed. A mass spectral analysis of the mixture at the end 60 of the reaction (in the absence of socialime) indicated the production of COg- The original orange solution became yellow U m a x = 330 nm, e = 3290) indicating reduction to a Rh* state with absorption maxima similar to that found for the anionic species [Rh(C0) 2Br 2]~ (Chapter II I ) . These observa-tions suggest an overall stoichiometry for the reaction given by the follow-ing equation". H20 + Rh H I(C0) + 2C0 >Rh T(C0) 2 + C02 + 2H+ (4.1) (bromide ligands have been omitted) 4.3. Kinetics of the Reaction in HBr Media Typical uptake plots shown in Figure 15 exhibit a short i n i t i a l region of slow gas absorption which was possibly attributable either to dissolution of [Cs]2[Rh(C0)Brg];or to a slow build up of a Rh* species; followed by a second region of more rapid CO uptake. All analyses of kinetics were thus based on this second region. The uptake plots in Figure 15 appear sigmoid in shape suggesting autocatalytic behaviour. Maximum rate measurements were made over a series of experiments at different i n i t i a l rhodium concentrations (Table IX). When the maximum rates were plotted agianst the i n i t i a l rhodium concentration at constant [CO], a linear relation was not observed. It can be shown that a rate proportional to the product [Rh***][Rh*] becomes a maximum when [Rh***] = [Rh*] = l / 2 [ R n ] T o t a i . A plot of the maximum rates (Figure 16) against the product [Rh***] x [Rh*] did however result in a linear dependence. Hence, the dominant term in the rate is expressed by 0 2.0 4.0 .6.0 8.0 10.0 12.0 14.0 Time x 10 , sec Figure 15. Rate plots for the reduction of [Rh(C0)Br5]2T 60°, 616 mm CO, [Rh i n(C0)]: (0) 0.0207M; (A) 0.0086M, 0.5M HBr. 62 Table IX Carbonylation of [Rh (CO) Br 5] 2~ Summary of kinetic data in 0.5M HBr at 60° [Rh]xl0 2, CO [C0]xl0 4, Maximum rate x 105, k 1 b M mm M* Msec - 1 M-isec - 1 0, .86* 616 5. .38 1.27 0, .48 1. .52a 616 5. .38 2.65 0, .35 2, .07* 616 5 .38 4.13 0. .39 1. .05 711 6. .21 2.82 1. ,03 1. ,05 616 5, .38 2.76 1. .17 1. ,05 308 2, ,69 2.09 0. ,83 1. .05 76 0. ,66 0.43 0. ,17 *[C0] computed from solubility data of S e i d e l l 1 ^ 0 assuming the solubility to be the same as in pure water, d i f f e r e n t sample of [Cs] 2 [Rh (CO) Br 5] . bAbove 600 mm CO, k 0' = k 0. 63 -d[CO]/dt = k2' [Rh I I I][Rh I] (4.2) The fact that the plot in Figure 16 does not pass through the origin suggests that a rate term resulting from the i n i t i a l stage of the reaction is not completely insignificant. The rate law can be expressed as: -d[C0]/dt = [ R h m ] n + k2' [Rh I I I][Rh I] (4.3) where k-j1 and kg' contain a l l physical and chemical variables other than the dependence on rhodium. Information concerning the nature of the i n i t i a l reaction was obtained from studies carried out in the presence of added ferric ion (sec. 4.5.). A value of kg1 calculated from the slope of the linear plot (Figure 16) is 0.32 M~* sec - 1. In order to determine the CO dependence, maximum rate measurements were made over a series of experiments in which the partial pressure of CO (p'CO) was varied while the i n i t i a l [Rh***] was maintained constant (Table IX). The rate was essentially independent of CO pressure above 600 mm, but showed a CO dependence at lower pressures (Figure 17). Thus, k^ ' at higher CO pressures is a true second order rate constant and the rate law becomes'. -d[C0]/dt = k^ [Rh 1 1 1]" + k 2 [Rh***][Rh*] (4.4) The complete flattening off of Figure 17 also indicates that the k-j1 term is independent of CO (sec. 4.4.). A method of calculating k 2 considers the final part of the overall reaction rate: 64 Figure 16. Dependence of maximum rate on [Rh ] x [Rh ], (60°, 616 mm CO, 0.5M HBr), 65 0 200 . 4 0 0 ' ' 600 "'' 800 p'CO, mm Figure 17. Effect of CO pressure on maximum rate, 60°,.0/0105M Rh i n.(C0), 0.5M HBr. 66 -d[CO]/dt = k 2 [Rh I I I][Rh I] (4.5) Integration of this expression yields equation (4.6) which takes no account at a l l log[Rh*] - l o g [ R h I H ] = k 2 [ R h ] T o t a l t/2.3 + constant (4.6) of the f i r s t reaction. Plots of log[Rh*] - log[Rh***] versus time give a straight line of slope k 2[Rh] T o t a-| /^.3 (Figure 18). As expected,these plots deviate from linearity during the f i r s t part of the reaction. k 2 values calculated by this method are summarized in Table IX for the system at higher CO pressures, and are in reasonable agreement with the value deter-mined from the slope of Figure 16. Quite consistent values for each sample lot are obtained. 4.4. The I n i t i a l Reaction In order to study the i n i t i a l part of the reaction during which [Rh1] was being formed, i t was necessary to "extend" that region of the uptake plot to avoid possible interference by the dissolution rate of [Cs]2[Rh(CO)Brg]. If as suspected, the f i r s t part of the reaction consisted 2- I of decomposition of [Rh(C0)Br5] by water to a Rh species, the presence of an oxidant such as Fe*** should lead to an overall catalytic process during which Rh* is oxidized to Rh***, followed by absorption of CO to give Rh I I J(C0) (see Chapter III) and resumption of the catalytic cycle. 67 0 2.0 4.0 6.0 8.0 Time x 10 , sec Figure 18. Plots of log [[Rh^C0) 2Br 2 n-log[Rh I n(C0)] vs. time for various i n i t i a l Rh^^CO) concentrations (60°, 616 mm CO, [Rh mC0]: (0) 0.0207M; (A) 0.0152M, 0.5M HBr). 68 4.5. Kinetics of the I n i t i a l Reaction in the Presence of Fe*** In the presence of a socialime tube, mixtures of [Rh(C0)Br,.] and iron(III) perchlorate in 0.5 M HBr gradually absorbed CO. The visible spec-trum of the final solution (A = 330 nm, e = 3460) again corresponded closely to [Rh(C0) 2Br 2]~. The total concentration of iron was determined 2+ 143 spectrophotometrically on the final solution as the Fe(o-phen)3 complex, 5 144 I 4 38 after f i r s t precipitating the Rh carbonyl species ' as [ln-C 4H g) 4N][Rh(C0) 2Br 2]. The Fe1** was almost completely reduced to Fe** (97%) during the reaction with [Rh(C0)Br5]2"/C0. The uptake plots (Figure 19) were unusual in shape. After ca_. 500 seconds, CO absorption was linear giving a total uptake of (^- - y) moles of CO per n moles of Fe*** and y moles of Rh*** respectively. The second part of the uptake plot (from the end of the linear region to the levelling off region) corresponded to a 3:1 mole ratio of C0:Rh. If excess Fe*** was 2-added to the [Rh(C0)Br5] solution at 60° under N2, C02 evolution was obser-ved. Quantitative experiments on the evolution showed that 1 mole of C02 was evolved per mole of Rh. In the absence of CO (ie. in the presence of an oxidant Fe***/N2, or air), a green solid (reflectance spectrum A m a x = 485 nm; A m a x = 583 nm) ?-precipitated from solutions of ca_. 0.020 M [Rh(C0)Br5] in 0.5 M HBr at 60°. Analysis of the compound gave Rh, 15.66; Br, 54.17%. Calcd. for [Cs] 3[Rh 2Br g] Rh, 15.55; Br, 54.34%. The far i r spectrum of the solid showed v R h _ B r at 256 and 254 cm"*. When dilute solutions ca_. 0.002 M [Rh(C0)Br 5] 2" in 0.5 M HBr were exposed to air at RI, the final visible spectrum featured a maximum at 532 nm, e = 170; characteristic of Rh*** bromo species. 0 1 .0 2.0 3.0 4.0 -3 Time x 10 , sec ?- T T T Figure 19. Rate plots for the [Rh(C0)Br5] catalyzed reduction of Fe , 60°, 616 mm CO, 0.046M Fe 1 1 1 , [Rh H ICC0)]: (0),0;0105M; (A) 0.0052M, 0.5MHBr. 70 The linear part of the uptake plot (Figure 19) was used to inves-tigate the kinetics in this region. Variation of [Rh(C0)Brg] while the [CO] and [Fe***] are maintained constant indicated a f i r s t order dependence in rhodium (Figure 20). Experiments varying [Fe***] show the reaction to be independent of oxidant. The absorption spectrum for the solution measured when ca_. half the Fe*** has been reduced showed a maximum at 550 nm, e = 100, indicating that a l l the rhodium is essentially present at [Rh(C0)Br,-] . The reaction becomes almost independent of CO at pressures above 600 mm and the rate decreases only slightly with pressure between 300 and 500 mm (Figure 21). The rate law in the high CO limiting reaction may be formulated as: -d[C0]/dt = k^Rh 1 1 1] (4.7) k-| calculated from the slope of the linear plot of Figure 20 is 0.0014 sec - 1 and values of k-| are summarized in Table X. In the absence of CO, the C0g evolution is consistent with a reac-tion such as H 0 Rh***(C0) + 2 F e m 1 > Rh**1 + 2Fe** + 2H+ + CO,, (4.8) The co-ordinated CO is acting overall as a 2 equivalent reductant. Under CO, regeneration of Rh***(C0), leads to a catalytic reduction of Fe***; this occurs during the i n i t i a l linear region according to the net stoichiometry: f H20 + Rh***(C0) + nFe*** + {\ - 1)C0 >Rh*** + nFe** + C^O,, + nH+ (4.9) 71 Figure 20. Dependence of i n i t i a l rate (Fe reduction) on [Rh(C0)Br 5] 2", 60°, 616 mm CO, 0.5M HBr. 72 Table X Summary of kinetic data for the [Rh (COjBrv] catalyzed reduction of Fe by CO in 0.5M HBr at 60° [Rh]xl0 2, M [ F e H I ] x l 0 2 , M CO mm [C0]xl0 4, M* 5 Ini t i a l rate x 10 , Msec-1 k, ' a x 10 2, 1 -1 sec 1.05 7.7 616 5.38 1.54 0.15 1.05 4.6 616 5.38 1.40 0.13 0.76 4.6 616 5.38 1.10 0.15 0.52 4.6 616 5.38 0.75 0.15 1.05 4.6 462 4.04 1,23 0.12 1.05 4.6 462 4.04 1.29 0.12 1 .05 4.6 308 2.69 1.10 0.11 *[C0] computed from solubility data of Seidell assuming the solubility to be the same as in pure water. aAbove 500 mm CO, k-,1 - k,. 74 Reduction of the i n i t i a l Rh***(C0) complex to a Rh1 species i s required (sec. 4.7.1.) prior to electron transfer to Fe***. 4.6. Carbonylation of Bromorhodate(III) Complexes after Reduction of Fe*** After the linear CO uptake region, (Figure 19) which corresponds to the reduction of Fe***, a second region of CO uptake occurs for the carbonylation of Rh*** to produce [Rh*(C0) 2Br 2]~. The stoichiometry for this stage corresponds to: H 0 Rh*** + 3C0 > Rh*(C0)2 + C02 + 2H+ (4.10) The rates for carbonylation of Rh*** in the presence of a sodalime tube at 60°, measured from the slopes of the second uptake region,are summar-ized in Table XI. The results show a f i r s t order dependence in rhodium, when [CO] is maintained constant. The reaction becomes independent of CO at pres-sures above 500 mm and the rate decreases only slightly with pressure between 300 and 500 mm (Figure 22). A similar pressure dependence was observed for the kinetics during the reduction of Fe*** (sec. 4.5.). The rate law for the second uptake region at higher CO pressures may be formulated as: -d[C0]/dt = k3[Rh1**] (4.11) Since the stoichiometry for this stage of the uptake plot is 3C0:Rh, -d[C0]/dt =-3d[Rh***]/dt, -d[Rh***]/dt = k3[Rh***]/3. (4.12) Table XI Carbonylation of Rh*1* Summary of kinetic data after reduction of Fe*** in 0.5M HBr at 60° [Rh]xl0 2, M CO mm [C0]xl0 4, M* Rate x 10 5, Msec"* k 3 , a x 10 2, sec"l k 3 7 3 b x 102, sec"l 1 .05 616 5.38 4.72 0.45 0.15 1.05 616 5.38 4.75 0.45 0.15 0.76 616 5.38 3.46 0.45 0.15 0.52 616 5.38 2.15 0.41 0.14 1 .05 462 4.04 4.32 0.41 0.14 1 .05 462 4.04 4.38 0.42 0.14 1.05 308 2.69 3.47 0.33 0.11 *[C0] computed from solubility data of Sei del 1 1^ D assuming the solubility to be the same as in pure water. aAbove 500 mm CO, k3' = k 3 > bAbove 500 mm CO, k,'/3 = k,/3. 76 77 4.7. Discussion 4.7.1. Reduction of [Rh(C0)Brv]2" by Water The autocatalytic production of [Rh(C0) 2Br 2]~ from [Rh(C0)Br 5] 2" in the absence of Fe*1* (Figure 15) almost certainly proceeds via a mechanism similar to that discussed in Chapter III ; involving bridged [Rh*** Br"'"Rh*] intermediates. The i n i t i a l formation of Rh* is thought to require reduction of CO by R hI** ( C0) H2° ) Rh* + C02 + 2H+ (4.13) k l Rh1 + 2C0 ^ > Rh*(C0)2 (4.14) 136 137 I water through an "insertion reaction". Rh is then stabilized by rapid reaction with CO (equation 4.14). Evidence for the decomposition of Rh 1 1 1 and Ir*** carbonyl complexes by water has been reported. 3 , 4'* 2 8 Information on the i n i t i a l decomposition of Rh***(C0) was obtained by studying the effect of an oxidant, Fe***, on the i n i t i a l rate. The fact that C02 was evolved even under N 2 in the absence of CO verified the existence of a reductive step (equation 4.15). The catalytic reduction of Fe*** was independent of [CO] and [Fe***] which strongly indicates that a Rh* species (formed by decomposition of [Rh(C0)Br5] ) is the actual reducing species. The following mechanism would be consistent with the data: [Rh***(C0)Br 5] 2- H k 2° > R h I + C 02 + 2 H + ^* 5> 78 R h I + 2 F e i n f a s t > Rh 1 1 1 + 2Fe** (4.16) Rh 1 1 1 + CO — > [Rh I n(CO)Br 5] 2" (4.17) The rate-determining step involves aquation of the pentabromocarbonylrhodate (III) ion. The spectral and kinetic data show that the rhodium(III) catalyst is regenerated as the pentabromocarbonyl species, although the nature of the co-ordinated ligands on the intermediate species in reactions (4.15 to 4.17) is not known. Since (^  - 1) moles of CO are absorbed per mole of Rh (CO) during the reduction of n moles of F e 1 1 1 , i t appears that the reductant is an unco-ordinated Rh1 species. This also implies that at the completion of ferric reduction, the rhodium(III) complex present contains no carbonyl. Based on the stoichiometric result, the reduction of Fe*** must be probably faster than co-ordination of Rh* by CO,since unco-ordinated Rh* in acid solution produces metal? In the case of a Rh***(C0) complex in DMA, i t has been suggested 3 that decomposition proceeds via hydroxide. R h I n(C0) ° H~ ( H 2 0 ) ^ !>(C0)2H)] >Rh* + C02 + H+ (4.18) Since the acid strength was not varied in the present system, i t is uncertain whether attack at the carbonyl group involves water or hydroxide. Furthermore, i t is d i f f i c u l t to determine whether the hydroxide (or water) need be co-ordinated to the metal before migration to the carbonyl (ie, the "insertion mechanism"), as seems usually postulated in similar aqueous systems such as 79 145 146 147 ethylene reduction of metal ions, or hydration of acetylene; ' there is the possibility that free hydroxide attacks the co-ordinated carbonyl group directly. After transfer of two electrons to the metal, the intermed-iate carboxylate undergoes a rapid decomposition to form Rh1, C02, and H+. The non-complementary electron transfer reaction between Rh* and Fe*** could take place by a two step mechanism involving a transient Rh** species, or alternatively by a one step termolecular reaction. The latter mechanism has rarely been postulated for reactions involving only metal ions 140 148 and would likely be very slow. ' In such a case, the stabilization of a Rh* species by CO would be expected. The electron transfer could proceed by an inner sphere mechanism involving a bromide bridged intermediate [Fe***'"'*Br''''Rh*], or by an outer sphere mechanism. During reduction of Fe***, one mole of CO is absorbed by Rh*** regenerating the catalyst, Rh***(C0). Above CO pressures ca_. 500''mm, this reaction (equation 4.17) must occur more rapidly than reaction (4.15), but at lower CO pressures this step (equation 4.17) appears to possibly become rate-determining. This accounts reasonably for the variation of rate with CO pressure (Figure 21). Thus, the slope of the linear region in Figure 21 at lower CO pressures represents an estimate of the rate for the reaction between CO with a Rh*** complex. The rate constant at 60° calculated from the slope of Figure 21 is 4.5 Nf* sec"*, which is close to the value of 2.6 M * sec * at 60° extrapolated from the Arrhenius plot (Figure 8) for the direct carbonylation of [RhBr 4(H 90) 2]" (Chapter I I I ) . The somewhat larger rate constant may imply a carbonylation reaction involving a different bromorhodate(III) complex. The fact that [Cs] 3[Rh 2Br g] is isolated during C02 evolution experi-o 80 merits under N£ indicates that these solutions contain bromorhodate(III) species. [Cs] 3[Rh 2Br 9] has been prepared by the addition of stoichiometric quantitities of CsBr to aqueous solutions of rhodium tribromide. 1 4 9 4.7.2. The Autocatalytic Reaction ([Rh(C0)Br5]2"/C0) The autocatalytic stage of the reaction (Figure 15), may be rep-resented by the following mechanism: [Rh(C0).2Br2]~ + [Rh(C0)Br 5] 2: ^ [(H 20)Br 2(C0) 2Rh I-Br-Rh 1 1 1 (C0)Br 4] 3' H2° j (4.19) I » [(H 20)Br 2(C0) 2Rh H I- Br-f- Rh^OBrJ 3" (4.20) 4-II H >[Rh I n(C0) 2Br 3(H 20)] + Rh1 (4.21) III III >RhJ + C02 + 2H+ Rh1 + CO [Rh^COJBr,]2" fast J IV (4.22) (4.23) K(k4/k ) IV + CO , [ R h ^ C O ) ^ ] " + Br" (4.24) At the higher CO pressures, the independence of rate on CO shows that at these conditions, k ] 1 and k2' (equation 4.3) must be essentially 81 independent of any CO term. Thus, the rate-determining step for the auto-catalytic stage involves a bimolecular reaction between [Rh(C0) 2Br 2]~ and [Rh(C0)Br5] . The detailed mechanism (equation 4.19 - 4,24) is thought to involve a bridged transition state [Rh 1" "Br"'"Rh111] analogous to that described in Chapter I I I . The overall rate law at high pressures of CO for the region of maximum rate is represented by k 1[Rh] T k [ R h ] 2 -d[C0]/dt = :.' 2 1 + 1 4 1 (4.25) The observed CO dependence at lower pressures of CO (Figure 17), could arise from a pre-equi1ibrium involving formation of the [Rh(C0) 2Br 2]~ 2-complex from species such as [Rh(C0)Br3] , or from reaction (4.24) becoming rate determining (cf. Chapter I I I ) . At low pressures of CO, i f reaction (4.24) is a rapid equilibrium, the rate law simplifies to k]K[Rh]T[C0] k 2K 2[Rh] T 2[C0] 2 " d [ C 0 ] / d t = (2K[C0j+[Br])+ (2KrC0MBr]) 2 <4-26) A value of K, determined from the maximum rate (Table IX) at 0.0105 M [Rh] and 0.66 x 10~4 M [CO], and substituting the values of k-j (Table X), and k 2 (Table IX) into equation (4.26), is ca_._ 2 x 10 3. However, at low CO pressures, such an equilibrium (4.24) implies that the product would be Rh^CO), and yet the measured stoichiometry is s t i l l 2C0:Rh. Thus, the effect of an equilibrium such as equation (4.24) being f e l t at low CO pressures appears unlikely. A more probable explanation is that the reaction of Rh^CO) with CO to give Rh I(C0) 2 becomes rate-deter-mining. 82 4.7.3. Carbonylation of Bromorhodate(III) Complexes After Reduction of Fe III Unlike the reaction of [RhBr 4(H 20) 2]" with CO studied in Chapter II I , the carbonylation of Rh 1 1 1 which follows the F e 1 1 1 reduction does not exhibit autocatalytic behaviour. A comparison of the rate constants, k^/k^. for the reduction of F e 1 1 1 and the subsequent carbonylation of the Rh 1 1 1 species shows them to be in the ratio of 1:3. Since -d[Rh I H]/dt = k 3[Rh H I]/3 (4.27) and k3/3 - k], (4.28) the same rate-determining step (reaction 4.15) is almost certainly involved. This reaction is then followed by rapid uptake of 2 moles of CO; this accounts Rh1 + 2C0 > Rh J(C0) 2 (4.29) for the factor of 3 between the rate constants. Furthermore, at CO pressures below ca_. 500 mm, the step (4.17) appears to become rate-determining (as described in sec. 4.7.1.). The rate constant calculated from the slope of the linear region of Figure 22 i s 12.6 M-1 sec"1, again a factor of almost 3 greater than the value of 4.5 M-1 sec"1 calculated from Figure 21 (sec. 4.7.1.) 4.8. Reproducibility of Kinetic Results [Cs]2[Rh(C0)Brg] was prepared according to method of Cleare and G r i f f i t h 5 substituting RhBr3-2 H"20 for hydrated Na3RhBrg. Rhodium(III) t r i -83 bromide was boiled in a i r with a mixture of concentrated HBr and formic acid for 15 minutes. The original red colour of the Rh*** solution became yellow and f i n a l l y orange, On addition of CsBr, the salt slowly separated as orange-red crystals. Attempts to wash the solid with water resulted in darkening of the product, presumably as a result of decomposition of the rhodium carbonyl to 122 metal. When concentrated HBr was used as the wash solvent, no decomposi-tion was observed. Analysis Found: C, 1.63; Br, 49.60%. Calcd. for [Cs] 2[Rh(C0)Br 5]: C, 1.57; Br, 50.19%. The solid showed a carbonyl stretch at 2063 cm-* as 4 5 reported in the literature.' Aqueous HBr solutions of the salt exhibited an absorption maximum at X = 435 nm, e = 1050; and A = 550 nm, e = 119. max max 5 Cleare and G r i f f i t h have postulated that the anionic species 2_ [Rh(C0)Brg] is formed from rhodium(III) tribromide by a reaction which pro-ceeds via formato and formato-bromo carbonyl intermediates. An anionic Rh* carbonyl species presumably [Rh^CO^B^]" w a s identified by i r and absorption spectroscopy. The yellow solution, i n i t i a l l y obtained from the reaction of RhBr.j-2 H^ O and HC00H in HBr shows an absorption spectrum with A = 330 nm, e = 3060. Addition of (n-C^Hg^NBr to this solution precipitated the yellow dibromodicarbonylrhodate(I) salt, which exhibited carbonyl stretching freq-38 -1 -1 uencies at 1990 cm" and 2070 cm" (see Introduction, sec. 1.2.). The role of Rh* species in the catalytic substitution of Rh*** com-plexes has already been discussed (Chapter I I I ) , and the importance of such species has been demonstrated for the carbonylation of [Rh(C0)Brg] . The irep r o d u c i b i l i t y of the kinetic results for this system using different prep-arations of [Cs]2[Rh(C0)Brg] may be attributed either to the existence of 84 trace quantities of a Rh species,probably present as [Cs] [Rh(C0) 2Br 2], or. more l i k e l y , to the presence of small amounts of metal arising from the de-composition of the salt during its preparation. 85 Chapter V CATALYTIC ACTIVATION OF MOLECULAR HYDROGEN BY BROMORHODATE(III) COMPLEXES IN AQUEOUS HYDROBROMIC ACID SOLUTION 5.1. Introduction Rhodium(III) trichloride in aqueous 3 M HCl is known to homo-geneously activate hydrogen for the reduction of Fe1** i o n . 6 ' 7 The activa-tion parameters for a rate-determining step involving reaction of H2 with [RhCl 5(H 20)] 2 _ (AH t = 24.6 kcal/mole, 9.0 eu) 6 are very similar to those estimated for the i n i t i a l stage of carbonylation of the same species; a step that involves reaction of CO with [RhCl 5(H 20)] 2~ (AH T" = 25.9 kcal/mole, 4 2 A S ' = 13.2 eu). Corresponding reactions in DMA again yield similar para-meters for the hydrogenation (AH ^  = 17.3 kcal/mole, A S ^  = -10.8 eu)* 5 0 and carbonylation reactions ( A H T = 17.0 kcal/mole, A S T = -8.3 eu). 5.2. Catalytic H2 Reduction of Fe*** by [RhBr 4(H 20) 2]~ Excess Fe(C104)j-6 H20 in 0.5 M HBr was added to a solution of RhBr3-2 H20 in 0.5 M HBr which contained principally the 4:1 bromo complex (Amax = 5 3 8 ; e = * 7 3) ( F i9ure 2). Equilibrated solutions of [RhBr 4(H 20) 2]~ activate hydrogen under mild conditions {4 1 atmosphere H 2 > ^ 40°) for the homogeneous reduction of Fe***. Typical hydrogen uptake rate plots are shown in Figure 23. The plots are linear up to a stoichiometry corresponding to 1 H2 per 2Fe***. Simultaneous spectrophotometric measurements confirmed that the hydrogen taken up is consumed in the reduction of Fe*** to Fe** and that the concentration of Rh*** remains constant. The visible spectrum at the 86 Figure 23. Rate plots for Rh 1 1 1 catalyzed reduction of F e 1 1 1 by hy, (•) 40°, 706 mm hy, (A) 45°, 690 mm hy, (0) 55°, 645 mm h^, 0.028M Rh 1 1 1, 0.066M F e 1 1 1 , 0.5M HBr. 87 beginning of hydrogenation and near the end of hydrogenation was found to be very similar,exhibiting a peak in the region of 536 nm, e = 171. When al l the F e 1 1 1 had been reduced to Fe 1 1, reduction of Rh 1 1 1 to metallic rhodium occurred and was marked by an increase in the rate of H2 uptake caused by autocatalysis by the metal. In i t i a l rates are calculated from the linear slopes of H2 uptake plots. The present system corresponds very closely to the [RhCl^(HgO)] system studied earlier, and the kinetic data (Table XII) for the linear rates similarly show a f i r s t order dependence on Rh*11 (Figure 24) and H2 (Figure 25), and are independent of Fe*** concentration. The uptake plots may be fitted by a rate law such as: -d[H2]/dt = k ^ R h 1 1 1 ] ^ ] (5.1) Since the concentration of Rh*** and H2 remain constant while the ferric species is being reduced during any given set of conditions, the kinetics are overall pseudo-zero order. Values of k-j for a series of experiments are summarized in Table XII. The rate constant, k-j, determined from the slope of Figure 23 or Figure 24 is 0.17 M~* sec -*, which is in close agreement with the values in Table XI. Kinetic measurements over the temperature range 40-55° (Table XIII) yielded a resonable Arrhenius plot (Figure 26), and the activation parameters AH ^ = 22.7 ± 1.7 kcal/mole;.AS ^ = 10.6 ± 5.1 eu. 5.3. Discussion 5.3.1. Catalytic Reduction of Fe 1 1* by [RhBr 4(H 20) 2] -It has been found that in acid solution [RhBr / 1(H 90) 9] w i l l activate -88 Table XII Rhodium tribromide catalyzed reduction ferric ion Kinetic data at 40° in 0.5M HBr [Rh]x [Fe]x H2 [H2]x Rate x kl 102,M 102,M mm 104,M* 106,Msec_1 M"lsec~l 2.80 6.6 706 6.62 3.23 0.17 2.80 6.6 524 4.92 2.49 0.18 2.80 3.3 524 4.92 2.34 0.17 2.80 6.6 353 3.31 1.68 0.18 2.80 3.3 353 3.31 1.51 0.17 2.40 6.6 706 6.62 2.69 0.17 1 .68 6.6 706 6.62 1.90 0.17 *The concentration of Hg was calculated using solubility data 151 of Wiebe and Gaddy correcting for the effect of HBr. The latter correction (-3% for 0.5M HBr) was based on room temp-erature data averaged over a series of acids at this acidity. 89 90 91 Table XIII Rhodium tribromide catalyzed reduction of fer r i c ion Temperature dependence of k-| ([F e 1 1 1 ] = 0.066M; 0.5M HBr) Temp. °C [Rh]x 102,M H2 mm ..H2x 104,M* k l M _ 1sec _ 1 40 2.80 706 6.61 0.17 45 3.00 690 6.38 0.30 50 3.00 670 6.12 0.40 55 3.00 645 5.88 0.92 *The concentration of H? was calculated using solubility data of Wiebe and Gaddy131 correcting for the effect of HBr. The latter correction (-3% for 0.5M HBr) was based on room temperature data averaged over a series of acids at this .... 152 acidity. 92 93 molecular H2 for the homogeneous reduction of ferric ion. A common feature of the mechanisms of many homogeneously catalyzed hydrogenation reactions appears to be a rate determining heterolytic splitting of H2 by the catalytic species with formation of a metal hydride complex,which reduces the substrate in a subsequent fast s t e p . 6 ' 7 ' 1 5 3 The present kinetic data are consistent with such a mechanism: [RhBr 4(H 20) 2]~ + H 2 ?=L=± [HRhB^CH^]~ + H + Br" (5.2). k - l [HRhBr3(H20)2] + 2FeBr3 — — > [RhBr4(H20)2]"+HBr + 2FeBr2 (5.3) Direct evidence for an equilibrium such as shown in the f i r s t reaction, equation (5.2), has been obtained for a corresponding ruthenium(III) chloride 56 57 system, by isotopic exchange experiments using deuterium. ' The heterolytic splitting involves a net substitutional process (eg. replacement of bromide ligand by a hydride ion), without change in the formal oxidation state of the metal. Hence, the reactivity of a particular complex is thought to be governed by i t s substitution l a b i l i t y and the stab i l i t y of the hydride formed. The presence of a suitable base to stabilize the released proton can also promote activity. In the case of the Rh*** chloro systems,7 the ease of formation of the hydride complex for a series of complexes followed essentially the order expected for replacement of chloride ion with increasing negative charge;* 5 4 [RhClg] 3" > [Rh(H^O)Cl5]^" > [Rh(H 20) 2Cl 4]\ There seems a distinct possibility that for certain metal ion com-plexes, the net heterolytic hydrogen splitting may result via oxidative 94 addition of H2 to give a dihydride followed by a reductive elimination reac-155 tion (eg. for a divalent ion). M**X + H2 > [MIVH2X] > M HH + HX (5.4) However, such a process seems unlikely for the co-ordinatively saturated Rh 1 1 1 complexes. In view of the well-known interconvertability of rhodium (III) hydrides and rhodium(I),* 5 6 Rh I HH" * Rh1 + H+ (5.5) i t is quite possible that the rapid step with the ferric substrate (cf. equa-tion (5.3)) may involve Rh1 rather than Rh 1 1^", ie. equation (5.5) followed by Rh1 + 2 F e I H >Rh I I I + 2Fe*1 (5.6) The carbonylation studies (Chapter IV) show that Rh* species do readily reduce Fe***. No information is available on the nature of the fast reac-tions subsequent to the rate-determining interaction of H2 with Rh***. In the absence of ferric substrate, rhodium(III) bromide complexes are a l l rapidly reduced to the metal by H2 at 40° in HBr solution. The equilibrium represented by reaction (5.7) must.still exist, but the hydride complex is rapidly converted by some process to metallic rhodium (5.8) Decomposition of the hydride almost certainly Rh**1 + H2 . k l - Rh**IH" + H+ (5.7) k - l 95 I I I - *s Rli 'H * > Rh° (5.8) involves formation of Rh1 (equation (5.5)) oossibly followed by dispropor-t i o n a t e . 2 3 , 1 5 7 2RhT ^ Rh° + Rh 1 1 (5.9) 96 CHAPTER VI THE REACTION OF MOLECULAR OXYGEN WITH DICHLORODICARBONYLRHODATE(I)•• IN AQUEOUS ACID SOLUTION 6.1. Introduction 1 2 James and Rempel ' have reported on the kinetics of the carbon monoxide reduction of hydrochloric acid solutions of RhCl^*3 HgO to dichloro-dicarbonylrhodate(I) species, [Rh(C0) 2Cl 2]~. The present studies revealed that upon exposure to a i r , acid solutions of the anionic species, [Rh(C0) 2Cl 2]", were converted f i r s t to [Rh(C0)Clg]2~, and f i n a l l y to mix-tures of [Rh(H 20) 6_ nCl n]( n" 3^"with the evolution of C02. The kinetics of the reaction of [Rh(C0) 2Cl 2]~ with 0 2 have been studied in detail. Oxidation of [Rh(C0) 2Cl 2]" to [Rh(C0)Clg]2~ was found to exhibit autocatalysis similar to that observed during the carbonylation 1 2 of RhCl^*3 H20. ' Subsequent oxidation of co-ordinated carbon monoxide in 2_ [Rh(C0)Clg] appears to require decomposition of the complex by water or hydroxide to form a Rh* species, which in the presence of 0 2 is rapidly transformed to Rh1*1 species. 6.2. The Reaction of [Rh(C0) 2Cl 2]" in Aqueous HC1 Solution with 0 2 Yellow solutions of [Rh(C0) 2Cl 2]~ in 3M HC1 showed l i t t l e change in visible spectrum at R.T. U m a x - 326 nm, e = 3550) (Figure 27) when equilibrated for 2 days at 80° under 1 atmosphere of CO. Solutions under vacuum at R.T. similarly gave an absorption maximum at 326 nm, e = 3450. 97 Wavelength, nm Figure 27. Absorption spectrum of !Rh(;C0)2Cl2]" in 3M HCl. 98 Similar absorption peaks have been reported previously for [Rh(C0) 2Cl 2]~ 40 in aqueous and non-aqueous media. A solid was isolated by the addition of a methanolic solution of Ph^AsCl to the starting yellow solution. Ir spectra in the carbonyl region (V CQ = 2061, 1979 cm"1) observed for the 38 complex corresponded to those for [Ph 4As][Rh(C0) 2Cl 2]. On exposure to 0 2, the dicarbonyl solutions between 55° and 70° gradually changed from yellow to orange. After about 2500 sec, the spectra are similar to that shown in Figure 28 (X m a v = 394 nm, e = 400; x „ = 502 nm, e = 80); these slow max max changes were followed quantitatively (sec. 6.3.). Two gas absorption experi-ments in the presence of a sodalime tube are shown in Figure 29; there is no net measurable gas absorption or evolution within the f i r s t 2500 seconds of reaction, although the solution is orange at this stage and almost certainly contains a Rh*** carbonyl (see below). Mass spectral analysis of the gas mixture above the solution at the end of the 2500 second period indicated the presence of C02. An orange solid {v^ = 2104 cm-1) could be precipitated at this point by the addition of an aqueous solution of CsCl. The electronic (essentially the same as in Figure 28) and i r spectral data for the impure Cs salt are in close agreement with those reported for the [Rh(C0)Cl 5] 2" species. [RhCl3-3 H20 in LiCl/DMA absorbs III 3 1 mole of CO producing a Rh (CO) complex with absorption maxima at X-max = 395 nm, e = 540; A m a x = 520 nm, e = 78 ; salts of [Rh(C0)Cl 5] 2" exhibit a carbonyl band at ^ 2095 cm-1, approximately the same frequency as that 4 5 observed for the isolated compound. ' ] The isolation of a Rh*** carbonyl and the detection of C02 in the f i r s t stage of reaction suggest the stoichiometry given by equation (6.1), with the C0? not being absorbed by the sodalime. 99 10.0 Time x 10 , sec Figure 29. Uptake plots for the oxidation of [Rh(C0)2C12] in the presence of a sodalime tube, 70°, 560 mm 0 2, [Rh1]: (O) 0.060M; (A) 0.020M, 3M HC1. 101 2H+ + Rh^COjg + 0 2 ^Rh I H(CO) + C02 + H20 (6.1) After the i n i t i a l 2500 seconds, slow gas absorption was observed up to a total measured uptake corresponding closely to a 3/2:1 mole ratio of gas to rhodium as shown in Figure 29. At the 3/2:1 uptake stage, the colour of the solution had changed to red (X = 404 nm, e = 118; X = 512 nm, e = 119) (Figure 30), H i d A max 154 suggesting that a chlororhodate(III) species had been formed. The i r spectrum of an impure pink solid precipitated from solution as the Cs salt at this stage showed no carbonyl stretch. These observations indicate an overall stoichiometry given by equation (6.2), with the C02 now having been absorbed. 2H+ + RhI(C0)z +" 3/2 0 2 - Rh 1 1 1 + 2C02 + H20 (6.2) It seems certain that the sodalime eventually absorbs the C02, but relatively slowly. Surprisingly, there appears to be essentially no C02 absorption over the f i r s t 2500 sec, but the overall stoichiometry is consistent with equation (6.2), where a l l the C02 has been absorbed! Some similar kinetic and stoichiometric gas absorption studies were found quite successful in the presence of sodalime for some carbonylation reactions of rhodium(III) bromides which produce C02 (Chapter I I I ) , and other workers in 2 this laboratory have similarly encountered no d i f f i c u l t i e s . A probable explanation in these oxygenation studies is that the 0 2 present 'poisons' to some extent, the sodalime surface. The C02 possibly has to compete for an absorption site. The nature of the uptake plot of Figure 29 suggests that 102 Wavelength, nm Figure 30. Absorption spectrum of rhodium(III) chloro species in 3M HCl. 103 the CO2 absorption is an autocatalytic type process. Although the stoichiometric arguments are not very satisfactory, the reactions outlined in equation (6.1) and (6.2) undoubtedly occur, as shown by separate studies of the two processes (sec. 6.3., 6.4., 6.5.). Clearly, plots such as those shown in Figure 29 are unsuitable for kinetic analysis, since they are complicated by an apparently relatively slow C02 uptake by the sodalime, particularly during the stage when the Rh^^CO) is converted to chlororhodate(III) species. In the absence of a sodalime tube, there was a net overall evolu-tion for the reaction of 0 2 with Rh***(C0) corresponding to a (1/2:1) mole ratio of gas to rhodium. This is consistent with reaction (6.2) "minus" that of (6.1). Rh H I(C0) + 1/2 0 2 - R h 1 1 1 + C02 (6.3) Supporting the stoichiometrics of reactions 6.1, 6.2., and 6.3 is the fact that the reaction of 0 2 with [Rh 1(C0) 2C1 2]" to give Rh 1 1 1 studied in the absence of sodalime gives an overall evolution of 0.5 moles of gas per rhodium with evolution curves similar to those of the type shown in Figure 31, but preceded by the expected induction period for reaction 6.1 (see sec. 6.5.). Experiments reacting 3M HCl solutions of [Cs] 2[Rh(C0)Cl 5] (reflec-tance spectrum AM = V / = 398 nm, AM,„ = 502 nm; solution spectrum A „ = 393 nm, max max max e = 453; A = 500 nm, e = 60) with 0 2 in the presence of a sodalime tube f i n -a lly resulted in a total measurable uptake corresponding to a 1/2:1 mole ratio of 0 2 to rhodium, again consistent with reaction (6.3). In similar 104 Time x 10" , sec Figure 31. Kinetic plot for the loss of Rh I n(C0) (0.040M Rh 1 1 1, 560 mm 0., 3M HC1, 70°): (O), C02 evolved; (•], log [Rh i r i(C0)]. 105 experiments with no sodalime, mass spectral analysis at the end of the reaction confirmed the presence of C0£- The final red solution showed an absorption spectrum (A m a u = 403 nm, e = 104; A „ = 511 nm, E = 103) similar max max to the product (a Rh*** chloride) obtained from the prolonged reaction of [Rh(C0) 2Cl£j~ and 02- It is interesting to note, in this regard, that attempts by Thomas and Stanko* 5 8 to grow single crystals of [(CHgJ^N^Rh^COjClg] by solvent evaporation from acid solutions of the pentachlorocarbonylrhodate(III) compound yielded instead the salt, [(CHg)4N] [RhCI 4(H 20) 2]. 6.3. Kinetics of the Formation of Rh***(C0) According to Reaction (6.1) The kinetics of the f i r s t stage (eg. 6.1) were studied by uv and visible spectroscopy using rhodium concentrations in the range 0.0002 -0.0200M. The reaction was studied by following the absorbance increase in the electronic spectrum at known intervals at a fixed wavelength, 393 nm 2_ (xm*v f° r [Rh(C0)Clr] ) on bubbling 0 o through the solution. The spectra, max o c. as a function of time, exhibited isosbestic points at 312 nm and 381 nm. The increase in O.D. shows autocatalytic behaviour (eg. Figure 32 at [Rh] = 0.0035M) and no simple dependence on rhodium concentration was observed. This is readily evident from the fact that at higher Rh concentrations i t becomes d i f f i c u l t to detect the autocatalysis (eg. Figure 33 [Rh] = 0.020M). Oxidation of 0.020M [Rh(C0) 2Cl 2]~ (Figure 33) was complete in ca_. 30 minutes, while a 0.002M solution of the complex required 2 hours to go to completion. A first-order plot of optical density versus time in Figure 32 shows an i n i t i a l approximately linear portion followed by a divergence from linearity, the rate being too rapid for simple f i r s t order. Based on our findings for 106 Time x 10" , sec Figure 32. Kinetic plot for the loss of [RhCCO^Clgl" at 393 nm (0.00348M Rh1, 560 mm 0^ 3M HC1, 70°): (O), O.D. plot (1 cm optical c e l l ) ; (•), log [A^-A]. 107 0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 _3 Time x 10 , sec Figure 33. Kinetic plot for the loss of [Rh(C0)2CT2]~ at 393 nm (0.020M Rh1, 560 mm 0 2, 3M HCl, 70°); (O), O.D. plot (1 mm optical c e l l ) ; (•), log[Rh I H(C0)] - logdRh^CO^Clg]"}. 108 autocata,lysis due to interaction of Rh and Rh complexes (Chapter I I I ) , such autocatalysis could result here in a similar manner. The correspon-ding rate equation would be: -dCRhHcOJ^/dt = k-j'LWccOjg] + kgCRh^COl^LRh111^)] (6.4) where the f i r s t term is concerned with the rate of production of the auto-catalytic species, [Rh***(C0)], and the second term is the rate of auto-catalytic reaction. The dependence of the reaction on 0 2 concentration could not be conveniently measured by spectrophotometric methods. It was assumed, however, that k-|1 contained a f i r s t order dependence in 0^, since the f i r s t term of the rate law is concerned with oxidation of [Rh*(C0)2C12]~. The kinetics of the i n i t i a l reaction were determined from the i n i t i a l part of the plots at various Rh concentrations between 55°-70°. The limited data at 70° (Table XIV) together with other data at 55° with lower rhodium concentrations (Table XVII, sec. 6,4.) show a f i r s t order dependence in [Rh]. Values of k-|1 and k-| are given in Table XIV. Measurements over the temperature range 55-70° yield a good Arrhenius plot, Figure 34, and the activation parameters AH-^ = 9.3 ± 0.8 kcal/mole and AS-^ = -33.2 ± 2.4 eu. 6.4. The Autocatalytic Reaction The 0 2 independence of k 2 was demonstrated indirectly by bubbling N 2 through 1:1 mixtures of [Rh 1(C0) 2C1 2]~ and freshly formed [Rh 1 1 1(C0)C1 5] 2~ Subsequent 0 o treatment gives faster changes in absorbance (measured at 393 Table XIV Oxidation of [Rh(CO) 2Cl 2]~ Summary of kinetic data in 3M HC1 (O.D. measurements at 393 mm) Temp. °C [Rh]x103, M Og (mm) [0 2]xl0 4, M* k] ' x 104, sec"* k l M~lsec _l 55 3.00 660 6.13 1.54 0.25 55 3.00 660 6.13 1.62 0.26 60 3.00 636 5.58 1.75 0.31 60 3.00 636 5.58 1.73 0.31 65 3.00 600 5.58 2.19 0.43 65 3.00 600 5.04 2.10 0.42 65 3.00 600 5.04 2.15 0.43 70 2.00 560 4.53 1.96 0.43 70 2.48 560 4.53 2.10 0.47 70 3.00 560 4.53 2.17 0.48 70 3.48 560 4.53 2.19 0.48 *[02] was calculated using solubility data of Seidell 159 correcting for the effect of HC1. The correction (-22% for 3M HC1) was estimated from data at 15° and 25° for various HC1 concentrations. no I l l nm, Figure 35) than was measured previously (Figure 33), indicating the presence of a reactive intermediate. This suggests that the 0 2 reacts faster with a Rh*'""Rh*** intermediate (or a decomposition product - see sec. 6.6.2.), than the rate of formation of the intermediate i t s e l f (ie. 0 2 is not involved in the rate determining step). Analysis of the k 2 term may be carried out in a number of ways, as discussed previously for the carbonylation of [RhBr 4(H 20) 2] ~ (Chapter III). (a) Maximum rates for the formation of Rh n i(C0) can be measured from the slopes of optical density - time plots (Figure 32). The relation shown in equation (6.5) was used to compute [Rh J I I(C0)] [Rh i n(C0)] + [Rh !(C0) 2] = [ R h I ( C 0 ) 2 ] i n i t i a l (6.5) It can be readily shown that a rate proportional to the product [Rh^Rh 1 1 1] becomes a maximum when [Rh1] = [Rh 1 1 1] = l / Z E R h 1 ] ^ ^ (6.6) Maximum rate and i n i t i a l rate data were used in a series of experiments (Table XV) using different i n i t i a l rhodium concentrations at constant 0 2 pressure,at conditions where the autocatalytic behaviour was observable; this depends on the rhodium concentration, sec. 6.3. The following relationships hold: Maximum rate = ^'[Rh 1] + k 2[Rh I][Rh I H] (6.7) 112 Time x IO"3, sec. Figure 35. Kinetic plot for the loss of [Rh 1(C0)C1 3] 2~ at 393 nm (formed by bubbling N2 through 1:1 mixtures of [Rh(C0) 2Cl 2] and [Rh(C0)Cl 5] 2" see sec. 6.6.2.). (0.020M Rh, 560 mm 02» 70°, 3M HC1) (O) O.D. plot (1 mm optical c e l l ) ; (•) log [A^-A]. 113 I n i t i a l rate = ^ ' [ R h I ] T o t a l (6.8) A plot, Figure 36, [Max. rate] - [Init. rate] against the product [Rh1] x III 2 [Rh ] at the point of maximum rate gave a straight line with a zero inter-cept, in agreement with relationships (6.6) to (6.8). A value for k 2 of 0.107M~*sec~* at 70° was calculated from the slope of Figure 36. (b) The ratio (R) of the maximum to i n i t i a l rate is given by equation (6.9), R = k 2[Rh I I I]/2k 1' + 1/2 (6.9) where [Rh***] is the concentration of the autocatalytic species at the point of maximum rate. Values for k 2 calculated by this method are summarized in Table XVI. They agree well with values determined by method (a), (c) The following expression (6.10) can be obtained by integration of the rate law (6.4). log(k 1' + k2[Rh***]) - log[Rh*] = (k ]' + k 2[Rh] T t ^ ) t / 2 . 3 + constant (6.10) where k-|1 = k-j [0 2] and t is the reaction time. By using appropriate values of k-|1 and k2, determined by the previous methods (a) and (b), plots of 1og(k-|* + k2[Rh***]) - 1og[Rh*] versus t should give a straight line of slope (k^1 + k 2[Rh]y 0j. ai )/2.3. The concentration of each of the species [Rh*(C0) 2Cl 2]" and [Rh 1 1 1(C0)C1 5] 2 _ is given by the following expressions: [ R h I ] . f c S _ ; m n h [ R h , ] i n i t i a i . fcM ( 6 J 1 ) 114 [Rh^xtRh 1 1 1] x ]0 6, M2 .. , Figure 36. Dependence of [maximum rate] - [ i n i t i a l p£}te]/2 on [Rh^xCRh 1 1 1], 70°, 560 mm 0 2, 3M HC], Table XV Oxidation of [Rh(C0) 2Cl 2]~ I n i t i a l and maximum rates in 3M HC1 Temp. [RhJxlO3, Init. rate Max. rate [ M a x ' rate]-[In1t. rate] °C M xlO 7, M sec"1 xlO 7, M sec"1 7 m s e c?-, 55 3.00 4.62 4.28 1.97 0.088 55 3.00 4.86 4.41 1 .98 0.088 60 3.00 5.25 4.67 2.05 0.091 60 3.00 5.19 4.72 2.12 0.094 65 3.00 6.57 5.57 2.29 0.102 65 3.00 6.30 5.34' 2.19 0.097 65 3.00 6.45 5.36 2.14 0.095 70 2.00 5.84 3.07 1.11 0.111 70 2.48 5.20 4.39 1 .79 0.118 70 3.00 6.51 5.63 2.38 0.106 70 3.48 7.48 6.98 3.24 0.108 70 20.0 115 0.115b 0.076C 0.112d Calculated using method (a). Calculated using equation (6.12). c Calculated using method (c). Calculated using method (d). 116 Table XVI Oxidation of [Rh(CO)2C12]" Ratio of maximum: i n i t i a l rates in 3M HC1. Temp. °C [Rh]xl0 3, M Max. rate ,„ Init. rate x l u V x 10 4, sec"1 M-l^sec-1 55 3.00 9.16 1.54 0.085 55 3.00 9.07 1.62 0.088 60 3.00 8.90 1 .75 0.091 60 3.00 9.13 1.73 0.095 65 3.00 8.49 2.19 0.102 65 3.00 8.48 2.10 0.097 65 3.00 8.32 2.15 0.095 70 2.00 5.27 1.96 0.106 70 2.48 8,44 2.10 0.116 70 3.00 8.64 2.17 0.105 70 3.48 9.34 2.19 0.109 a Calculated using method (b). 117 where e-| =150 and = 450 are the molar extinction coefficients respec-tively for the Rh* and Rh*** complexes at 393 nm. A typical plot (computer drawn) is shown in Figure 37. Excellent linear plots of the l e f t hand side of equation (6.10) versus time up to ca_. 90% completion were obtained. A value of k 2 determined by this method for 0.003M Rh is 0.076M~*sec~*. However, the rate constants are only indicative of the order of magnitude. k-|' and k 2 should be varied, producing a family of curves until a consistent set of rate constants is obtained. (d) Using an appropriate [Rh*] concentration ( 0.02M) so that k 2 [Rh*][Rh***] >> k-|'[Rh*] at the maximum rate, the rate law may be simpli-fied to give: -d[Rh*]/dt = k2[Rh*][Rh***] } (6.12) which can be integrated directly to give the expression log[Rh***] - log[Rh*] = k 2 [ R h ] j o t a l t/2.3 + constant (6.13) [Rh]y0£ai represents the total rhodium concentration, [Rh*] is the concen-tration of [Rh*(C0)2C12]~ reactant, [Rh***] is the concentration of [Rh***(C0)C1g]2~ product, and t is the reaction time. [Rh***] and [Rh*] are calculated using (6.11). A plot of equation (6.13), (Figure 33) shows a good f i t for most of the data. k 2 calculated by this method is 0.112M~*sec~* in good agreement with values calculated by previous methods. It should be noted that i f the i n i t i a l Rh concentration i s very low (ca_. 10~4M), the f i r s t term of the rate law should become dominant and 118 Time x 10" , sec 119 the rate expression is reduced to a simple f i r s t order dependence in rhodium. . -d[Rh!]/dt = ^ '[Rh 1] (6.14) The rate of formation of product [Rh(C0)Clg] by a direct 0 2 reaction is now faster than the rate of production via the autocatalytic reaction. Log plots for the optical density data at 326 nm (*max of [Rh(C0) 2C1 2]~), Fig; 38, were found to be linear over ^ 40% of the reaction. Values of k-j determined from these plots (Table XVII) are somewhat larger than the values calculated at 393 nm; however the agreement is reasonable, since different analytical methods and very much lower rhodium concentrations were used. Kinetic measurements over the temperature range 55°-70° with analysis by methods (a) and (b) yielded a good Arrhenius plot, Figure 39, for k 2 and the activation parameters, AH^ = 2.6 ± 0.6 kcal/mole and AS 2^ = -56 ± 2 eu. 6.5. Decomposition of [Rh(C0)Cl 5]^" in 3M HC1 When [Rh 1(C0) 2C1 2]" is reacted under 0 2 in the absence of soda-lime, gas evolution is observed corresponding to 1/2:1 mole ratio of C02: Rh (see sec. 6.1.). At lower [Rh], ca_. 0.020M, an induction period was sometimes observed prior to C02 evolution, which was of the same duration as the time required for oxidation of [Rh 1(C0) 2C1 2]~ to [Rh 1 1 1(C0)C1 5] 2~ (ca. 1500 seconds for [Rh] = 0.020M). Since the decomposition reaction of Rh J I I(C0) is very much slower than the autocatalytic oxidation of 120 -0.2 -0.4 Log [A -A] -0.6 -0.8 o o -*.o o -1 .2 1 1 1 1 0 1.0 2.0 3.0 4.0 5.0 _3 Time x 10 , sec. Fig. 38. Kinetic plot for the loss of [Rh I(C0) 2Cl 2]" at 326 nm (1 cm optical c e l l ) (0.0002M Rh, 660 mm 0 2, 55°, 3M HCl). 121 Table XVII Summary of kinetic data in 3M HCl at 55° (O.D. measurements at 326 nm) ([0 2] = 6.13 x 10"4 M) RhxlO4, M k1 'xlO 4, -1 sec M"1 k l sec 2.00 2.34 0 .38 3.00 2.58 0, .42 4.00 2.74 0. .45 -1 122 -0.95 h -0.97 \--0.99 r C M CD o -1.01 -1.03 r -1.05 h -1.07 2.90 2.95 3.00 (1/T) x IO 3, K -1 3.05 Figure 39. Arrhenius plot for oxidation of [Rh^CO^Ciy" by [Rh I H(C0)C1 5 ] 2 " in the presence of 0^> 3M HCl, • 123 [RhA(CO)2C12^1"» the induction period was not detectable at higher concen-trations of [Rh]. Evolution during the second stage, i.e. the production of Rh*** from Rh***(C0) therefore corresponds to 1 mole of C02 evolved per mole of [Rh** I(C0)Cl 5] 2". At constant 0 2 pressure, the second stage of reaction exhibited pseudo-first order kinetic behaviour as shown in Figure 31. 2d[C02]/dt = -d[Rh A i lC0]/dt = k 3'[Rh l i lC0] (6,15) The formation of Rh*** chloride products obeys the rate law given in equation (6.15). Evolution of C02 was extremely slow so that values of k^1 were usually estimated from i n i t i a l slopes (Table XVIII). Similar results were obtained when k3' was determined from the slopes of log plots. The f i r s t order dependence shown by the log plots is further substantiated by i n i t i a l slope data for varying Rh concentration (Figure 40). The rate constant k 3 ( was found to be essentially independent of 0 2 pressure between 430 - 630 mm, so at these conditions the rate becomes: -d[Rh***C0]/dt = k3[Rh***C0] (6.16) -4 where k 3 is a true f i r s t order rate constant. A k 3 value of 1.04 x 10 sec - 1 was estimated from the slope of Figure 40 in good agreement with the value obtained from the i n i t i a l rate data (Table XVIII). At low 0 2 pressures, i n i t i a l rate data indicated that the rate steadily decreased with decreasing 0 2 and approached a f i r s t order dependence in 0 2, Figure 41. At low 0 2 pressure and high [Rh] concentration, Rh metal was produced. 124 Table XVIII Formation of Rh 1 1 1 chloro species from Rh H I(C0) Summary of kinetic data at 70° in 3M HCl [Rh]xl0 2, M °2 mm 0 2xl0 4, M* In i t i a l ratexlO , M-l sec~l k 3 x l 0 4 , sec-1 k4 M"l sec"l 2.00 560 4.53 1.12 1.12 3.00 560 4.53 1.65 1.10 4.00 560 4.53 2.84 1 .42 5.00 560 4.53 3.28 1 .30 6.00 560 4.53 3/46 1 .06 7.00 560 4.53 4.07 1.06 7.00 560 4.53 4.25 1 .20 3.00 628 5.07 1 .72 1.15 3.00 486 3.92 1 .68 1.12 3.00 456 3.68 1 .57 1 .05 3.00 428 3.46 1.56 1.04 3.00 392 3.16 1 .34 3.00 344 2.78 1.21 0.29 3.00 276 2.24 0.91 0 .27 7.00 424 3.43 1.65a 0.27 a Metal is produced. *[02] was calculated using solubility data of S e i d e l l ^ correcting for the effect of HCl. The correction (-22% for 3M HCl) was estimated from data at 15° and 25° for various HCl concentrations. 125 127 Experiments were carried out in which the [Cl~] was varied while the acidity was maintained constant. This was accomplished by decreasing [HC1] and adding sufficient HCIO^ to provide a 3M acid media. The value of k 3 was found to be essentially independent of [Cl~] (Table XIX) (Figure 42). The effect of acid on the reaction was studied by varying HCl/LiCl while maintaining a total [Cl~] of 3M. k 3 showed an inverse dependence on acid concentration (Table XX). A plot of k3 vs. [H +]~* is given in Figure 43. The variation of [Cl~] and [H +] was studied using a new stock solution of [Rh(C0)2Clg]~ which may explain the faster i n i t i a l rates (a factor of ^2) which were obtained when compared to the rates in Table XVIII. Kinetic measurements over the temperature range 55-70° (Table XXI) yielded a good Arrhenius plot, Figure 44, for k 3 and the activation para-meters A H ^ = 8.4 ± 1.0 kcal/mole and A S ^ = -51.9 ± 3.1 eu. 6.6. Discussion 6.6.1. Formation of Rh^fCO) According to Reaction (6.1) The f i r s t term of the rate law (equation 6.4) is concerned with the direct production of the autocatalytic species [Rh(C0)Cl5] using 0 2, and is f i r s t order in both [Rh 1(C0) 2C1 2]~ and 0,,. A plausible rate deter-mining step is thought to involve oxidative addition of 0 2 to produce an III 2 octahedral intermediate [Rh (C03)(C0)C13] with chelating carbonate which decomposes in acid media evolving CO,,:70 Rh^COJg + 0 2 k-! + [Rh I I I(C0 3)(C0)] (6.17) [Rh H I(C0j(C0)] ^ - Rh m(C0) + C09 + Ho0 (6.18) J fast L 1 128 Table XIX Effect of [Cl -] on the rate of reaction of Rh I H(C0) at 70° ([Rh] = 0.040M; [H +] = 3.0M) [CT] M Ini t i a l M ratexlO 6, 3 sec"l k 3 x l 0 4 , sec"l 3.0 3, .61 1.81 2.4 3. .70 1.85 1 .8 3. ,43 1.71 1.2 3. ,86 1.93 a Fresh stock solution of [Rh(C0) 9Cl 9]". Table XX Effect of [H +] on the rate of reaction of Rh H I(C0) at 70° ([Rh] = 0.040M; [Cl"] = 3.0M) [H +] M In i t i a l ratexl0 6, a M sec~l k 3 x l 0 4 , sec _l 3.0 3.61 1.80 2.4 3.74 1.87 1.8 4.07 2.03 1.5 5.25 2.62 1.2 6.11 3:05 a Fresh stock solution of [Rh(C0) ?Cl 9] 129 4.0 — . o 3.0 2.0 1 .0 0 1 1 1 1 1 i 0 0.5 1.0 1.5 2.0 2.5 3.0 [Cl -] , M Figure 42. Effect of [Cl~] on formation of Rh 1 1 1 chloro species from [Rh(C0)Cl 5] 2-, 70°, 0.040M [Rh], 560 mm 0 2, 3M [H +]. 130 3.0 2.5 2.0 — / o 1 .5 1.0 0.5 n — 1 1 1 1 1 0 0,2 0.4 0,6 [H+]"1, M"1 0,8 1,0 Figure 43. Plot of k 3 vs. [H*]' 1, 70°, 0.040M [Rh], 560 mm 0 2, 3M [ C l - ] . 131 Table XXI Formation of Rh 1 1 1 chloro species from Rh^^CO) Temperature dependence of k3 in 3M HCl ([Rh] = 0.070M) Temp. k3xicr, °C sec-"1 70 1.20 65 0.95 60 0.82 55 0.71 132 Figure 44. Arrhenius plot for formation of Rh chloro species from Rh H I(C0), 3M HCl. 133 Oxidation could proceed through i n i t i a l formation of an oxygen complex, followed by an insertion type reaction (Scheme I). Scheme I k ?****0 Rh I(C0) 2 + 0 ? ] > [(CO)Rh^ ] '2 ' "2 :o [(C0)Rh***v. / ] / (6.19) S / 0 \ [(CO)Rh ^ + T T T \ >C = 0] + 2H —>Rh 1 U(C0) + C02 + H20 (6.20) An alternate reaction path for the present system could involve oxidation of a metal centre (but not via oxidative addition to a carbonate) as shown in the following scheme. Scheme II k, T 1^ TTT Rh^ C O ) . + 1/2 Oo •—> Rh (C0) o + H90 (6.21) c 2H+ L R h H I ( C 0 ) 2 + H20 > Rh*(C0) + C02 + 2H+ (6.22) Rh*(C0) + 1/2 0 2 — — > Rh H I(C0) + H20 (6.23) Here, oxidation of the Rh* in reactions 6.21 and 6.23 could involve direct electron transfer to 0 2 in an outersphere mechanism (eg. to give superoxide or peroxide).* 6 3 The former mechanism (Scheme I) seems very probable since square planar d complexes have been reported to undergo atom transfer reactions in which both the oxidant and reductant are co-ordinated to the metal. 7* 134 The oxygen adduct would be expected to rearrange to a metal carbonate directly or via formation of a perester-1ike intermediate. The reaction of C02 with (PPh 3) 2Pt0 2 involves prior formation of a peroxycarbonate complex*60'*6* (see Chapter VII). P h3 P> Ph3P 0 ptCT | + co, o Ph3P. Ph3p p t C = o PPh, Pt(C0 3)(PPh 3) 2 + OPPh (6.24) Several platinum metal complexes containing bidentate carbonate d b) 162 may be prepare y adding 0 2 to M(C0) complexes 6 9' 7 4' 8 3 or by adding CO to M02 complexes. I V C Oxidation of co-ordinated CO in OsCl(C0)N0(PPh3)2 pro duces 6 8 the characterized 0s(C0 3)ClN0(PPh 3) 2 complex which can react with 1 mole df HC1 to form 0s(0H)Cl 2N0(PPh 3) 2 and further HC1 to give OsCl 3N0(PPh 3) 2. The existence of a monocarbonylrhodium(I) complex has been speculated upon at low pressures of CO during the carbonylation of Rh*** 2 complexes in 3M HC1 to [Rh(C0) 2Cl 2] . Rapid co-ordination by one mole of CO is assumed in order to stabilize Rh*, but this is- thought to be followed by slow addition of a second mole of CO. A Rh* monocarbonyl-chloro species is formed when oxygenated solutions of [Rh(C0) 2Cl 2]~ are refluxed in DMA, but this is by a process probably involving solvent decarbonylation 135 (Chapter VII). An impure yellow solid was isolated as the Ph^As salt. The compound in CH2C12 had a single CO stretching mode at 1979 cm-1 in the region expected for Rh(I) carbonyls (eg. Rh(C0)Cl(PPh 3) 2 shows a CO band 1 6 4 at 1978 cm-1 in CgHg. It is of significance, however, that oxygenation of [Rh(C0) 2Cl] 2 in LiCl/DMA produces one mole of Li^CO^ per mole of rhodium (Chapter VII). Furthermore, a Rh*** carbonato complex has been isolated by treatment of the [Rh(C0) 2Cl] 2 dimer with t-butylhydroperoxide (Chapter VII). The Arrhenius parameters obtained for k-| (A H^ = 9.3 kcal/mole, A S ^ = -33.2 eu) are also similar to those reported for oxidative addition reactions involv-ing Ir* and o 2 5 4 ' 1 6 5 ' 1 6 6 ( A H ^ = 8.5 to 10.8 kcal/mole, A S ^ = -32 to -42 eu). These findings provide strong evidence in support of the oxidative addition mechanism, Scheme I. 6.6.2. The Autocatalytic Reaction The overall autocatalytic reaction may be represented by: + T TTT ®9 TTT 2H + Rh i(C0) 2 + Rli ^ C O ) £ 2Rh 1 X(C0) + C02 + H 2 0 (6.25) When 1:1 mixtures of [Rh(C0)Cl 5] 2~ and [Rh(C0) 2Cl 2]~ are reacted together under N 2, the subsequent rate of oxidation of the resulting species by 0 2 is accelerated (Figure 35) compared to the "normal" rate measured without the prior Rh***-Rh* reaction. This suggests that the rate determining step in the Rh***-Rh*-02 reaction is independent of 02-136 This step then involves a bimolecular reaction between [Rh1(CO)2C12]~ III 2-and [Rh (CO)Clg ] and is thought to involve formation of a mixed valence bridged transition state [Rh1 Cl Rh 1 1 1] analogous to that described for the Rh1 catalysed reduction of Rh1** by carbon monoxide.2 Rh* species have also been postulated as the active catalyst in the nucleophilie substitution reactions of some pyridine, ' ' and 22 III chloro complexes of Rh . The mechanism involves chloro-bridged dimers derived from a rhodium(III) and rhodium(I) complex. Thus [H20Rh(py)4 -- 12 Cl - Rh Cl 40H 2] is believed to occur in the ethanol catalyzed conversion of K 2[Rh(H 20)Cl 5] to trans -[Rh(py)^Cl 2]C1, and an analogous species is postulated in the conversion of trans -[Rh(en) 2Cl 2] + into trans -[Rh(en)2-(N2H^)]C12. The bis(dimethylglyoxime) complex of rhodium(III) is reduced in I 16 aqueous alkaline ethanol to Rh by a similar autocatalytic process. The corresponding scheme in the present system i s : H20 + [Rh*(C0) 2Cl 2]" + [ R h I I I ( C 0 ) C l 5 ] 2 - **. [(H 20)Cl 2(C0) 2Rh*-Cl-Rh 1 1 1 (C0)C1 4] 3' I (6.26) I [(H 20)Cl 2(C0) 2Rh*** - C l 4Rh*(C0)Cl 4] 3" (6.27) II II ^[Rh***(C0) 2Cl 3(H 20)] + Rh*(C0) (6.28) III HI - [Rh*(C0)Cl J 2 " + C0o + 2H+ l 3 r + C02 + 2H (6.29) IV 137 2H+ + IV + 1/2 0 2 — • [ R h m ( C O ) C l 5 ] 2 " + H20 (6.30) The planar [Rh(C0) 2Cl 2]~ complex forms the bridged intermediate (through i t s axial position) with the substitution-inert Rh***(C0) complex and also co-ordinates a sixth ligand (probably H20) giving I. A two electron transfer then occurs with subsequent breaking of the labile Rh* ligand bond to give the Rh 1 1 1 product, III, and a Rh1 monocarbonyl. Compound III contains both co-^  ordinated carbonyl groups and H20 and would undergo spontaneous decomposition to the Rh1 product;IV (see Chapter I I I ) ; IV must then probably react with oxygen to give further catalyst (product), Rh***(C0). With this mechanistic scheme, reaction of [Rh 1(C0) 2C1 2]~ with [Rh 1 1 1(C0)C1 5] 2" in the absence of I 2-0 2 should give rise to [Rh (C0)C13] product, and at the same rate as the III 2-formation of [Rh (CO)Clg] measured in the presence of 0,,. 111 2-When a mixture containing a slight excess of [Rh (CO)Clg] and [Rh*(C0)2C12]~ was reacted under N 2 >a slow decrease in absorbance at 393 nm ( A m a x for [Rh(C0)Clg] ) was observed. At the completion of this reaction (equations 6.26-6.29), i t was assumed that the only species in solution were T ?- T T T ? [Rh (C0)C13] and [Rh 1 1 1(C0)C1 5r (the latter being given by the amount [Rh***(C0)Cl 5] 2" exceeded [Rh*(C0) 2C1 2j" i n i t i a l l y ) . The estimated molar extinction coefficient for [Rh 1(C0)C1 3] 2 _ is 165 at X 393 nm. Using this extinction coefficent, k 2 can be determined by adapting the integrated form of the rate law (equation 6.14), which is valid when the autocatalytic reaction predominates for formation of Rh***(C0), taking into account the fact that the product wi l l be [Rh*(C0)C13]2": 138 log[Rh I H] - log[Rh*] = kg { [Rh 1 1 1] - [ V ] } i n i t i a l t/2.3 + constant (6.31) where {[Rh A i l] - [Rh1]}. n i t i a l i s t h e d l ' f f e r e n c e i n the concentration of the two anions [Rh***(C0)C1g]2~ and [Rh*(C0)gClg ]~prior to reaction under Ng. The value of kg for two experiments at 70° calculated from plots such as Figure 45 gave an average result of 0.112 ± 0.004 M~*sec~\ which is in close agreement with values obtained by other methods for data measured under 0g atmospheres (Table XV). The activation parameters determined for kg, AH^g = 2.6 kcal/mole. AS^ = -56 eu may be compared to those reported for other redox reactions involving bridged intermediates,* 6'* 4 0'* 4* although the enthalpy of activation is somewhat smaller (by ^ 5 kcal/mole) and the entropy of activation more negative (by ^ 20 eu). A negative entropy seems reasonable for interaction 142 between two negatively charged species. It is of interest to note that Stanko and coworkers 7 0 have also determined an activation energy of 2.5 kcal/mole for the 0g oxidation of [Rh(C0)gClg]" in 0.66M HC1. In contrast to the present studies which demonstrated an autocatalytic process, these workers reported that the reaction was simply f i r s t order in rhodium (no ex-perimental details of the work were published). The pseudo-first order rate constant for a constant concentration of 0g (1 atm.) was 1.94 x 10~4 sec"* at 25°. I 2-Oxidation of [Rh (00)01^] by 0g gives a pseudo-first-order rate -3 -1 constant of 2.25 x 10 sec at 70° (Figure 35), a factor of 10 greater than the value measured for the oxidative addition of 0g to [Rh*(CO)gClg]~ (sec. 6.4.) (k^ = 2.15 x 10"4 sec - 1 at 70°). A greater degree of Rh-CO u-139 140 39 ? back bonding in [Rh(C0) 2Cl 2] compared to that found in [Rh(C0)Cl3] (since two CO ligands are involved) will l i k e l y reduce the electron density at the metal centre and thus decrease the tendency for the metal to undergo 1 oxidative addition. A similar electronic effect has been suggested to explain the a b i l i t y of RhCl(PPh 3) 3 to oxidatively add H2, while RhCl(C0)(PPh3) 2 does not undergo oxidative addition with h^. 1 6 9 6.6.3. Decomposition of [Rh 1 1 1(C0)C1 5J 2~ III ? The decomposition of [Rh (C0)C153 is independent of 0 2, at least at the higher pressures studied, and must involve water to i n i t i a l l y reduce [ R h H I ( C 0 ) C l 5 ] 2 " to a Rh1 species. Rh1 in the presence of 0 2 is then rapidly oxidized to Rh 1 1 1: Rh H I(C0) _ J ^ ! - Rh1 + C09 + 2H+ (6.32) k '2 3 2H+ + Rh1 + 1/2 0 2 ± — - R h H I + H20 (6.33) The reduction of [Rh 1 1 1(C0)C1 5] 2~ by water (equation 6.32) was verified independently by visible spectroscopy (followed at 393 nm}Am,v for [Rh(C0)Cl 5] 2 -). I n i t i a l rates for the decomposition of [Cs] 2[Rh(C0)Cl 5] under N 2 in 3M HCl at 70° gave a k 3 value of 0.59 x 1:0~4 sec"1 in agreement with the values obtained from evolution experiments (Table XVIII). The inverse acid dependence for k3» Figure 37, indicates that hydroxide rather than water may be mainly involved in the decomposition of [Rh(C0)Clg] . The reduction process may be written via a carboxylate 141 intermediate as shown: Rh I H(C0) 2t! - [Rh(C09H)] Rh1 + CO- + H+ (6.34) k K3 The absence of any Cl dependence for k^  rules out a pre-equilibrium reaction such as: H20 + [ R h ( C 0 ) C l 5 ] 2 " ^ = = * [Rh(C0)Cl 4(H 20)] _ + Cl" (6.35) A [Rh(C0)Cl 4(0H)] 2~ + H+ (6.36) The hydroxide attack in reaction 6.34 may or may not involve prior co-ordination to the metal. A direct SN2 attack to yield a [Rh(C0)Cl 4(0H)] 2~ intermediate, or direct attack at the co-ordinated CO both seem feasible. It should be noted however that there does not appear to be a direct dependence on [OH-]; the plot of k 3 vs. [H +]~* also shows an acid indepen-dent path. The possibility that added OH" might give rise to different species (eg. hydroxide bridged dimers) of different activity was not investi-gated,al though this is considered unlikely at the relatively high acidities used. An inverse acid dependence for the decomposition of Rh***(C0) in DMA by water suggested the involvement of hydroxide. Forster has also synthesized Rh1 dicarbonyl halides by the reduction of the corresponding Rh 1 1 1 monocarbonyl 4 using carbon monoxide in aqueous media. Again, the presence of water seems necessary to reduce Rh***(C0) to Rh* before further co-ordination of CO. 142 The activation parameters A H | = 8.4 kcal/mole and A S | = -51.9 eu were determined from Figure 44. The complex inverse acid dependence of strongly indicates the involvement of hydroxide (or Rh hydroxy species) in the rate-determining step (6.32) (discussed above). No adjustment was made to account for the acid dependence of k^ . At low 0 2 pressures, the oxidation of Rh* may become involved in the rate-determining process. The precipitation of Rh metal under conditions of high i n i t i a l [Rh] (0.070M) and moderate p'02 (424 mm) strongly suggests a build-up in concentration of Rh*. It is known that Rh* chloro species in the absence of stabilizing ligands tend to disproportionate to metal. 7 When 0.030M Rh was reduced, no metal was observed under the same conditions of low pressure,and metal was never observed at pressures ^ 600 mm. This perhaps indicates that the rates of reactions (6.32) and (6.33) are of com-parable magnitude and that the system is more accurately represented by two consecutive f i r s t order processes. On the gas evolution apparatus, the observed results would then be a combination of the evolution (1 mole of CO2) and absorption (1/2 mole 0 2) processes. At both higher and lower oxygen pressures, the gas evolution data, however, gave reasonable f i r s t order plots. There is also the distinct possibility that bridged Rh*-Rh*** species would be involved at lower Or, pressures; these would also be susceptible to 0^ oxidation. Further investigation is required to give a more detailed under-standing of the 0 2 dependence. The apparent value of k^  determined from the slope of Figure 41 is 0.30 M~* sec"* in good agreement with the values obtained from the i n i t i a l rate data (Table XVIII), although this cannot simply represent the process of equation (6.33). It is again possible to postulate that oxidation of Rh* (with no 143 carbonyls) (reaction 6.33) occurs via a Rh1 - 0 2 complex. Though such a species has not been detected in this system, Gil lard has recently noted the inhibitory effect of 0o in Rh1 catalyzed nucleophilie substitutions III 15 25 I at Rh . ' Oxygen can effectively compete for Rh centres and decrease the Rh 1 1 1 substitution rate either by tying up the rhodium(I) as a Rh1 -Og complex or by oxidation to Rh***. Oxygenation of [Rh(C0)2Cl]2 i n Li Cl/DMA shows an esr spectrum corresponding to a Rh** - 02~ species (Chapter VII) and provides evidence for a Rh* - 0 2 complex. Similar observations have also been made for the oxygenation of [RhCl(C gH 1 4) 2] 2 in LiCl/DMA media. 8 , 9 144 Chapter VII THE REACTION OF MOLECULAR OXYGEN WITH DICHLORODICARBONYLRHODATE(I) IN DMA 7.1. Introduction The reaction between [Rh(C0) 2Cl 2]~ and 0 2 was studied in a non-aqueous solvent, N,N-dimethylacetamide (DMA) for comparison with studies on the corresponding oxidation in 3M HCl (Chapter VI). A number of metal complexes containing co-ordinated DMA have been prepared and 169-174 characterized , and i r studies have indicated that complexation occurs through the oxygen atom of the carbonyl group in the amide 1 7 5. 1 nc It was anticipated that the increased donor strength of DMA relative to H20 would lead to more stable intermediates which might be useful in elucidating the oxidation mechanism in both aqueous and non-aqueous media. Water is known to play a vital role in reductive carbonylation 2 3 processes ' . 7.2. The Reaction of [Rh(C0) 2Cl] 2 in LiCl/DMA with 0 2 [Rh(C0) 2Cl] 2 in LiCl/DMA prepared under vacuum at R.T. gave a yellow solution with an absorption maximum at 343 nm, e = 3150 (Figure 46). The complex exhibited almost no absorption in the visible region between 400 - 750 nm. • ,iThe> i r spectrum of such solutions showed two strong carbonyl bands at 2060 and 1981 cm"1. The [Rh(CO)2C1]2 dimer in DMA in vacuo (no added chloride) gave a brown solution having an absorp-tion maximum at 336 nm, e = 3200 (Figure 46) and also a broad absorption 145 300 320 340 360 380 Wavelength, nm Figure 46. Absorption spectrum. [Rh(C0) 2Cl] 2 in 0.5M LiCl/DMA, [Rh(C0) 2Cl] 2 in DMA. 146 between 400 - 550 nm in the visible region. The solution i r spectrum again showed carbonyl bands at 2060 and 1981 cm"1. It should be noted that possible bridging carbonyls below 1800 cm-1 are not detectable, due to the strong solvent absorption bands. The dimer, [Rh(C0) 2Cl] 2 in hexane exhibits carbonyl stretching frequencies* 7 7'* 7 8 at 2105(m), 2089(s), about 2080(vw), 2035(s), and 2003(w) cm"*. The reaction of [Rh(C0)2C1]2 with 0 2 in LiCl/DMA was followed at 75° as a function of'time using i r spectroscopy (Figure 47). 0^ was bubbled through the solution, and at frequent intervals, samples were pipetted into a NaCl liquid cell and !the i r spectrum was measured against a DMA reference. The i n i t i a l spectrum showed two strong car-bonyl bands at 2060 and 1981 cm"*. After 35 minutes, a new peak appeared at 823 cm *, while both carbonyl bands simultaneously reduced in intensity. After 3 hours, the bands at 2060 and 1981 cm"* had completely disappeared while the band at 823 cm"* had continued to intensify with time. At R.T., no reaction towards 0 2 was indicated by i r measurements. The reaction between [Rh(C0) 2Cl] 2 and air in LiCl/DMA at R.T. was monitored by esr spectroscopy (Figure 48). The solution was exposed to a i r for a known time interval and immediately frozen in liquid nitrogen to a glass for an esr measurement. The sample was then thawed L at R.T. for further reaction with a i r . It was assumed that any reaction between the metal complex and air proceeded only at R.T. and virtually ceased while frozen. No signal was detected for a sample containing the rhodium complex in LiCl/DMA prepared under vacuum. After exposure of the solu-tion to air for 2 minutes, a spectrum (A) was observed. During the f i r s t 147 \ 0 min " " Y •f \ f 35 min. i ft A 90 min. V, 180 min 2060 cm-1 'I- 1 1 i 1981 cm"' 823 cm"' Figure 47. Reaction of [Rh(C0) 2Cl 2]" with 0 2 in 0.5M LiCl/DMA ! at 75° followed by i r spectroscopy (0.0340M Rh; 733 mm 0 2). 148 gM =2.10 gx=2.oo 9r 2; i ; ,6 g2=2.035 g3=1.974 Figure 48. Esr spectra of [Rh(C0) 2Cl 2]" with 0 2 in DMA (0.2M LiCl) measured at 77°K. The reaction time is the exposure time to air at room temperature (23°); 0.055M [Rh]. 149 10 minutes, a new spectrum (B) built up. (A) slowly decreased in inten-sity while (B) further intensified. At room temperature, spectrum (B) col lapsed into a single broad peak;g = 2.03^. These esr changes were not accompanied by detectable i r changes,. indicating that only^small percent-ages .of paramagnetic species, are^-present in solution. The system was also studied using gas absorption techniques. [Rh(C0) 2Cl] 2 in LiCl/DMA slowly absorbed 0 2 at temperatures between 60-75° yielding complex plots similar to Figure 49. Detailed analyses of the gas uptake plots suggested three distinct stages in 0 2 absorption: I induction period II a stoichiometric absorption of 0 2 (02:Rh = 2:1) III catalytic oxidation. 7.2.1. Stage I: Induction Period During the i n i t i a l stage of reaction the oxygen uptake plots consistently showed a region of negligible gas absorption, which was attributed to an induction period (Figure 49). The duration of this induction period was inversely dependent on rhodium concentration for the same i n i t i a l [Cl~] (Table XXII). As the concentration of the metal complex increased, the induction period decreased. A similar inverse dependence was also observed as [Cl~] was varied at constant [Rh] (Table XXIII). The induction period, however, was essentially indepen-dent of oxygen pressure (.p\02.) over the range 175 - 733 mm. Only small differences in induction period were observed even at lowp"0 2 (Table XXII). It became d i f f i c u l t to discern whether these were actual 0 1.0 .- 2.0 : 3.0 4.0 5.0 6.0 7.0 8.0 . Time x 10 _ - j, sec Figure 49. Rate plots for oxygenation of [Rh(C0) 2Cl 2]~, 75°, 733 mm 0 2, [Rh1]: (A). 0.0236M, (O) 0.458M, 0.5M LiCl/DMA. I induction period II stoichiometric absorption of 0,,,(02:Rh =2:1) III catalytic oxidation. Table XXII Oxidation of [Rh(C0)2Cl,£]" in CDMA Summary of kinetic data at 75° [Rh] x 10 2, 0 o [LiCl] Induction period Linear rate x 10 5, k l x 1 q 3> -1 -1 -1 M mm M rmin M sec M sec 1.64 733 0.53 20 0.67 0.77 2.36 733 0.53 13 1.12 0.90 3.47 733 0.52 11 1.68 0.93 3.63 733 0.53 10 1.88 0.98 4.58 733 0.56 7 2.47 0.96 6.76 733 0.53 3 3.53 0.98 3.59 536 0.53 9 1.77 0.94 3.62 369 0.53 9 1.70 0.89 3.57 175 0.53 7 1.45 0.76 Table XXIII Oxidation of [Rh(C0) 2Cl 2]~ in DMA Effect of [Cl"] on rate at 75°, 733 mm 0 2 [Rh] 10 2 '[LiCl] [LiClOJ Induction period Linear rate x 10 5, k' bxl0 3, kn x IO3, M M M min, - l 1 - i L i _ i M sec sec M sec 2.36 0.53 13 1.12 0.49 0.90 2.48 0.38 17 0.86 0.35 0.91 2.57 0.26 14 0.65 0.26 0.98 2.56 0.12 39 0.31 0.12 - 1.01 3.58a nil 160 • very slow 3.47 0.52 11 1.68 0.49 0.93 3.57 0.39 8 1.31 0.36 0.94 3.58 0.26 12 0.86 0.24 0.93 3.57 0.13 25 0.44 0.12 0.95 3.54 0.39 0.13 14 1.34 0.38 1.00 3.60 0.25 0.26 12 0.98 0.27 1.09 3.54 0.13 0.39 23 0.51 0.14 1.10 Different uv spectrum, Figure 46. \ = ^ [Cl]. 153 changes in induction period or whether they should be attributed to the increased sensitivity of the apparatus at low pressure. The yellow colour of the starting solution remained essentially unchanged throughout the duration of the induction period. The spectrum at the end of stage I was similar to the i n i t i a l spectrum of the complex in LiCl/DMA ( A = 343, e = 2800). An esr signal was detected and was the same,but 1/20 as Intense as.the^spectrum.measured at the.end of stage/II fsee Sec. 7.2.2.). Addition of radical scavengers DBPC and DPPH to the i n i t i a l reaction mixture at 75° yielded uptake plots with extended induction periods ofLthetype shown in Figure 50. The effect of inhibitors on the induction period is summarized in Table XXIV. At a 10:1 mole ratio of DPPH:Rh, gas uptake was extremely slow with less than 2.8 x 10 moles of 0^  being absorbed after 7500 seconds for a [Rh] ca. 0.035M and 733 mm 0 2 . If a LiCl/DMA mixture was shaken under 0,, at 75° for a few hours prior to introduction of [Rh(C0) 2Cl] 2, a very slow absorption of 0^  was observed and a weak new i r band appeared at 823 cm-1. After the metal complex had been added, gas uptake was almost instantaneous. Exposure of .the solvent to oxygen in two experiments, so that 0.0080M and 0.0028M 0 2 were absorbed respectively, eliminated the induction period (Figure 50) in the f i r s t case, and reduced the induction period by a factor of 2.5 in the latter, compared to the observed induction period in the absence of any oxygen pretreatment (Table XXIV). ** CS. Si CM L O O O O O O • rn Q _ Q _ s : • —^ P— oo o • p— o _ J • o s: L O • SS o as. CM L O o L O o oo . cu E o (/» E . • o C O oo oo . o 1 o X L O • « CU €\ D -E Q _ Q h-r , -CM s : o o CM i— o o • c_>: t. t o sz • CS. .c: , '—1 M— s : o L O L O . . .• c oo o o • "1— •. +-> o (0 E ' cu CD • >> X • o : . 4-> s- . c o cu ; • st-; E-' +-> t/i (O . +-> CU o s-( Q - cu CU Q . +J (C CM Q ; O o L O cu S-H * OL * psquosqe 155 Table XXIV Oxidation of [Rh(C0)2Clo]" in DMA Effect of additives on rate at 75°, 0.53M LiCl, 733 mm 0 [Rh] x 102, M Additive Induction period min 5 Linear rate x 10 , M sec - 1 3.49 - 12 1.98 [DPPH] M 3.55 0.010 33 1.71 3.46 0.038 30 1.33 3.56 0.073 30 1.28 3.57 0.382 25 very slow [DBPC] M 3.57 0.017 25 1 .78 3.51 0.087 33 1.58 3.45 0.352 25 1.39 [OgJxlO2, M* 3.52 0.80 0 2.13 3.68 0.28 5 2.26 1.75 0.96 7 0.76 * 0 2 absorbed during pretreatment of LiCl/DMA. 156 7.2.2. Stage II: Stoichiometric Absorption of 0^ Following the i n i t i a l induction period, [RhfCOjpCllg in L i C l / DMA readily absorbed 0^ at measurable rates over the temperature range 60-75°. Typical gas uptake plots, Figure 49, consisted of an extended linear region which gradually levelled off at about a 2:1 mole ratio of 0^:Rh. Vpc analysis of the gas mixture above the solvent at the end of the reaction indicated the production of COg. The retention time for the component other than 0^ in the gas mixture was identical (retention time, 6.4 minutes, s i l i c a gel column, 40°, 150 ma, pressure setting = 10) with that for a sample of pure CO,,. Thus 1 mole of COg was probably produced for every 3 moles of 0,, consumed. At the completion of this stage of the reaction, a white solid had precipitated which could be filtered from the solution mixture. The 179 properties of the isolated compound corresponded to those for LigCO^ ( i r KBr disc v C Q = 1431, 1481 cm"1). Found: C, 15.68X (Sailed, for Li 2C0 3: C, 16.24%. •j on The solid burned with a bright red flame indicating the presence of L i . Addition of CaClg^HgO to an aqueous solution of the solid gave a 1 oi white precipitate , CaC03> Based on the formula weight of Li^CO.^, the mole ratio of LigCO-^ /Rh produced during the reaction was always very nearly 1:1. The final green-brown f i l t r a t e ( x m 3 V = 444, z = 326; xmav = 544, e = 145; X m = v = 670, e = 55) (Figure 51) gave a symmetrical esr signal at max 157 \ \ o I I I : — I L _ L L 400 450 500 550 6,00 -650 7Q0 Wavelength, nm Figure 51. Absorption spectrum after oxygenation.-at the end of 2:1 ((yRh) uptake, -after catalytic oxygenation (18 hours), 75°, 733 mm 0 2, 0.5 M LiCl/DMA. 158 R.T. with a g value of 2.03^. At liquid nitrogen temperature, the signal was s p l i t with g^  = 2.11g, g 2 = 2.03,., and g 3 = 1.97^. A solu-tion i r spectrum indicated the absence of any carbonyl groups. The presence of DBPC in the ratio of 0.5:1, 2.5:1 and 10:1 mole ratio of DBPC:Rh slowed the linear rate of Or, absorption by 10%, 20%, and 30% respectively (Table XXIV). Similar experiments using DPPH in the ratio of 0.3:1, 1:1, and 2:1, DPPH:Rh retarded the reaction rate by 15%, 30%, and 35% respectively (Table XXIV). Addition of DPPH in a 10:1 mole ratio inhibited 0^  uptake completely. Pretreatment of LiCl / DMA with 0^  prior to addition of the metal complex had l i t t l e effect on the rate of 0 2 uptake (Table XXIV). In a l l cases Li'2C03 was detected as a product. 7.2.3. Stage II I : Catalytic Oxidation During the third stage, the rate of oxygen absorption was far too slow to provide accurate data for a complete kinetic study (Figure 49). Oxygen uptake was non-stoichiometric and suggested catalytic oxidation of DMA. A typical experiment conducted at 75°, 0.0164M Rh, 733 mm 0 2,resulted in a total gas uptake corresponding to a 6:1 mole ratio of 02:Rh after 70 hours. The rate of uptake decreased as the reaction proceeded. An attempt was made to qualitatively investigate the catalytic reaction and characterize i t s products. A solution sample analyzed by vpc yielded 2 peaks of. retention time 2.0 minutes and 14.0 minutes (dinonyl phthalate column, 100°, 150 ma, pressure setting = 10). 159 Comparison of these retention times with samples run under identical conditions identified the two components as H^ O and DMA respectively. No oxidation products of DMA were detected. The visible spectrum of the final reaction solution showed absorption bands (A = 440 nm, e = 382; r max = 5 5 0 nm, e = 148, A „ = 670 nm, e = 93) (Figure 51) very similar max max to the spectrum obtained at the completion of stage II (Sec. 7.2.2.). A green solid which could be precipitated by adding methanolic Ph^AsCl to the final solution w i l l be described in Chapter VIII. A solution i r revealed that the band at 823 cm"1 was more intense than observed at the end of stage I I . The solution was s t i l l paramagnetic with an esr signal identical to that previously discussed (Sec. 7.2.2.). The intensity of the esr spectrum at the end of stage III increased with increasing tCl"]. For [Rh] ca. 0.035M the intensities of the esr signal relative to an intensity of 1 at 0.53M LiCl were: ^ 1 at 0.38M Cl"; 1/5 at 0.13 C l " , and 1/30 in the absence of added Cl". 7.3. Kinetics of the Reaction of [Rh(C0) 2Cl] 2 with 0 2 in LiCl/DMA Oxygen uptake plots were analyzed by measuring slopes in the linear region. The rate of oxygen uptake showed a good f i r s t order dependence on rhodium between 0.0164 and 0.0676M [Rh] (Figure 52) (Table XXII) at a constant iO^i and tCl"]. The rate law may be formu-lated as: -d[0 2]/dt = k^ ERh111] (7.1) where k^  contains a l l chemical and physical variables except the depen-dence on Rh. 161 Plots of rate/[Rh], k j , versus [Cl"] , Figure 53 (Table XXIII), are linear and pass very nearly through the origin. Thus the rate law may be written: -d[0 2]/dt = k^Rh1] [Cl"] . (7.2) In the absence of added [Cl ] , no 0^ absorption was detected up to 160 minutes; this was followed by a slow oxygen uptake producing a weakly paramagnetic orange solution with no absorption maxima in the visible region. When mixtures of LiCl/LiClO^ were employed to maintain constant ibnic strength (I .v 0.52MKt slightly faster rates were obtained although the f i r s t order dependence on [Cl~] was s t i l l pre-served.- . . . ... ' : - ".. ..: • • • i... • In order to analyze for^the oxygen dependence of the reaction, rate measurements were made for a series ofcexperiments in which the i n i t i a l partial pressure of 0 2 (p"02) was varied (Table XXII). k-j was essentially independent of 0 2 pressure between 175 and 733 mm (at 175 mm, k1 was s t i l l 80% of that at 733 mm). Values of k-j are summarized in Tables XXII and XXIII. The value of k1 at 75° calculated from Figure 52 was ca. 0.96 x 10 M sec -3 -1 -1 Values of k-j calculated from Figure 53 were 0.92 x 10 M sec and -3 -1 -1 0.92 x 10 M sec (I ^  0.52M) respectively. Kinetic measurements over the temperature range 60-75° at 1 atm. pressure (Table XXV) yielded a good Arrhenius plot, Figure 5'4, and the activation parameters AH^ = 16.5 ± 1.0 kcal/mole and AS* = -25.2 ± 2.8 eu. 162 0 0.1 0.2 0.3 0.4. 0.5 [Cl"] , M Figure 53. Dependence of k-j on [Cl~], ((O) variation of [Cl~]; (•) variation of [Cl~] at constant ionic strength,! ca. 0.52M), 75°, 733 mm Op, DMA. 163 Table XXV Oxidation of [Rh(C0) 2Cl 2] ~ Temperature dependence of k-j Temp. [.Rh] .x''102, . .[LiCl] Induction period k] x 10 3 ; °C M M min. sec - 1 75 3.47 0.52 11 0.93 70 3.54 0.53 14 0.68 65 3.67 0.53 22 0.45 60 3.56 0.52 25 0.30 164 165 7.4. Preparation and Characterization of [RhCO^h^O^OH] Addition of [Rh(C0) 2Cl] o (0.35g) to 20 ml. of a 70% Bu* 00H solution produced an exothermic reaction with vigorous gas evolution. An orange powder which precipitated from the yellow solution was collected by f i l t r a t i o n and washed with CH2C12 and ether to remove reactant mater-i a l , (yield 0.38g, 97%). Analysis of the compound gave C, 5.70; H, 2.44; Rh, 47.43%. Calculated for Rh 2C 20 1 2H l o; C, 5.55; H, 2.31; Rh, 47.69%. The solid is formulated as LRhCOgCh^ O^ OHj' 2. A symmetric dimeric structure is proposed containing bridging hydroxy! groups and -bidentate carbonato groups in order to maintain octahedral co-ordination 3+ around the metal. Potentiometric studies on the OH /[Rh(H20)g] system 182 indicated the possible existence of a similar diol bridged dimer , [Rh(0H)(H 20) 4] 2 +. An i r spectrum of [RhC0g(H20)20H] in a nujol mull showed an intense OH stretching mode in the region at 3400 cm-1 and an H0H bending mode at 1618 cm-1. CO stretchings were observed at 1560 cm-1 (partially obscured by an H0H bending mode) at 1270 cm-1 and 1161 cm-1. These 166 results are similar to the frequencies observed for bidentate Co 1 1 1 carbonate complex 1 8 3, [Co(NH3)4 COglCl (1593, 1265, 1030 cm"1). A mass spectrum of the solid showed an intense peak at a m/e ratio of 44, with the intensity increasing as the instrument temperature was raised. The peak can be attributed to thermal decomposition of the Rh compound producing C0 2, which was detected in the mass spectrometer as COg. No peak assignable to the parent m/e was observed. The orange powder slowly dissolved on heating in concentrated HCIO^ to give a yellow solution. This was then evaporated almost to dryness and redissolved in HCIO^. The visible spectrum of this solution showed absorption maxima at x = 314 nm, e = 75.0 and x = 402, r max max E = 58.3, similar to the spectrum reported 1 5 4 for [Rh(H 20) g] 3^ x m a x = 311, e = 6 7 ' 4 ; Xmax = 3 9 6 ' e = 6 2' 0 )-7.5 Discussion The [Rh(C0) 2Cl] 2 dimer has been reported to undergo bridge-splitting reactions with complexing agents such as [Rh(C0) 2Cl] 2 + 2L 2Rh(C0)2ClL (7.3) .38 .. ., 184 ; , m t r i l e nitrogen donor ligands halides 3 8, n i t r i l e 1 8 4 , pyridine, primary amines 3 4' 1 8 5, and some polymeric 27 167 The Ph 4As + salt of [Rh'(C0)2Cl2]" has been isolated and the i r 38 carbonyl stretching frequencies have been measured ( v C Q = 2060, 1975 cm"1). Based on the similarity in i r spectra between the isolated dichloro-anion and [Rh(C0) 2Cll 2 made up in LiCl/DMA, the most probable species in these solutions is the chlorocarbonylrhodium anion, [Rh(C0) 2Cl 21~. Solutions of the rhodium dimer in purified DMA in the absence of added Cl" would be expected to exist as mononuclear complexes containing a co-ordinated solvent molecule i.e. Rh(C0)2ClDMA. However, i r data again suggest that the major species present is [Rh(C0) 2Cl 2]". This unexpected formation of the anionic complex may be attributed either to a Cl" impurity, or. more likely, to traces of water in the sol-vent. Studies of bridge-splitting reactions of [Rh(C0) 2Cl] 2 under CO with polystyrene bonded amine resin polymers yielded, in general, the expected monomeric analogues, except in experiments where moist solvent 27 or added water was employed . In these cases, the products were trans-formed into a mixture of [Rh(C0) 2Cl 2]" and small amounts of the anionic p aggregate, [Rh 1 2(C0) 3 Q] . The isolation and characterization of.products from the reaction mixture of [Rh(C0)2Cl21 ~ with 0 2 proved to be of considerable d i f f i c u l t y . The green-brown solution (at the end of stage II and stage III) indicated the presence of a paramagnetic Rh1* species. Previous workers have reported g values for Rh 1 1 complexes 1 1 3' 1 1 5 co-ordinated by various ligands to be in the range g "= 2.03 to g = 2.11. Electron spin resonance 9 studies on the oxygenation of the rhodium(I) ..cyclo-octene complex, 168 [Rh(CgH-j4)2C1 ] 2 in DMA containing Cl" also show the existence of a Rh 1 1 complex, and the spectrum reported is identical with that observed in the present work for the green-brown solution. Visible spectral studies on the final solution U = 440 nm, e = 382; A m 3 v = 550 nm, e = 148, A „ = 670 nm, e = 93) suggest the max max III 3-presence of a dimeric Rh chloro complex, [RhgClg] , in addition to Rh 1 1 species. The (C 2H 5) 4N + salt of this dianion .(A = 435, e = 308.8, A = 540, e = 98.0) has been prepared and characterized by ma x Work and Good 1 8 6, while the corresponding Ph^As+ salt in DMA (Chapter VIII) showed similar absorption maxima, but with larger extinction coefficients = 446 nm, e = 413; A = 554 nm, e = 146). The visible max max spectrum of the green-brown solution exhibited a further unusual broad absorption between 630 and 700 nm with A = 670 nm, e = 93, that ma x could not be assigned to the Rh*** species. Such weak d-d transitions at low energy are often associated with the typical green colour of Rh** , 95,98 complexes 3_ The preparation and complete characterization of [Rh 2Cl g] and a new Rh** complex w i l l be discussed in Chapter VIII. It was assumed that the final solutions contained mixtures of Rh** and Rh***. Based on their visible spectra, the solutions appeared to contain Rh** and Rh*** in a ratio of 6:1. 7.5.1. Induction Period The observation that a new infrared band appeared at 823 cm"*, soon after exposure of [Rh(C0) 2Cl 2]~ to oxygen, and the continued 169 intensification of this absorption band even after complete oxidation of both carbonyl groups suggests that the metal complex must be involved in the oxidation reaction. The position of the band indicates that the new species probably contains a peroxidic linkage. In the absence of the rhodium carbonyl complex, oxygenated DMA solutions also exhibit a band at 823 cm-1, and oxidize I" (giving a positive test for I^ in the pre-sence of aqueous starch). Freshly prepared DMA or DMA stored under did not show the i r band or give the iodide reaction. Ir absorption frequencies in the range 845 to 875 cm"1 due to the -0-0- twisting vibra-tion have been observed for a series of hydroperoxides formed during the 187 thermal autoxidation of N-alkyl- and N,N-dialkylamides . It seems clear, therefore, that a hydroperoxide is produced during the oxidation of DMA and i t s formation is promoted by the metal complex. Several workers have concluded that the major reaction in amide oxidation is oxygen attack at the methylene or methine adjacent to nitro-1 O O "j O Q gen,forming peroxy radicals ' . By analogy with these systems, oxi-dation of DMA may involve CH3C0N(CH2« )CH3 which then leads to formation of CH3C0N(CH200H)CH3, and possibly CH3C0N(CH3)CH0 and CH^ONHChy The a b i l i t y of radical scavengers to inhibit,and hydroperoxides to initiate,the reaction confirms the involvement of free radicals in a mechanism such as: RH —»-R* + H- (7.4) R- + 0 2—*R0 2. (7.5) R = CH3C0N(CH2)CH3 R02--+ RH—*R02H +..R- (7.6) 2R02«—^nonradical products (7.7) 170 Depending on the reaction conditions, these peroxy radicals may either be removed from the system by mutual interaction or propagate chain 189 reactions by hydrogen atom transfer from a further mole of substrate The induction period may be associated with the formation of small amounts of hydroperoxide by the thermal reaction with oxygen. Amides with one or two methylene groups adjacent to nitrogen 190 are most readily oxidized while N-methyl amides are less susceptible DMA has been found by other workers to be unusually resistant to oxida-190 tion in the absence of initiators under extreme conditions . In the absence of rhodium complex, prolonged treatment of DMA with 0 2 showed no absorption of. gas unless LiCl was added (rate ^ 1 x 10~7M sec" 1, 75°, N 191 0.5M LiCl). The exact role of LiCl as an initiator is uncertain , 192 193 although i t may be involved in the decomposition of peroxide ' 7.5.2. Stoichiometric Oxidation Following the production of peroxy radicals and hydroperoxide, plausible reactions with the metal complex and consistent with the kinetics and?observations could occur as follows: [Rh I(C0) 2Cl 2]" + C l " q = | ^ I R h I ( C 0 ) 2 C l 3 ] 2 " (7.8) [Rh I(C0). 2Cl 3] 2" + R02- 2 — • [Rh^COjClg] 2" -" + RO- + C02 (7.9) 4 .RhT(C0) + 0 2 >Rh I H(C0 3) (7.10) 171 I I I 2 L L ' + T + Rh^^COg) + 2R00H »• Rh1 + 2R00- + 2H + LigCOg (7.11) Rh1 + 0 2 -RhIOz (7.12) Some reaction:- (slow or fast), with Cl" is indicated, since the rate of oxygen uptake is directly related to the added C l " Concentra-te tion. Low spin four co-ordinate d complexes can add an additional ligand to form five co-ordinate complexes. Cationic complexes of the +•• 194 69 type Rhb2(C0)3 ' where L = BPh3 have been prepared . Collman has made MCl(C0)b complexes, where M is Rh, Ir , andhb is the tripodal ligand CH3C(CH2PPh2)3; the complexes may be four or five co-ordinated in solu-tion. Chen and Hal pern have shown that IrjCl (C0)L"2,L = PMe2Ph, readi ly adds further phosphine ligands to give five co-ordinate complexes IKCI (C0)b3 and Ir(C0)bJ. Complexation of olefins byCRh(C0)2 (acac)] and [Rh(C0)2 (3-trifluoroacetylcamphorate)] is presumed to involve formation of five co-ordinate species 1 9 6. Five co-ordinate complexes have been proposed as intermediates 8 197-199 for ligand substitution reactions of d square planar complexes Nucleophilic displacement of stibine from [Rh(1,5 cyclo-octadiene) ,Cl(SbR3)] (R = p-tolyl) by amines occurs in two steps. The f i r s t cor-responds to the rapid formation of a five co-ordinate species; the second, which is slower, to i t s decomposition to give the product [Rh(l,5-cyclo-octadiene)Cl(am)] 2 0 0 , 2 0 1. [Rh(C0)(SbR3)2Cl] (R = p-tolyl) has been reported to add a ligand according to the equilibrium [Rh(CO)(SbR3)2Cl] + SbR3 ^  [Rh(CO) (SbR^Cl ] (7.13) 172 In the present system formation of a stable species I 2-[Rh (CO)2C13] would not be expected, since the complex does not contain any stabilizing ligands other than CO. However, i t s presence as a reactive intermediate must be considered. -3 -2 At higher concentrations ofnmetal (10 - 10 ty), direct reac-202 203 tion between peroxy radicals and the metal complex becomes possible ' The reaction (step 7.9) may entail prior co-ordination of the peroxy 203 radical to the metal [Cl 3Rh I(C0) 2] 2" + R00- [Cl 3(C0) 2Rh H00R] 2" (7.14) Subsequent carbonyl insertion into the rhodium peroxide bond could R00-\ LQ Cl 3(C0)Rh H—CO ». [Cl 3(C0)Rh n—C-00R]—^.[Rlj^COjClg] 2" + C02 + R0-\, 7C (7.15) produce;, a perester complex which undergoes a rapid decomposition to form Rh1, C02,and R0-. The polymerization of methylmethacrylate in the presence of Rh* species and dibenzoylperoxide has been postulated to proceed by an II 204 ini t i a t i o n system involving an unstable Rh benzoate intermediate 75 II Emanuel has reported that in the presence of Mn (stearate) 2 inhibition of hydrocarbon autoxidation occurs via intermediate formation of a III 12 3+ peroxy Mn complex, [S^ MnOOCHR R ] which then decomposes to [St 2Mn0H] 3 + and the respective R*C0R2 ketone. 173 Several novel peroxy complexes of the general formula IrX (00But)2C0L2(X = C l , L = PPh3, AsPh3 or PPh2Me and X = Br, " L = PPh3 or AsPh3) and . IrX(00CPhMe2)2(C0)(PPh3)2 (X = Cl or Br) have been isolated from the reaction of Bu^ OOH or PhMe2C00H with the respec-tive trans-IrXtC0)L2-.^ complexes. Similar reaction of Bu^ OOH with trans-IrIC0(PPh 3) 2 yields Irl^OOBu 1) (CO) ( P P h 3 ) 2 2 0 5 . Although reaction between [Rh(C0) 2Cl 3] and peroxy radical is most probable, similar 86 reactions between the metal complex and «0H or RO* cannot be excluded, Some evidence for the existence of a rhodium(I) monocarbonyl complex has been obtained. When solutions o.f:[Rh(C0)2Cl2]" in LiCl/DMA which have been oxygenated for ca./il8 hours at 75° are refluxed for 5 minutes in DMA, an impure yellow solid can be precipitated as the Ph^As+ salt. The i r spectrum of the solid in CH2C12 shows a strong carbonyl band at 1979 cm"1 similar to the carbonyl stretching frequency reported 1 6 4 for RhCOCl(PPh3)2 ( v c Q = 1978 cm"1 in CgHg). The formation of the rhodium(I) monocarbonyl probably involves decarbonylation of the solvent, DMA. I 2-[Rh (C0)C13] could rapidly absorb oxygen according to the following scheme:(cf. p. 133); 0-. Rh!(C0) + 0 2 Rh-CO -0. Rh III A Rh 1 1 1 > 0 \ 0 / (7.16) 174 The oxygen adduct could undergo rearrangement to the metal carbonate which is known to be formed at some stage. Production of the carbonate may be direct or might involve prior formation of a perester-1ike complex. Isolation of a stable peroxycarbonate intermediate from the reaction of CC^ with Ph3P. Ph3P Pt + CO, Ph3P, Ph3P Pt^ .C=0 0' PPh, J Pt(C0 3)(PPh 3) o + 0PPh3 (7.17) (PPh 3) 2Pt0 2, and the addition of oxygen by the same complex to the carbonyl group of aldehydes and ketones to give a pseudo-ozonide lend P h ~ P \ .0 0 ,R ;iPt, 1 ;Ph P , ;3 J support to the previous suggested sequence , u u' I D 1 ,<L'J{J' Several platinum metal carbonato complexes have been prepared by similar atom transfer reactions in which both the oxidant and reductant are co-ordinated to 68 74 162 69 the metal ' ' . Coll man and coworkers have isolated a stable car-bonato complex, IrCl(C0 3)L L = CH3C(CH2PPh2)3 by reacting IrCl(C0)L with oxygen. Attempts to prepare the corresponding Rh carbonato complex were unsuccessful, with C09 being evolved during the course of reaction. An 175 anionic rhodium(I) species [Rh(C03)(CO)(Ph^P),,]" has been recently 207 reported A rhodium(III) carbonato complex has been prepared from the reaction of [Rh(C0) 2Cl] 2 with Bi/oOH (Section 7.4.), supporting the involvement of such species in the oxidation of Rh* carbonyls. The production of one mole of LigCO^ per mole of Rh during the reaction of [>Rh(C0)2Cl2-]'~ with 0 2 in LiCl/DMA also certainly indicates that a car-bonato complex has been formed. It is possible that Rh***(C03) w i l l decompose hydroperoxide into free radicals (step 7.11) by a Haber-Weiss 85 mechanism , thereby also init i a t i n g the chain reaction. It is d i f f i c u l t to predict whether reduction of Rh*** proceeds by two stepwise one-7fi electron transfer steps, or a two-electron step . The nature of the ligand co-ordinated to the metal may also affect the electron transfer process. It is surprising that Li 2Q0 3,is not decomposed by the protons formed during the reaction. It appears that H + is being consumed in solution. The formation of hydrogen peroxide by protonation with weak 64 acids of metal oxygen adducts has been reported . Such an effect is suggested in the present system, since a 2:1 mole ratio of p-toluenesul-fonic acid:Li'2C03(suspension)(0.071M:0.041M) in 0.5M LiCl/DMA reacts quite slowly to liberate C0 2, in the absence of Rh complex at 75°. Although the detailed mechanism of electron transfer (step 7.11) is s t i l l in doubt, prior complex formation between the hydroperoxide and catalyst has been indicated in other systems 7 5' 7 8' 8 6' 8 7. The rate of autoxidation of diphenylmethane catalyzed by Rh(CO)(PPh,)9Cl decreased 176 in the presence of cyanocarbon olefin complexing agents 7 7 with increas-ing electrophilicity of the olefin. The blocking of an available bond-g ing site on the d complex was thought to indicate involvement of metal hydroperoxide intermediates. Rh1 chlorides in LiCl/DMA rapidly absorb oxygen up to a 1:1 Q mole ratio at ambient conditions according to the following equation : T K T Rh1 + 0 2 , — Rh x0 2 (7.18) The solution i r spectrum shows a band at 895 cm"1 attributable to the 0-0 stretch. In the present system, the Rh* produced by a Haber-Weiss type mechanism (step 7.11) would similarly be expected to form a molec-ular oxygen complex. The overall stoichiometry up to the end of the second stage can be written (equations 7.8-7.12): 2R00H + R- + [Rh(C0) 2Cl 2]" + 302 > Rh !0 2 + C02 + CO,' -: u L, + RO- + 2R00- + 2H+ (7.19) The rate law derived from the mechanism assuming a steady state concen-tration for [Rh(C0) 2Cl 3] 2" i s : k-ikptRh1] [Cl] [R00]] " d [ ( V / d t = ;c': k ^ R . [R00-] <7-2°) Previously measured values of k 2 for reaction of peroxy r a d i c a l s 7 5 with Co 1 1 or Mn11 are in the range of 10 4 - 10 6M - 1 sec" 1. The 177 rate data (Table XXIV) indicate l i t t l e difference in rate for the system which was i n i t i a l l y free of hydroperoxide,compared to systems which con-tained hydroperoxide that was prepared "in situ " to different concentra-tions. The rate then appears to beiessentially independent of peroxide; thus i t may be assumed that k^ CROO*] » k ^ . The rate law reduces to: -d[0 2]/dt = k^Rh 1] [Cl] (7.21) The reaction seems to occur via a rate-determining formation of a reactive five co-ordinate complex, [Rh^CO^Clg] . Activation para-meters :for the ceact*ontAbt=16.5 kcal/mole and. AS^=-25.2/euuare ,sim-^lar 2 w,to the activation enthalpy and entropy changes commonly observed II 208 £ for substitution reactions of square planar Pt complexes (AHr from 8.4 to 23.0 kcal/mole and AS^ -14 to -34 eu),which are known to proceed via an associative mechanism. The large negative entropy is consistent with a slow step involving an expansion in the co-ordination number of the metal. 7.5.3. Catalytic Oxidation The final stage of reaction involves a very slow and inefficient catalytic oxidation of the solvent (rate ^  5 x 10~Sl sec" 1, 75°, 0.0458M Rh). Two general paths exist for oxidation of organic substrates, using molecular oxygen namely, dissociative oxygen insertion (oxygen atom transfer), or a free radical chain mechanism. There are very few examples for which direct oxidation of the organic substrate by molecular oxygen 178 has been demonstrated, although i t has been reported that Rh 1(PPh 3) 3Cl promotes oxidation of terminal olefins to methyl ketones by oxygen atom transfer as a result of the particular a b i l i t y of the rhodium complex to 84 activate oxygen . In contrast, the oxidation of cyclo-octene in the presence of rhodium(I) has been suggested to occur by a free radical mechanism1^. peroxide decomposition, or via direct chain ini t i a t i o n at the metal 209 80 centre has not yet been clearly established . Stern has reported that Pd° catalyzed autoxidation of cumene occurs by direct chain i n i t i a -79 tion, while Sheldon has postulated that chain i n i t i a t i o n proceeds by decomposition of the hydroperoxide s t i l l present in solution. related to the capacity of the complex to absorb oxygen. A plausible mechanism, based partly on the findings of other workers in this labor-atory is as follows: Whether the free radical chain mechanism proceeds via hydro-The mechanism for the autoxidation of DMA containing Rh may be Rh'O (7.22) ini t i a t i o n propagation 2R00- ••non radical products (7.26) termination R = CH3C0N(CH2)CH3 179 The rhodium(I) oxygen complex can exist in equilibrium with a superoxide species. Rh 0~ catalyzes direct chain i n i t i a t i o n by hydrogen abstraction from the substrate yielding a hydroperoxide intermediate. [Rh**]00H~ w i l l regenerate the catalytic species by interaction with a peroxy radical (step 7.25) or form an inactive Rh*1 species! [RhII]00H" *-Rh n + 00H" (7.27) The f a l l off in rate of the catalytic oxidation is probably due to production of Rh 1 1, which may be further oxidized to Rh 1 1 1. Both Rh 1 1 and Rh**1 species were identified in the oxidized solutions. The superoxide Rh**0^ is believed to be the species detected by esr spectroscopy (Figure 48) (spectrum A). This appears to be a composite of two components: one component with = 2.10 and gj_. = 2.00; 210—212 g a v = 2.03, which is similar to spectra recorded for the 0 2 ion . The second component (gj, g 2, g^) appears similar to spectrum (B) (g 1 = 2.116, g 2 = 2.035, g 3 = 1 .97^) which has been attributed to a Rh** species. A relatively weak interaction between the Rh** and 0 2 centers would give a spectrum essentially the superimposition of 0 2 and Rh**. II -The g 3 value of the latter is somewhat modified in the Rh 0 2 species and Rh II is labeled g^ in the (A) spectrum; the g^ and g£ values for the component in (A) appear almost indistinguishable from g1 and g 2 > Autoxidation of cumene in the presence of Ag° has been postu-lated to occur through a Ag*02 species, as a result of^the a b i l i t y of 213 silver to chemisorb oxygen in a inon-dissociative manner . An esr signal has been observed in the presence of Ag°/02, and cumene at 180 g = 2.016, which has been attributed to the cumene peroxy r a d i c a l ^ 1 4 . Chain ini t i a t i o n by cobalt(II) and copper(I) complexes has also been suggested to occur by formation of superoxide oxygen complexes. A comparison of the.intensity of the esr signal recorded at the conclusion of each gas uptake experiment as a function of the [Cl~] indicates that esr detectable Rh1* was less readily formed at low [Cl~]. This effect may be evidence for a Rh** dimer involving Cl bridges (see Chapter VIII). An equilibrium of the type: dimer ~ monomer + Cl~ (7.28) suggests that as [Cl~] is decreased, increasing amounts of monomeric Rh* would be formed. Monomeric Rh** may be more susceptible to oxidation to Rh1**. 181 Chapter VIII PREPARATION AND CHARACTERIZATION OF [Ph 4As] 2[Rh 2Cl 6(DMA) 2l AND [Ph 4As] 3[Rh 2Cl 9] 8.1. Introduction The unusual paramagnetic behaviour of oxygenated solutions of [Rh(C0) 2Cl 2]" in LiCl/DMA indicated the existence of a stable Rh 1 1 species. Most of the previously reported compounds of divalent rhodium are diamagnetic dimers, containing a strong Rh-Rh bond. (See Introduc-tion, Sec. 1.7.) Few paramagnetic complexes of Rh** have been isolated and thoroughly characterized. The square planar salt L~(n-C4Hg)4N]2 [Rh(MNT)2].. (maleonitriledithiolate dianion, [S 2C 2(CN) 0] 2~) has been prepared and i t s magnetic properties investigated** 5'** 6. The reduction of rhodium(III) halides with bulky phosphine ligands yields monomeric and dimeric complexes of paramagnetic rhodium(II)* 1 1 - 1* 4. Other genuine paramagnetic Rh** species are the bis-n-cyclopentadienyl complex**7, Rh(C 5H 5) 2, the hexamethylbenzene complex 1 1 8, [Rh{Cg'$.CH3)6}']'2+-.and. the 119 tetraphenylporphyrin complex , Rh(TPP). In the present studies, an anionic complex of divalent rhodium was isolated from oxygenated solutions of [Rh(C0) 2Cl 2]" in Li C l / DMA as the tetraphenylarsonium salt. The compound was characterized by absorption, and esr spectroscopy, magnetic moment, electrical conductance, and~molecular weight measurements. The complex in DMA solution was slowly oxidized by atmospheric oxygen to a known Rh*** species. The Rh*** compound was also f u l l y characterized, including a preliminary structure determination by x-ray diffraction. 182 8.2. Experimental 8.2.1. Preparation of [Ph 4As] 2[Rh 2Cl g(DMA) 2] 1.03 g of LiCl was dissolved in 35 ml DMA in a two-necked flask fitte d with a gas bubbler. The mixture was saturated with 0 2 and heated to 75° on an o i l bath. 0.37 g of [Rh(C0) 2Cl] 2 was added while oxygen bubbling was maintained at 75° overnight (18 hours). The colour slowly changed from the i n i t i a l yellow, becoming dark brown after 3 hours. After cooling to room temperature, the solution was filtered to remove Ln^CO^. 2.58 g Ph^AsCl,dissolved in a minimum volume of CH^ OH, was added to the brown f i l t r a t e andthe mixture was briefly stirred at room temperature. The solution was cooled in an ice bath and treated with 60 ml of ether to yield a brown o i l . The ethereal layer was decanted and the;oil was further extracted with four more portions of ether. On addition of 50 ml of ice cold H20 to the o i l , a fine green powder precipitated. After cooling the suspension for 48 hours, the solution was f i l t e r e d , and washed successively with H20 and ether, and dried in vacuo overnight (yield 0.68 g, 48%). A pure compound was obtained by redissolving the crude product in DMA and repeatedly washing with ether. Calcd. for [Ph 4As] 2[Rh 2Cl g(DMA) 2] : C, 49.45; H, 4.27; N, 2.06; C l , 15.67; Rh, 15.16%. Found: C, 49.23; H, 4.32; N, 1.74; C l , 15.79; Rh, 15.39%. 183 8.2.2. Preparation of [Ph 4As] 3[Rh 2Cl g] 2.19 g RhCl^'SH^O was dissolved in 50 ml of DMA containing 2.62 g of L i C l . The red solution was stirred at 80° and became orange-brown after 1 hour. Upon cooling the solution to room temperature, 10.0 g of Ph^AsCl, dissolved in a minimum volume of CH^ OH, was added to the mixture. Treatment of the orange-brown solution with 150 ml of H^ O then gave a brown precipitate. The solid was collected by f i l t r a t i o n , washed with ice cold H20 and ether, and dried in vacuo overnight (yield 5.37 g, 77%). Calcd. for [Ph 4As] 3[Rh 2Cl g]: C, 51.60; H, 3.58; C l , 19.08; Rh, 12.30% Found: C, 50.23; H, 3.66; C l , 18.91; Rh, 12.18%. 8.2.3. Reaction of [Ph4As]2[Rh2Clg(DMA)2] with HOAc/NaOAc 0.070 g of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] was dissolved with mild heating for 1 hour in a mixture of 40 ml glacial HOAc/NaOAc (pH -v 6). The green solution was concentrated by evaporation down to small volume. The colour of the solution slowly changed from green to yellow and f i n a l l y back to green. The visible spectrum of the concentrated solu-tion exhibited absorption maxima at 470 nm and 632 nm. On cooling, green-blue crystals precipitated from solution. The visible spectrum of a solution of the green-blue solid in H,,0 showed two absorption maxima (A ,„ = 450 nm, e = 156; A , „ = 590 nm, e = 279). The visible max max spectrum is similar to that reported for [Rh(vCH3.C02)2H20]2. 184 8.2.4. Oxidation of [Rh 2Gl 6(DMA) 2] 2" Oxidation of DMA solutions of [Ph4As].2[Rh2Clg(DMA)2] by atmospheric oxygen could be followed by visible and esr spectroscopy at room temperature. The colour slowly changed from green to orange. The decrease in absorption maximum at 670 nm (Figure 56) was accompanied by a corresponding decrease in intensity of the-esr signal (Figure 59). The final visible spectrum showed absorption maxima at A = 444, e = 243 and A = 540, e = 99. Addition of an excess of LiCl retarded the oxida-tion rate. 8.2.5. Attempts at Growing Crystals of [Ph 4As] 2[Rh 2Cl g(DMA) 2] Slow evaporation of saturated solutions of the salt, [Ph 4Asl 2 [Rh2Cl6(DMA)2] in CH2C12:CH30H, 1:1, yielded brown-green needles of 2 mm edge. Microanalysis and a preliminary structure determination by x-ray diffraction indicated that Rh 1 1 had been oxidized to tervalent rhodium and reprecipitated as [Ph 4As] 3[Rh 2Cl g]. Calcd. for [Ph 4As] 3[Rh 2Cl g]: C, 51.60; H, 3.58; Cl, 19.08; Rh, 12.30% Found: C, 49.91; H, 3.60; C l , 19.43; Rh, 12.60%. 8.2.6. Esr of .Rh 1 1, in a [Ph 4As] 3[Rh 2Cl g] Host Lattice Crystals obtained by the previous method gave an esr signal as a result of paramagnetic impurity of Rh**. Single crystals were mounted 185 on a polystyrene wedge in such a way as they could be rotated around one of the principal axes. Measurements were performed to obtain the angular variation of g values. 8.3. Discussion 8.3.1. Preparation of [Ph 4As] 2[Rh 2Cl g(DMA) 2] From the postulated mechanism for the oxidation of [Rh(C0) 2Cl 2] in LiCl/DMA by molecular 0 2 (Chapter XII), formation of Rh 1 1, Rh 1 1 super-oxide, or Rh*** species would be expected. Scheme I T TT - TTT ? Rh/Og v Rh 1 X0 2^ ^ R h 1 1 ^ For some preparations,mixtures of both Rh l x and Rh l u complexes have been observed. The Rh** complex was detected by i t s characteristic esr spectrum (Sec. 8.3.9), and the Rh*** complex was assigned on the basis of i t s visible spectrum (Sec. 8.3.7). Mixtures of the two species [Ph 4As] 2[Rh 2Cl 6(DMA) 2] and [Ph 4As] 3[Rh 2Cl g] in CH2C12could be separated by thin layer chromatography on fluorosil plates using methanol as developer ( R f R h I I = 0.65) (Rf:.RnIII = 0.0). Traces of acid (DMAH+) 186 generated during the oxidation process may affect the equilibrium by favouring the formation of rhodium(III) hydroperoxides (Scheme I ) . 2- 217 The hydrido complex [RhH(CN)4H20] has been reported to react with 0 2 producing the hydroperoxo complex [Rh(02H)(CN)4H20] The ethylenediamine and ammine substituted y-amido-u-peroxodicobalt "2-18 complexes undergo protonation in acidic solution : . 3+ / N H 2\ Co Co W c - l Co ' Co \ 9 / 3+ + H i l l * • / N H \ Co Co H H c-2 / N „ 2 X Co Co I OH 4+ -i'4+ followed by rapid isomerization to the corresponding hydroperoxo complex. James and Ochdai^ have postulated that oxidation of cyclo-octene in LiCl/DMA proceeds via intermediate Rh hydroperoxides. Under conditions in which acid production (DMAH+) is inhibited, Rh 1 1 could be formed exclusively. 8.3.2. Preparation of v . i [Ph 4As] 3[Rh 2Cl g] Previous preparative;methods for quaternary ammonium salts of 3-[Rh 2Cl g] have involved either extraction of the anionic complex from 187 aqueous media by organic solutions containing a quaternary ammonium h a l i d e s a l t 1 8 6 ' 2 1 9 , R4NC1; or by the direct addition of R4NC1 to an aqueous solution containing a cloro aquo complex of rhodium followed by solvent e v a p o r a t i o n 1 8 6 ' 2 2 0 ' 2 2 1 . Equilibrated DMA solutions, or LiCl/DMA mixtures of RhCl3'3H20 222 exhibited absorption maxima at X = 442 nm, e = 335 and x = 530 max max nm, e * 150, similar to those reported for [Rh 2Cl g] 3" (Table XXVI)."~ Isolatri.onnof the tetraphenylarsonium salt, [Ph 4As] 3[Rh 2Cl q] (reflectance spectrum A m 3 v = 450 nm and A „ = 555 nm) from these solutions verified max max that the anionic dimer is the major component in DMA media. The reaction mechanisms proposed for the catalytic hydrogenation of substituted ethy-l e n e s 1 5 0 by RhCl3-3H20 in DMA and the direct carbonylation 3 of RhCl3*3H20 in DMA should be reconsidered in view of:the existence of such species. 8.3.3. Infrared Spectra of [Ph 4As] 3[Rh 2Cl g] A comparison of the spectral data for Rh 1 1 and Rh 1 1 1 complexes suggests that structural similarities exist between the compounds. Far infrared vibrational spectra indicate that both species have terminal and bridging chlorine stretching modes. An ion of the typeO^Cl^T 1 - of D^ symmetry is expected to have two terminal stretching modes and two bridging stretching modes. However, only 3 bands are observed for the rhodium complex. The observed spectrum consists of the same number of • j o e 00"i bands, the same relative intensity, and position ' as those of the 188 Table XXVI Infrared spectra of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] and [Ph 4As] 3[Rh 2Cl g] Compound v(Rh-Cl) Reference v.cm"! [Ph 4As] 2tRh 2Cl 6(DMA) 2] b 330(s), 318(s), ?272(w) This work [Ph 4As] 3[Rh 2Cl g] 330(s), 319(s), 274(m) This work [Et 4N] 3[Rh 2Cl g] 331(s), 319(s), 277(m) 186 Abbreviations: s, strong; m, medium; w, weak; br, broad 'v(C-O) and ^ ^C_H^ of co-ordinated DMA overlapped with skeletal ring stretch of phenyls in Ph 4As +, (1620 - 1540 cm"1) (br). 189 [Et 4N] 3[Rh 2Cl g] (Table XXVI) and tEt 4N] 3[Cr 2Cl g]. In the case of the Cr complex, the absent bridged Cl stretching band is thought either to occur at a frequency lower than 200 cm"1, or to be obscured by the ter-223 mmal metal chlorine stretching mode . The latter explanation has 186 been thought to apply for the rhodium complex, since i t seems unlikely 3_ that the absent band occurs below the "bending" mode for [RhClg] which is found at 205 cm-1. 8.3.4. Infrared Spectra of [Ph4As]2[Rh2Clg(DMA)2] The infrared spectra of various transition metal salts contain-ing co-ordinated DMA have shown a considerable shift of the carbonyl band to lower frequency, suggesting that DMA is bonded through oxygen in these complexes 1 6 9" 1 7 1. The broad band between 1620 - 1540 cm"1 is most likel y due to the lowered C=0 and C-N stretching frequencies for co-ordinated DMA overlapped with -C=C- skeletal stretching modes of the Ph 4As + cation. Similar -C=C- vibrations were observed in the region of 1623 cm"1 for [Ph 4As] 3[Rh 2Cl g]. Based on thecobserved low magnetic moment, a dimeric structure of C 2 h symmetry is suggested for [Rh2Clg(DMA)2] , Group theory for this structure predicts 3 i r active ^ stretching modes. Two strong bands at 330 cm-1 and 318 cm"1 can be assigned to terminal chlorine stretching modes. A very weak band appears at 272 cm"1, which may be partially obscured by the terminal stretching modes in the region previously assigned to bridged-chlorine (Sec. 8.3.3.). The relatively weak absorption further supports the postulated symmetric 190 structure. A similar i r spectrum has been observed for the chlorine-bridged dimeric complex 1 1 4 [RhCl(siphos),,]0. 8.3.5. Electrolytic Conductances of [Ph4As]0[RhCl6(DMA)0] and [Ph 4As] 3[Rh 2Cl g] in DMA (Table XXVII) Molar conductances, A , measured in DMA at R.T. show that both [Ph 4As] 3[Rh 2Cl g] and [Ph 4As] 2[Rh 2Cl g(DMA) 2] are dissociated in dilute solutions. For a series of ionic transition metal complexes, containing co-ordinated DMA the following range of molar conductances17^ measured o 1 2 in DMA at 25° have been observed: 1:3 complexes (200 - 250 ohm" cm -1 -1 2 -1 mole ); 1:2 complexes (130 - 180 ohm cm mole ): 1:1 complexes - 1 2 -1 - 1 2 - 1 (<85 ohm cm mole ). A m values between 40-60. ohm .'cm ,.mole . have been obtained for L i C l , AgN03, ( nBu) 4NI, Ph 4AsCl, and p-toluenesulfonic acid in DMA media 1 6 9. Lower molar conductances are to be expected for the rhodium complexes, since both the tetraphenylarsonium cation and the binuclear anion are bulky, and therefore lack ionic mobility. For [Ph 4As] 2[Rh 2Cl g(DMA) 2], = 115 ohm"1 cm2 mole"1 f a l l s slightly below -1 2 the range established for 2:1 electrolytes. A A m value of 147 ohm cm mole"1 measured for [Ph^As]3[Rh2Clg] is considerably lower than the expected range for 3:1 electrolytes. A 1:1 electrolyte containing a binuclear anion e.g. the rhodium complex of N-phenylanthranilic acid, -1 2 H[Rh2(PAA)2Cl] has a molar conductance value in DMF of ca. 30 ohm" cm mole 1, which is approximately 1/2 the usual value for monomeric uni-224 univalent electrolytes in this solvent . 191 Table XXVII Electrical conductances of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] and [Ph 4As] 3[Rh 2Cl p] in DMA at 23° m Compound , -1 2 , -1 Ionic Nature ohm cm mole a M a i , u , e [Ph 4As] 2[Rh 2Cl 6(DMA) 2] 115 2:1 [Ph 4As] 3[Rh 2Cl 9] 147 3:1 192 8.3.6. Molecular Weights of [Ph 4As] 2[Rh 2Cl g(DMA) 2] and [Ph 4As] 3[Rh 2Cl g] in Solution (Table XXVIII) Molecular weight measurements for [Ph 4As] 2[Rh 2Cl g(DMA) 2] lend further support for the formulation of a dimeric structure for the complex. The observed molecular weight in CHC1^ of 1050 was intermediate between the expected monomeric molecular weight of 679.5 and the dimeric molecular weight of 1359. The compound appears to be partially disso-ciated into the cation and dianion. In contrast, a molecular weight of 377 measured for [Ph4As]^[RhgClg] in DMF corresponded to complete dissociation of the salt. The compound was insoluble in solvents (e.g. C gH g, CHCl^) in which dissociation would not be expected to occur. 8.3.7. Visible Spectra of [Rh 2Cl g] 3" (Table XXIX, Figure 55) Positions of absorption maxima, x . (Table XXIX) in the vis i -max ble spectra of [Ph 4As] 3[Rh 2Cl g] measured in both DMA and CH2C12 corres-3_ ponded to previously reported data for the ion, [Rh 2Cl g] , in CgHg 1 8 6 and CgHj-NO,,219. However, the molar absorptivities, e, were considerably greater in the present studies than the values obtained by 186 other workers . The spectrum of [(C 2Hj-) 4N] 3[Rh 2Cl g] measured in 221 CH3CN also features higher values of e. It is possible that the ion "J O C T T T as prepared by Work and Good contained some monomeric Rh chloro . 154 species which have similar absorption maxima but much smaller extinc-tion coefficients (e.g. RhCl 3" in HC1/HC10- has X = 411, e = 93.8; 3 6 4 max Amax = 5 1 8> e = 1 1 1 -5>-193 Table XXVIII Molecular weights of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] and [Ph 4As] 3[Rh 2Cl g] in solution Compound Solvent MW Comments [Ph 4As] 2[Rh 2Cl 6(DMA) 2] CHC13 1050 partially dissociated [Ph 4As] 3[Rh 2Cl g] DMF 377 completely dissociated Table XXIX Visible spectral data for [Rh:2'ei6(DMA)2]2" and [Rh 2Cl g] Ion Solvent A a max, nm b e Reference [Rh 2Cl 6(DMA) 2] 2" DMA 442(sh), 540(sh), 670 (286), (148), 101 This work [Rh 2Cl g] 3' DMA 446, 554 413, 146 This work [Rh 2Cl g] 3" CH2C12 444, 546 402, 155 This work [Rh 2Cl g] 3" C6 H6 435, 540 308.8, 98.0 186 [Rh 2Cl g] 3" C6H5N02 442, 546 219 [Rh 2Cl g] 3" CH3CN 440, 544 441, 130 221 Abbreviations: (sh) shoulder. k.Per rhodium. 195 Wavelength, nm Figure 55. Absorption spectrum of [Ph 4As] 3[Rh 2Cl g] in DMA. 196 A preliminary x-ray diffraction study shows that the complex ion consists of two distorted octahedra sharing a trigonal face (Sec. 8.3.10.). As a result, the ligand f i e l d spectrum may be simp!istically interpreted in terms of l:ow spin d&'lRh in an octahedral f i e l d . Similar-3_ i t i e s exist between the electronic spectrum of [Rh2Clg] and that 3_ reported for the [RhClg] ion. Displacement of absorption maxima and 3- 3-increased band intensities for [Rh 2Cl g] relative to [RhClg] may be due to a charge-transfer contribution to the essentially pure d-d tran-sitions resulting from distortion of the octahedral co-ordination around the metal. Similar comparisons have been made between the electronic spectra of [Cs] 3[V 2Cl g] , [Et 2NH 2] 3[V 2Cl g] and the [VClg] 3 - , i o h 2 2 3 > 2 2 5 - 2 2 7 , 8.3.8. Visible Spectra of [Rh2Clg(DMA)2]2" (Table XXIX, Figure 56)' 7 2-The electronic spectrum for the d [Rh2Clg(DMA)2] ion is considerably more d i f f i c u l t to interpret than the preceeding case. The ion is best described as consisting of two mutually independent square planar "RhCl^" units sharing equatorial sites and a weakly interacting solvent molecule, DMA, in the axial position. The ligand f i e l d strength "jyo pop. of Cl as compared to DMA is a factor of 1.5 - 2 times greater ' , hence the geometry of the complex w i l l be significantly influenced by the inplane equatorial ligands and less influenced by the weaker axial ligand. Two possible orbital energy schemes may be proposed for the tetragonally distorted complex. 197 198 d x 2 - / d ? 2 <I> xy xz dz2 (a) d 2 z xy (b) Configuration a) has been suggested for octahedrally co-ordinated Rh 229 ions stabilized in a zinc tungstate la t t i c e , while configuration b) has been a t t r i b u t e d 1 1 6 ' 2 3 0 to [Rh(MNT)2]2" and [RhCl 2(CN) 41 4~. For configuration a),the visible bands might be assigned to transitions from the lower f i l l e d d , d , and d ? orbitals to the d or d ? ? xz yz zc xy x^-y^ orbital. Similarly, the expected transition for configuration b) should occur from the f i l l e d d x z , d y z , and d ^ orbitals to the dz2 and d x2_y2 orbital. 199 8.3.9. Magnetic Properties and Esr of [Ph4As]2[Rh2Clg(DMA)2] A consideration of the magnetic properties and esr spectrum for [Ph^As]2[Rh2Clg(DMA)2] may be useful in differentiating between the postulated electronic configurations. The magnetic moment of the com-plex (0.95 BM at 22.2° ) determined by the Gouy method (Table XXX) is subject to some uncertainty owing to a large diamagnetic correction. However, the value is much lower than the spin only value of 1.73 BM expected for a single unpaired electron. The acetate compounds of rhodium JRhfCKjCO^U ^containing different terminal ligands are dia-magnetic 9 7 (ca. 0.5 BM/Rh) with a strong Rh-Rh bond 2.4A°) 8 9' 9 0. The unusually low magnetic moments noted for some monomeric phosphine and siphos complexes of Rh 1 1 in the solid state can be attributed to intermolecular metal-metal i n t e r a c t i o n s 1 1 1 " 1 1 3 . A similar type of 119 interaction may also exist in the divalent rhodium porphyrin complex ( y e f f ~ 1.2 B M ) . The esr spectrum of the species formed during the oxygenation of [RhCl(CgH 1 4) 2l 2 in LiCl/DMA exhibited signals with g values similar to those observed for [Ph 4As]^[RI^Cl^DMA)^ in DMA (Table XXXI), (Figs. 57-59) Detailed analysis of the former spectrum indicated a planar structure 234 with the odd electron occupying a d orbital . xy If the unpaired electron in [Ph 4Asl 2tRh 2Cl g(DMA) 2] is present in a d x y orbital situated in the plane between the equatorial Cl ligands, a weak face to face interaction may be possible between the half-occupied orbitals on the metal atoms. The resulting weak coupling of 200 Table XXX Molar susceptibi l i t y and magnetic moment of [Ph 4As] 2[Rh 2Cl 6(DMA) 2] T ' G n n6 a ,b x M.xlO cgsu a p e f f BM 295.4 377.5 0.95 136.2 560.6 0.78 aPer rhodium. b -6 Including a diamagnetic correction of 399 x 10" cgsu per monomer unit; this fs.scomppsedr" of a value of 27 x 10"6 cgsu 2 3 1 for Rh 1 1, 20.1 x 10"6 cgsu 2 3 1 vSor C l , 253 x 10"6 cgsuc for Ph 4As + and 58 x 10"6 cgsu d for DMA. Estimated from the experimental value of 177 x 10 cgsu for PhgAs and 231 233 Pascal's constants ' for substitutuents. Estimated from the experimental value of 34 x 10~6 cgsu 2 3 2 for CH3C0NH2- and 233 Pascal's constants for substituents. 201 electronic spins leads to the observed large reduction in magnetic moment. If however, the odd electron is located in a d 2 o r b i t a l , no significant interaction between the metal centers would be expected, hence a larger magnetic moment should be observed. A comparison of the g values calculated from the esr spectra for [Ph^As]2[Rh2Clg(DMA)2] in a strongly co-ordinating solvent, DMA, a weakly co-ordinating solvent CH^Cl^. as well as the values reported q from the oxygenation of [Rh(CgH 1 4) 2Cl] 2 in DMA/LiCl indicate the presence of a common paramagnetic species. In DMSO, a shift in the position of the signals (Table XXXI) suggests probable replacement of the co-ordinated DMA by DMSO. Table XXXI Esr spectra of [Ph 4As] 2[Rh 2Clg(DMA)2] Solvent ^isotropic 3 g • i • i 3amstropic 9i> g2> b'g 3, g d  3av Reference polycrystalline s o l i d 3 2 .103, 2.031, .K97.2 2.035 This work CH2C12 2.037 2 .103, 2.031, 1.971 2.035 This work DMA 2.034 2, .103, 2.031, 1 .971 2.035 This work DM SO 2.041 2, .111, 2.041, 1.982 2.044 This work LiCl/DMAC 2. .n 4, 2.048, 1.96g 2.044 234 determined at 298°K. determined at 77 K°in the glass state. cOxygenated solution of [Rh(CgH 1 4) 2Cl]_ 2. d gav = 1 / 3 ( g l + g2 + g3]-203 4-204 206 The electron paramagnetic resonance study of Rh** in a [Ph^AslgCRhgClg] lattice w i l l not be discussed in any detail. No hyper-fine coupling was observed since linewidths (20 - 25 gauss) were broader than the expected splitting due to interaction with the Cl nucleus. The magnetic axes containing the principal g values have unique directions with reference to the crystallographic axes. The measured g values are: g-| = 2.095, gn = 2.028, g^ = 1.968. 3_ 8.3.10. Structure of the [Rh^Clg] Ion The crystal structure of [Rh^Asl^tRh^Clg] has been determined by x-ray diffraction using Patterson and Fourier techniques. The compound crystallized in the monoclinic system, space group P2q/c, with lattice constants a = 13.23, b = 20.95, c = 26.23A0, and 3 = 103.63°. The 3 measured density P f | o t = 1-57 g/cm agrees with that calculated for four 3 formula units in the unit c e l l , P c a - j c c j = 1-60 g/cm . Preliminary refinement of the structure to a value R -v 0.08 based on approximately 3500 reflections (Mo K«) located the positions of a l l Rh and Cl atoms. The molecule possesses features in common with those of the 235-239 previously determined structures s , CsgCr 2Clg, CSgWgClg, and 3_ Cs-jTlgClg. jhe essential geometry of the [Rh^Clg] ion is,depicted in Figure 60. The ion contains two octahedral ly co-ordinated "RhClg" units sharing a common face. Co-ordinative saturation around the metal and the unusually long Rh ••••Rh distance of 3.11 A° clearly indicate the absence of any metal-metal interaction. The rhodium atoms are displaced 207 208 from the centers of the octahedra, so that the Rh-Cl distance is shorter when the chlorine atom belongs to one octahedron (2.32 (1) A°), than when i t is shared between two (2.36 (2) A 0 ) (Table XXXII). The Rh-Cl t bond length is not significantly different from the sum of-the elemental covalent radii (1.34 + 0.99 = 2.33A0) or from other III 240 Rh -Cl bond lengths recently observed; the two bonds in trans-[Rhpy4Cl0]-N03-HN03 are equal at 2.340 (7) A 0 , the three bond lengths 2 4 1 in 1,2,6-trichlorobis-(6-phenylenediimethylaminedimet.hyl.arsine)rhodium( III) average 2.33 (1) A 0 , and the Rh-Cl bonds in [(CH^N] [RhCl4(.H20)o] have an average l e n g t h 1 5 8 of 2.33 A°. The average Rh-Cl^ distance of 2.36 (2) A° in [Rh 2Cl g] 3" e'iso similar to the Rh!-Cl bond lengths 3 6 in [RhCl (CO)^ (2.35 A 0 ) ; [RhCl(C 8H 1 2) 2l 2 (2.38 A 0 ) 2 4 2 ; and [RhCl(C 2H 4) 2] 2 (2.40 A 0 ) 2 4 3 . However, none of the examples cited contain a chlorine bridge between two rhodium atoms of formal Rh 1 1 1 oxidation state. The greater positive charge of Rh 1 1 1 relative to Rh1 and hence smaller covalent effective radius w i l l result in a shortening of the Rh-Cl bond. The bond distances derived from the formally Rh1 complexes should be regarded as an upper limit to III III 244 the value expected in Rh systems. The average Rh -Cl^ distance trans to ff bonded carbon in [RI^Cl^CgH^O)^-CHgOH is 2.386(4) A 0 . The small shift of the Rh atoms away from the centers of the octahedra causes a pyramidal distortion resulting in a Cl, -Rh-ClL b b angle of 81(2)° as compared to the regular octahedral Cl^-Rh-Cl^. angle of 91(1)° (Table XXXII/I). The Rh-Cl b bond has been bent slightly away from colinearity with the trans-Rh-Cl^ bond by ca. 7°, the angle being 173(3)°. 209 Table XXXII Principal bond distances in [Ph^As]^[RhoCl g] A4.„mf.a Distance 1 3' c Atoms ^ A o j Rh-.-Rh 3.11 Rh-Cl t 2.32(1) Rh-Cl. 2.36(2) a C l t , terminal C l ; C l b , ^Averaged over a l l bond type. cNumber i;n parentheses figure quoted. bridging Cl. lengths of a common is rmsd in the last 210 Table XXXIII Principal bond angles for [Ph.As]o[Rh9ClQ] Atomsa n i b,c Angle Cl b-Rh-Cl b 81(2) Cl.-Rh-Cl. b t (cis) 94(2) Cl t-Rh-Cl t 91(1) Cl.-Rh-Cl + b t (trans) 173(3) Rh-Cl-Rh 82(1) a C l t , terminal C l ; C l b > bridging C l . bAveraged over a l l bond angles of a common type. cNumber in parentheses is rmsd in the last figure quoted. 211 235 A similar distortion was found in the case of Cs^Cr^Clg for which, the trans-Cl b-Cr-Cl t angle was 176c 8.3.11. Possible Structure for the [Rh 2Cl g(DMA) 2] 2" Ion Molecular models show that pseudo-octahedral symmetry around each Rh can be achieved (Figure 61) i f Cl atoms occupy positions about the metal so as to form a square planar dimeric "Rh^Clg" skeleton. Trans co-ordination of DMA at the axial position above or below the metal plane results in blocking the vacant co-ordination site of the neighboring Rh by the N-methyl group of DMA. This arrangement provides the closest possible ap^'oarh contact between the Rh and N-methyl group. 112-114 Trans square planar structures have been proposed for RhCl 2[PBu 2R] 2, RhCl 2{P(o-tolyl) 3> 2 and [RhBr 2(siphos) 2] . Vacant octa-hedral sites above and below the plane of the complex are blocked from attack by a t-butyl or methyl group. The apparent st a b i l i t y of these Rh** complexes therefore can be attributed to a kinetic effect i.e. blocking the vacant co-ordination sites, thereby hindering the approach of reagents. Snimilar explanations have been offered for the s t a b i l i t y of planar ortho-substituted aryl derivatives of Ni(II) , Co(II) , and F e ( I I ) 2 4 6 and complexes of theUype 2 4 7 [RhBr(.l-naphthyiyPR3)21 • When 112 RhCl 3-3^0 is reacted with less bulky tertiary phosphines such as PEt 3, PMe2Ph, PEt 2Ph, PPrlJPh, PBu2Ph, PBu3, PBi^Pr!}, which are less capable of blocking the vacant co-ordination sites, there is no evidence for Rh** complex formation. 212 213 8.3.12. Reaction, of [Rh 2Cl 6(DMA) 2] 2~ with HOAc/NaOAc The rhodium(II) acetate dimer is usually prepared by pro-longed refluxing (ca. 18 hours) of Rh(0H)3«H20 or [NH413[RhClg] with 93 95 96 acetic acid either in glacial acetic acid or ethanol ' . Reaction is extremely slow;as is commonly found for substitution inert Rh(III). The fact that [Ph^As]2[Rh2Clg(DMA)2] can be transformed to the acetate dimer under much milder conditions would be consistent with the metal already being in a reduced state, although the nature of the reduction step starting from Rh*1* compounds has not been elucidated. A similarly mild reaction has been observed between [Rh 2(H 20)g(H 2 C O 2 ] 4 + and NaOAc to give rhodium(II) acetate**^. 8.3.13. Oxidation of [Rh 2Cl g(DMA) 2] 2~ in DMA The reaction between solutions of [Rh2Clg(DMA)2] and atmos-pheric oxygen may proceed via a path involving cleavage of chloride bridges followed by oxidation of the monomeric Rh 1 1 units to Rh 1 1 1 3_ species. The final spectrum resembles that for the [Rh 2Cl g] ion (Figure 56). Addition of excess 01" possibly decreases the concentra-tion of monomer, with a resulting retardation in the oxidation rate. A more detailed investigation is required in order to completely elucidate the oxidation mechanism. 214 Chapter IX GENERAL CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK 9.1. Formation of Rh H I(CO) and Rh H IH~ RhBr^HgO in 0.5M HBr has been found to undergo reactions with CO and W^. Rate data and activation parameters suggest that a common mechanistic step may be involved for the i n i t i a l formation of a Rh H I(C0) species (Chapter I I I ) ; Rh 1 1 1 + CO — — R h I H ( C 0 ) (9.1) and for the activation of H2 (Chapter V): TTT 1^ T T T - + Rh 1 1 1 + H2 " Rh H + H (9.2) Kinetic data for reactions of RhCl3«3H20 in 3M HC1 with CO and H2 also 2 6 7 show a close similarity ' ' . Data for a l l these systems are summarized in Table XXXIV. Rate constants are recorded at the experimental temper-atures, as well as at a common temperature, to enable qualitative comparisons between systems. The CO and H2 reactions show a f i r s t order dependence in gas and metal complex with no appreciable dependence on halide. It seems most like l y that these reactions involve replacement of the halide ion by CO or H2 (which f i n a l l y co-ordinates as H~) via an S^ 2 mechanism. A less favourable activation entropy for the H2 reaction as compared with the CO reaction in the respective systems would be consistent with some charge separation in the transition state.(i.e. heterolytic splitting of H9 to H»«'H ) with a higher degree of solvation. Table XXXIV Summary of kinetic data for the reaction of CO and H with some rhodium(III) complexes System •«<1 '•- M"l?.sec_1 AH-i t ur-1 kcal/mole AS^ eu Reference 80° 40° [RhCl 5(H 20)] 2"/C0 a 0.54 0.005C 25.9 13.2 2 [RhCl 5(H 20)] 2"/H 2 a 0.56 0.005C 24.6 9.0 6,7 [RhBr 4(H 20) 2] 7C0 bo 0.14 26.9 23.5 This work [RhBr 4(H 20 ) 2 ]7H 2 J b: 0.17 22.7 10.6 This work a In 3M HCl. b In 0.5M HBr. Estimated from activation parameters. 216 Only limited comparisons can be made between the [RhBr^HgO^l and [RhCl5(H20)] complexes, because of their charge difference, and also since identical conditions of ionic strength were not maintained in the two systems. The presence of solute species such as H+, C l , or Br may affect the properties of the reactants in the activated xomplex. Rate constants at 40° indicate that the bromo complex is far more reac-tive toward CO and W^, than is the chloro complex. An associative mechanism which involves some bond breaking (as well as bond making) in the rate-determining step would generally favour bromo species compared to corresponding chloro species, since Br" forms a weaker bond to Rh than does Cl~. A lower activation entropy for the pentachloroaquorhodate(III) systems, particularly for the CO reaction may reflect the greater charge and smaller size of the activated complex as compared to the bromide system. The factors which affect the reaction rates are, however, not as simple as suggested above. For example, the carbonylation of 2-[RhClgfH^O)] shows a definite acid dependence suggesting involvement 2 of hydroxy species , while the bromide system does not show this effect. Thus, the rate constants are in some instances almost certainly composite constants incorporating acid dissociation constants. The l a b i l i z i n g a b i l i t y of 0H~ is well known and has usually been considered for sub-139 stitution reactions proceeding via a dissociative mechanism 217 9.2. The Autocatalytic Reactions The autocatalytic reduction of RhBr3«2H20 in 0.5M HBr to form [Rh(C0) 2Br 2]~ (Chapter III) involves formation of a mixed valence bridged intermediate, [Rh*"11 • • • Bi Rh*(C0)2]. Reductive carbonylation of RhCl3*3H20 in 3M HCl had been shown earlier to proceed in a similar manner . Data for the autocatalytic reaction (i.e. rate of formation of the bridged species = k2[Rh***][Rh*]) are summarized in Table XXXV. For the halide systems, [RhBr 4(H 20^"]-7[Rh(C0) 2Br 2]and [RhCl 5(H 20)] 2" /[Rh(C0^ 2C1 2]the bromo system was found to benmore reactive by a factor of about 10. For the process Rh***X" + Rh* -v Rh**I---X---Rh* j ^ — Rh* + "XRh***, (9.3) the measured k 2 would refer to step (1) or would be a composite constant comprising^Kk* with K small i.e. either (1) or (2) would be rate-deter-mining. Several factors may thus influence the efficiency of the bridging reaction. If the reaction is simply governed by the rate of nucleophilic attack by Rh***X~ at the reducing agent, the expected order of reactivity for corresponding species would be Cl > Br . However, electrostatic factors would be less favourable for the chloride systems, because of the greater negative charge on the pentachloro species com-pared to the tetrabromo species. Further, the ab i l i t y to transmit electrons and undergo homolytic bond dissociation w i l l be greater for the more polarizable Br" than for C l " ; this is in the order of decreasing Table XXXV Summary of kinetic data for autocatalytic .react.ionssC'. o;">s System k 2 M"1 sec - 1 A H 2 kcal/mole AS^ eu Reference 80° 60° 40° [RhCl 5(H 20)] 2"/[Rh(CO) 2Cl 2]" & 0.066 0.02C 0.005C 12.5 -29.2 2 [Rh(C0)Cl 5 ] 2 7[Rh(C0) 2Cl 2]" 9 0.094 2.5 -56 This work [RhBr 4(Hl 2Q) 2 ]7tRh(C0) 2Br 2]' b is Ct'^c 0.21 0.050 15.1 -16.5 This work [Rh(C0)Br 5l^ /[Rh(C0)2Br2] 550.4/1 .l)d This work a In 3M HCl. b In 0.5M HBr. c Estimated from activation parameters. d Different samples of [Rh(C0)Br^]Z~ 219 strength of the Rh i l i'"X"«Rh 1 bridge. Since the reaction between [RhCl 5(H 20)] 2 - and CO was studied in 3M HC1, while the [RhBr 4(H 20) 2]" system was studied in 0.5M HBr, i t is obvious that the chloride bridge w i l l be cleaved with greater frequency than w i l l the bromide bridge. In fact, the rate constants have been observed to decrease with increas-ing halide concentration. The bridge splitting reaction should severely limit the efficiency of the electron transfer process. The above arguments would then predict lower activity for the chloride system. 2- 2 For the [Rh^[RhCl^(H^)] system , a larger decrease in ac t i -vation entropy was observed than for the corresponding reaction with [RhBr4(H20)2]"and this is the factor responsible for the lower activity. Again, such an effect could be attributed mainly to a greater charge on the activated chloro complex. The possible role of hydroxy species in these reactions also complicates discussion of the overall measured activation parameters. Reductive carbonylation of [Rh(C0)Br 5] 2" in 0.5M HBr (Chapter IV) and oxidation of [Rh(C0) 2Cl 2]" by molecular 0 2 in 3M HC1 (Chapter VI) demonstrate autocatalytic behaviour with probable formation of bridged species of the type [(C0)RhIII---X---RhI(C0)2] (X = Br, Cl). The pentahalocarbonylrhodates are somewhat more reactive than the halo-aquorhodate(III) complexes toward [Rh(C0) 2X 2]~, as shown by the rate constants in Table XXXV. The thermodynamic stab i l i t y of the Rh* products after electron transfer and homolytic bond breaking may be greater for the carbonyl species compared to the rhodium(III) halide species which could contribute to faster rates (i.e. electron transfer 220 to the Rh center would be facilitated by a n acceptor CO ligand at that center) in the former case. The reactivity follows the polariza-b i l i t y pattern Br" > Cl" and again the halide concentration of the solution media may play a role. As mentioned, the oxidation of [Rh(C0) 2Cl o]" by molecular 0 2 in 3M HCl (Chapter VI) showed unusual behaviour, which was attributed to autocatalysis. Experiments at high and low concentrations of metal complex confirm a non-first-order dependence in rhodium. The kinetics indicate that oxidation occurs via oxidative addition of 0 2 to [Rh I(C0) 2Cl 2]~ followed by a more efficient sequence involving formation of a mixed valence bridged intermediate, [(C0)Rh***"'Cl"-Rh*(C0)2] . The activation parameters for the autocatalytic process would be expected to be somewhat similar to those obtained for the [RhClg(H20)] / [Rh(C0) 2Cl 2]~ system. The rate constants however showed very l i t t l e change with temperature and gave a low activation energy of ^ 2.5 kcal/ mole. This value should perhaps be treated cautiously in view of the limitations of the analytical method, which may not have adequately con-sidered the i n i t i a l oxidation of [Rh(C0) 2Cl 2]" by 0 2 to [Rh(C0)Cl g] 2" (see Sec. 6.3). 1 g Other workers have investigated these types of systems by f i t t i n g spectrophotometric data to equations such as (9.4), :log [Rh^-lbqjRh 1 1 1,^ Rh k ^ I R h f y g ^ t/2.3 + constant (9.4) which neglect the i n i t i a l formation of the autocatalytic species. A 221 further method of analysis would be to study the variation in rate as a III 2-function of the ratio of oxidant to the reductant (i.e. [Rh (CO)Clgl : [Rh*(CO)2C1g]~) in the region where the autocatalytic reaction dominates and equation (9.4)„is valid. 9.3. Decomposition of [Rh** 1(C0)X g] 2" (X = Br, Cl) The decomposition of Rh***(C0) complexes by water H20 (OH") Rh I H(C0) ^ Rh1 + C02 + 2H+ (9.5) with the liberation of C02 to form reactive Rh* species has been demon-strated for [Rh(C0)Br 5] 2" (Chapter IV) and [Rh(C0)Cl 5] 2" (Chapter VI). Subsequent reaction of Rh* depends on the nature of the reactants in solution. In the presence of an oxidant, Fe*** or 0 2, oxidation to Rh*** p results with reduction of the substrate. For the [Rh(C0)BrJ /CO system (Chapter IV), i t is l i k e l y that Rh* is involved in the autocata-l y t i c reduction of [Rh(C0)Br 5] 2" to [Rh(C0) 2Br 2]" via formation of a bridged intermediate. The a b i l i t y of Rh* to catalytically reduce Fe1*'* has been observed during the decomposition of [Rh(C0)Br,_] by water in the presence of CO (Chapter IV). The non-complementary electron transfer reaction between Rh* and Fe*** could take place through a [Rh*..«Br"'Fe***] bridged intermediate, or presumably by an outer-sphere mechanism involv-ing a transient Rh** species and two successive one-electron transfers. 222 Activation of the abundant and cheap gas carbon monoxide may have applications in extractive metallurgy for the reduction of specific metal ions. The study of the [Rh(C0)Cl 5] 2" system indicates that OH" rather than H^ O may be involved in the reduction of Rh***(C0). A com-parison of the rate constants at 60°, Table XXXVI, shows that the penta-bromocarbonylrhodate(III) complex is reduced at a rate 20 times faster than the chloro complex. Whether reduction requires prior co-ordination of H20 or (OH") is uncertain. 9.4. Autoxidation of DMA The autoxidation of DMA using [RI^CO^Cl" is not well under-stood, although i t may be related to the [Rh(CgH^)2Cl] ^/O^ system in 9 191 LiCl/DMA ' . The ab i l i t y of radical scavengers to inhibit the reaction demonstrates the involvement of free radicals, Ir spectra clearly indicate the production of hydroperoxides in solution. Both peroxy radicals and molecular 0^ may act as oxidants in thds system, and different oxidation products for CO (CO^  and CO^ ) have been identified. Esr experiments suggest that two different paramagnetic species are formed during the course of the reaction. The f i r s t , has tentatively been assigned to Rh 0^- (s.(analogous to that found in the [Rh(CgH^)2Cl]g/Og system); the second, may be a decomposition product of the Rh** superoxide and has been attributed to a Rh** complex. The Rh** species was isolated (Chapter VIII) as [Ph^Aslg [Rh?Clf.(DMA)9] . A study of the catalytic behaviour of this compound 223 Table XXXVI Summary of kinetic data for the decomposition of Rh I H(C0) species at 60° System sec k 1 [Rh(C0)Cl g] 2 i a 8.2 x 10"5 [Rh(C0)Br 5] 2" b 1.5 x IO"3 a In 3M HC1. b In 0.5M HBr. 224 (i.e. as a hydrogenation catalyst) would give further insight into the catalytic activity of rhodium species in different oxidation states. Rh** species have often been postulatedras intermediates for redox processes. The decarbonylation of DMA by oxygenated solutions of [Rh(C0) 2Cl 2]" (containing Rh**) is worthy of further study. These solu-tions probably contain a mixture of a chlorocarbonylrhodate(I) species (detectable by i r ) and some other chlororhodate species. A chloride -bridged Pt** carbonyl complex 248 is known and may be related to the carbonyl species in the rhodium P case, since both metals have a d eJec ' tnonicGConfaquratn ion. 9.5. Complexes of the Type [Rh 2X g] 3~ (X = Br, Cl) Dimeric species such as [Rh2Xg] (X = C l , Br) have been 149 220 known ' for about 50 years, yet these compounds have only recently been "rediscovered" and characterized using modern instrumental techniques. During the course of the present studies both [Cs] 3[Rh 2Br g] 225 (Chapter IV) and [Ph 4As] 3[Rh 2Cl g] (Chapter VIII) have been prepared, and the latter compound has been f u l l y characterized. Such binuclear anionic complexes may play a role in reactions usually attributed•t to simple halide complexes, particularly in non-3_ aqueous solvents. [Rh^Clg] may be a useful starting material for the preparation of some substituted Rh*** compounds. Cleavage of the chloride bridges in the dimer would produce a co-ordinatively unsaturated Rh*** species which could react with donor ligands. 3_ The dimeric structures proposed for both [RhgClg] and '2-[RhgClgtDMA)^] may be of some interest. Polarographic studies in DMA have been initiated in order to determine whether reduction of 3- II 221 [RhgClg] proceeds through a stable Rh dimer . The reduction seems to be complicated by Cl" ion; quite different results have been obtained at high Cl" than at low Cl". A Cl" effect has also been observed during the esr studies of Rh** (Chapter VII) as well as for the oxidation of Rh** to Rh* 1 1 (Chapter VIII). In LiCl/DMA, chloride-bridged species such as [RhfCgH^J^Cl] 2 and [Rh(C0) 2Cl] 2 have been found to be totally dissociated into the 191 monomeric species . The opposite effect has been observed for 3_ RhCl 3•3^0, for which the formation of the dimer, [Rh^Clg] , is favoured even in the presence of excess Cl" . 226 REFERENCES 1. B.R. James and G.L. Rempel, Chem. Comm., 158 (1967). 2. B.R. James and G.L. Rempel, J. Chem. Soc. (A), 78 (1969). 3. B.R. James, G.L. ."Rempel and (in. part) ^ . 1 . 1 . Ng,^ J:. Chem. Soc. (A), 2454 (1969). 4. D. Forster, Inorg. Chem., 8, 2556 (1969). 5. M.J. Cleare and W.P. G r i f f i t h , J. Chem. Soc. (A), 372 (1969). 6. J.F. Harrod and J. Halpern, Can. J. Chem., 3_7, 1933 (1959). 7. B.R. James and G.L. Rempel, Can. J. Chem., 44, 233 (1966). 8. B.R. James and F.T.T. Ng, Chem. Comm., 908 (1970). 9. B.R. James, F.T.T. Ng,and Ei. Ochiai, Can. J. Chem., 50, 590 (1972). 10. B.R. James and E. Ochiai, Can. J. Chem., 49, 975 (1971). 11. R.D. Gillard, J.A. Osborn, and G. Wilkinson, J. Chem. Soc, 1951 (1965). 12. J.V. Rund, F. Basolo, and R.G. Pearson, Inorg. Chem., 3, 658 (1964). 13. D.J. Baker and R.D. Gillard, Chem. Comm., 520 (1967). 14. R.J. Cozens, K.S. Murray, and B.Q. West, Chem. Comm., 1260 (1970). 15. R.D. Gillard, B. T. Heaton, and„D7Ht.^y1aughans;J,.;£hem:. Soc. (A), 18*0": (1971). 16. J.D. Miller and F.D. Oliver, J. Chem. Soc. Dalton, 2469 (1972). 17. J.D. Miller and F.D. Oliver, J. Chem. Soc. Dalton, 2473 (1972). 18. J.V. Rund, Inorg. Chem., 7, 24 (1968). 19. M.I. Bruce, B.L. Goodall, M.Z. Iqbal, and F.G.A. Stone, Chem. Comm., 661 (1971). 20. A.R.M. Craik, G.R. Knox, P.L. Pauson, R.J. Hoare, and O.S. M i l l s , Chem. Comm., 168 (1971). 227 21. R.L. Bennett, M.I. Bruce, B.L. Gooda11, M.Z. Iqbal and F.G.A. Stone, J. Chem. Soc. Dal ton, 1787 (1972). 22. B.R. James, M. Kastner and G.L. Rempel, Can. J. Chem., 47.. 349 (1969). 23. B.R. James and M. Kastner, unpublished results. 24. R.D. Gillard, HiDiscs Faraday Soc, 46, 90 (1968). 25. R.D. Gillard, B.T. Heaton, and D.H. Vaughan, Chem. Comm., 974 (1969). 26. W.P. G r i f f i t h , "The Chemistry of the Rarer Platinum Metals' (Os, Ru, Ir and Rh)", Interscience Publishers, New York, N.Y., 1967, pp/,372, 384, 401. 27. L.D. Roll man, Inorg. Chim. Acta, 6_, 137 (1972) and references therein. 28. F.E. Paulik, and J.F. Roth, Chem. Comm., 1578 (1968). 29. J.F. Roth, J.H. Craddock, A. Hershman, and F.E. Paulik, Chem. Tech., 1, 600 (1971). 30. P. Chi n i , S. Martinengo, and G. Garlaschelli, Chem. Comm., 709 (1972). 31. L. Cassar and J. Halpern, Chem. Comm., 1082 (1970). 32. L. Cassar, P.E. Eaton, and J. Halpern, J. Amer. Chem. Soc, 92_, 3515 (1970). 33. W. Hieber and H. Lagally, Z. anorg. Chem., 25J_, 98 (1943). 34. D.N. Lawson and G. Wilkinson, J. Chem. Soc, 1900 (1965). 35. J. A. McCleverty and G. Wilkinson, Inorg. Syn., 8, 211 (1966). 36. L.F. Dahl, C. Martell, and D.S. Wampler, J. Amer. Chem. Soc, 83, 1761 (1961). 228 37. B.R. James, Coord. Chem. Rev., 1_, 505 (1966). 38. L.M. Vallarino, Inorg. Chem., 4, 161 (1965). 39. C.K. Thomas and J.A. Stanko, Inorg. Chem., ljO, 566 (1971). 40. D. Forster, Inorg. Chem., VU 1686 (1972). 41. D. Forster, Inorg. Chem., 11_, 473 (1972). 42. D.J. Dahm and D. Forster, Inorg. Nucl. Chem. Lett., 6, 15 (1970). 43. M.J. Cleare and W.P. G r i f f i t h , J. Chem. Soc. (A), 2788 (1970). 44. R. Col ton, R.H. Farthing, and J.E. Knapp, Austral. J. Chem., 23_, 1351 (1970). 45. D. Forster and G. Beck, Chem. Comm., 1072 (1971). 46. B.R. James, "Homogeneous Hydrogenation", John Wiley & Sons Inc., New York, 1973. 47. J. Halpern, Adv. in Chem. Series, Vol. 70, Amer. Chem. Soc, Washington, D.C, 1968, p. 1. 48. J. Halpern, Ann. Rev. Phys. Chem., 1_6, 103 (1965). 49. \ J. Halpern, Proc. 3rd Intern. Congr. Catalysiis, Amsterdam, 1964, North Holland, Amsterdam, 1, 146 (1965). 50. J. Halpern, Chem. Eng. News, 44, 68 (1966). 51. J. Halpern, J. Harrod,and B.R. James, J. Amer. Chem. Soc, 88_, 5150 (1966). 52. J. Halpern, Disc Faraday Soc, 46, 7 (1968). 53. R.G.S. Banks and J.M. Pratt, J. Chem. Soc. (A), 854 (1968) and references therein. 54. L. Vaska, Acc. Chem.Res., 1_, 335 (1968), and references therein. 55. M. Calvin and W.K. Wilmarth, J. Amer. Chem. Soc, 78, 1301 (1956). 229 56. J. Halpern and B.R. James, Can. J. Chem., 44, 671 (1966). 57. J.F. Harrod, S. Ciccone, and J. Halpern, Can. J. Chem., 39_, 1372 (1961). 58. Ref. 46, pp. 260 and 301. 59. V.J. Choy and C.J.0'Connor, Coord. Chem. Rev., 9, 145 (1972/3). and references therein. 60. J.S. Valentine, Chem. Rev., 73_» 235 (1973) and references therein. 61. J.A. McGinnety, R.J. Doedens, and J.A. Ibers, Inorg. Chem., 6_, 2243 (1967). 62. J. Chatt and L.A. Duncanson, J. Chem. Soc, 2939 (1953). 63. M.J.S. Dewar, Bull. Soc. Chim. Fr., 1_8, C71 (1951). 64. 6. Wilke, H. Schott, and P. Heimback, Angew. Chem. Internat. Ed., 6, 92 (1967). 65. B.W. Graham, K.R. Laing, C.J. O'Connor, and W.R. Roper, Chem. Comm., 1272 (1970). 66. B.W. Graham, K.R. Laing, C.J. O'Connor and W.R. Roper, J. Chem. Soc. Dalton, 1237 (1972). 67. S. Otsuka, A. Nakamura, and Y. Tatsuno, J. Amer. Chem. Soc, 91, 6994 (1969). 68. K.R. Laing and W.R. Roper, Chem. Comm., 1568 (1968). 69. W.O. Siegl, S.J. Lapporte, and J.P. Collman, Inorg. Chem., 10_, 2158 (1971). 70. J.A. Stanko, G. Petrov, and C.K. Thomas, Chem. Comm., 1100 (1969). 71. J.P. Collman, M. Kubota, and J.W. Hosking, J. Amer. Chem. Soc, 89, 4809 (1967). 230 72. R.W. Horn, E. Weissberger, and J.P. Collman, Inorg. Chem., 9_, 2367 (1970). 73. J.J. Levison and S.D. Robinson, J. Chem. Soc. (A), 762 (1971). 74. J. Valentine, D. Valentine, Jr., and J.P. Collman, Inorg. Chem., l f j , 219 (1971). 75. N.M. Emanuel, Z.K. Maizus, and I.P. Skibida, Angew. Chem., Intern. Ed., 8, 97 (1969). 76. V.P. Kurkov, J.Z. Pasky, and J.B. Lavigne, J. Amer. Chem. Soc, 90_, 4743 (1968). 77. L.W. Fine, M. Grayson, and V.H. Suggs, J. Organometal. Chem., 22_, 219 (1970). 78. A. Fusi. R. Ugo, F. Fox, A. Pasini, and S. Cenini, J. Organomet. Chem., 26_, 417 (1971) and references therein. 79. R.A. Sheldon, Chem. Comm., 788 (1971). 80. E.W. Stern, Chem. Comm., 736 (1970). 81. J.P. Birk, J.H. Halpern, and A.L. Pickard, J. Amer. Chem. Soc, 90_, 4491 (1968). 82. J. Halpern and A.L. Pickard, Inorg. Chem., 9, 2798 (1970). 83. J.P. Collman, Acc. Chem. Res., 1, 136 (1968). 84. C. Dudley and G. Reed, Tetrahedron Lett., 52,, 5273 (1972). 85. F. Haber and J. Weiss, Naturwissenschaften, 20> 948 (1932). 86. A.J. Chalk and J.F. Smith, Trans. Faraday Soc, 1214 (1957). 87. T E.„.0chiai,. ..Tetrahedron^,2p_5 1819.(1964). 88. Ref. 26, pp. 328, 345 and 364. 89. M.A. Porai-Koshits, A.S. Antsyshkima, Dokl. Akad. Nauk SSSR, 146, 1102 (1962). 231 90. F.A. Cotton, B.G. Deboer, M.D. Laprade, J.R. Pipal , and D.A. Ucko, J. Amer. Chem. Soc, 92_, 2926 (1970). 91. J.N. Van Niekerk and F.R.L. Schoening, Acta Cryst., 6, 227 (1953). 92. I.I. Chernyaev, E.V. Schenderetskaya, A. Nazarova, Abstracts of Papers Presented at 7th International Conference on Coordination Chemistry, Stockholm, 1962. 93. L.A. Nazarova, I.I. Chernyaev, xind A.S .jMorozova.,_RussJ . Inorg. Ohem.,-11:, .1387 (1966). 94. L.A. Nazarova, I.I. Chernyaev, and A.S.^M.orozova.,jRuss.. J_. Inorg. Chem.^lTJ, 29LV (1965). 95. S.A. Johnson, H.R. Hunt, and H.M. Neumann, Inorg. Chem., 2_, 960 (1963). 96. T.A. Stephenson, S.M. Morehouse, A.R. Powell, J.P. Heffer, and G. Wilkinson, J. Chem. Soc, 3632 (1965). 97. V.I. Bel ova and Z.S. Dergacheva, Russ. J. Inorg. Chem., 16_, 1626 (1971). 98. L. Dubicki and R.L. Martin, Inorg. Chem., 9, 673 (1970). 99. I.I. Chernyaev, E.V. Schenderetskaya, and A.A. Koryagina, Russ. J. .".no- Inorg. Chem., 5_, 559 (1960). 100. I.I. Chernyaev. E.V. Schenderetskaya, A.G. Maiorova, and A.A. Koryagina, Russ. J. Inorg. Chem., TTJ, 290 (1965). 101. I.I. Chernyaev, E.V. Schenderetskaya, A.G. Maiorova, and A.A. Koryagina, Russ. J. Inorg. Chem., 11_, 1383 (1966). 102. S. Cenini, R. Ugo, and F. Bonati, Inorg. Ch-i-m. Acta, 1, 443 (1967). 232 103. S.A. Shchepinov, E.N. Salnikova, and M.L. Khidekel, Izv. Akad. Nauk SSSR, Ser. Khim. , 2128 (1967). 104. J. Halpern, E. Kimura, J. Molin-Case, and CS. Wong, Chem. Comm., 1207 (1971). 105. K.G. Caul ton and F.A. Cotton, J. Amer. Chem. Soc, 91_-, 6518 (1969). 106. F.A. Cotton and J.G. Norman J r . , J. Amer. Chem. Soc, 93_, 80 (1971). 107. J.W. Dart, M.K. Lloyd, R. Mason, and J.A. McCleverty, J. Chem. Soc. Dalton, 2046 (1973). 108. B.F.G. Johnson, J. Lewis, and P.W. Robinson, J. Chem. Soc (A), 100 (1970). 109. P. Legzdins, R.W. Mitchell, G.L. Rempel, J.D. Ruddick, and G.Wilkinson, J. Chem. Soc (A), 3322 (1970). 110. F. Maspero and H. Taube, J. Amer. Chem. Soc, 90, 7361 (1968). 111. C. Masters, W.S. McDonald, G. Raper, and B.L. Shaw, Chem. Comm., 210 (1971). 112. C. Masters and B.L. Shaw, J. Chem. Soc. (A), 3679 (1971). 113. M.A. Bennett and P.A. Longstaff, J. Amer. Chem. Soc, 9J_, 6266 (1969). 114. A.T.T. Hsieh, J.D. Ruddick, and G. Wilkinson, J. Chem. Soc. Dalton, 1966 (1972). 115. E. B i l l i g , S.I. Shupack, J.H. Waters, R. Williams, and H.B. Gray, J. Amer. Chem. Soc, 86, 926 (1964). 116. A.H. Maki, N. Edelstein, A. Davidson, and R.H. Holm, J. Amer. Chem. Soc, 86, 4580 (1964). 233 117. E.O. Fischer and H.H. Lindner, J. Organomet. Chem., ]_, 307 (1964). 118. E.O. Fischer and H. Wawerisk, J. Organomet. Chem., 5_, 559 (1964). 119. B.R. James and D.V. Stynes, J. Amer. Chem. Soc, 94, 6225 (1972). 120. R.D. Gillard, B.T. Heaton, and D.H. Vaughan, J. Chem. Soc (A), 734 (1971). 121. J. Gulens and F.C. Anson, Inorg. Chem., J2, 2568 (1973). 122. Ref. 45, footnotes. 123. D. Cozzi and F. Pantani, J. Electroanal. Chem., 2_, 72 (1961). 124. J.J. Byerley and E. Peters, Symp. Unit Processes Hydromet.;Met. Soc. Conf. fVol. 24, Gordon and Breach, New York, 1964. 125. J.J. Byerley and E. Peters, Can. J. Chem., 47, 313 (1969). 126. A. Seidell, "Solubilities of Inorganic and Metal Organic Compounds", 4th ed., Vol. 1, Van Nostrand, New York, 1958, p. 453. 127. D. Forster, Inorg. Nucl. Chem. Lett., 5, 433 (1969). 128. F. Basolo and R.G. Pearson, "Mechanisms of Inorganic Reactions", 2nd ed., Wiley, New York, 1967, p. 579. 129. J. Halpern and A.C. Harkness, J. Amer. Chem. Soc, 83, 1258 (1961). 130. J. Halpern and S. Nakamura, J. Amer. Chem. Soc, 83, 4102 (1961). 131. J. Halpern and S.F.A. Kettle, Chem. and Ind., 668 (1961). 132. J.K. Nicholson, J. Powell and B.L. Shaw, Chem. Comm., 174 (1966). 133. J. Powell and B.L. Shaw, J. Chem. Soc. (A), 583 (1968). 134. R.F. Heck, "Mechanisms of Inorganic Reactions", Adv. in Chem. Ser., Vol. 49, Amer. Chem. Soc, Washington, D.C, 1965, p. 181. 135. K. Noack and F. Calderazzo, J. Organomet. Chem., 10_, 101 (1967). 136. K. Swaminathan and G.M. Harris, J. Amer. Chem. Soc, 88, 4411 (1966). 234 137. C.K. Jorgensen, Acta Chem. Scand., JJ_, 151 (1957). 138. F. Basolo, M.L. Morris, and R.G. Pearson, Disc. Faraday Soc, 29, 80 (1960). 139. Ref. 128, p. 170. 140. J. Halpern, Quart. Rev., 15, 207 (1961). 141. Ref. 128, p. 496. 142. A.A. Frost and R.G. Pearson, "Kinetics and Mechanism", 2nd ed., John Wiley and Sons, New York, 1963, p. 143. 143. A.I. Vogel, "A Textbook of Quantitative Inorganic Analysis", 3rd ed., Longmans, Toronto, 1962, p. 787. 144. H. Irving, M.J. Cabell, and D.H. Mellor, J. Chem. Soc, 3417 (1953). 145. P. M. Henry, J. Amer. Chem. Soc, 86, 3246 (1964). 146. J. Halpern, B.R. James, and A.L.W. Kemp, J. Amer. Chem. Soc, 83_, 4097 (1961). 147. B.R. James and G.L. Rempel, J. Amer. Chem. Soc, 91_, 863 (1969). 148. J.0. Edwards, "Inorganic Reaction Mechanisms", W.A. Benjamin Inc., New York, 1965, p. 128. 149. P. Poulenc, Ann. Chim., 4 (11) 567 (1935). 150. B.R. James and G.L. Rempel, Disc Faraday Soc, 46_, 48 (1968). 151. R. Wiebe and V.L. Gaddy, J. Amer. Chem. Soc, 56, 76 (1934). 152. Ref. 126, Vol. 1_, 3rd ed., 1940, p. 557. 153. J. Halpern, Adv. in Catalysis, 11_, 301 (1959). 154. W.C. Wolsey, CA. Reynolds, and J. Kleinberg, Inorg. Chem., 2_, 463 (1963). 155. Ref. 46, p. 334. 235 156. R.D. Gillard, B.T. Heaton, and D.H. Vaughan, J. Chem. Soc. (A), 3126 (1970). 157. K. Thomas, J.A. Osborn, A.R. Powell,and G. Wilkinson, J. Chem. Soc. M ) , 1801 (1968). 158. C.K. Thomas and J.A. Stanko, J. Coord. Chem., 2, 211 (1973). 159. Ref. 126, Vol. 2, p. 1228. 160. C.J. Nyman, CE. Wymore, and G. Wilkinson, Chem. Comm., 407 (1967). 161. P.J. Hayward, D.M. Blake, C J . Nyman, and G. Wilkinson, Chem. Comm., 987 (1969). 162. T.E. Nappier, Jr . , D.W. Meek, R.M. Kirchner, and J.A. Ibers, J. Amer. Chem. Soc, 95_, 4194 (1973). 163. L.E. Bennett, Prog. Inorg. Chem., 1_8, 25 (1973). 164. J. Chatt and B.L. Shaw, J. Chem. Soc. (A), 1437 (1966). 165. L. Vaska and L.S. Chen, Chem. Comm., 1080 (1971). 166. L. Vaska, L.S..Chen, and W.V. Mill e r , J. Amer. Chem. Soc, 93_, 6671 (1971). 167. W.R. Roper and J.P. Collman, Adv. Organomet. Chem., 10_, 566 (1971). 168. J.A. Osborn, F.H. Jardine, J.F. Young, and G. Wilkinson, J. Chem. Soc. (A), 1711 (1966). 169. B.R. James,aR(JS& .McMiWlianr^and E. Ochiai, Inorg. Nucl. Chem. Lett., 8, 239 (1972). 170. W.E. Bull, S.K. Madan, and J.E. W i l l i s , Inorg. Chem., 2, 303 (1963). 171. A.J. Carty, Can. J. Chem., 44, 1881 (1966). 172. R.S. Drago, R.L. Carlson, and K.F. Purcell, Inorg. Chem., 4, 15 (1965). 236 173. R.S. Drago, D.W. Meek, M.D. Joesten, and L. La Roche, Inorg. Chem., 2, 124 (1963). 174. S.K. Madan and A.M. Donahue, J. Inorg. Nucl. Chem., 28_, 1617 (1966). 175. D. Cook, Can. J. Chem., 42, 2721 (1964). 176. V. Gutmann, Coord. Chem. Rev., 2, 239 (1967). 177. C.W. Garland and J.R. Wilt, J. Chem. Phys., 36, 1094 (1962). 178. A.C. Yang and C.W. Garland, J. Phys. Chem., 6J_, 1504 (1957). 179. F.A. Miller and CH. Wilkins, Anal. Chem., 24, 1253 (1952). 180. A.I. Vogel, "Macro and Semimn.cro Qualitative Inorganic Analysis", Longmans, Toronto, 1960, p. 614. 181. Ref. 180, p. 327. 182. J.S. Forrester and G.H. Ayres, J. Phys. Chem., 63, 1979 (1959). 183. K. Nakamoto, "Infrared Spectra of Inorganic Co-ordination Compounds", 2nd ed., Wiley Interscience, New York, 1970, p. 170. 184. R. Ugo, F. Bonati, and M. Fiore, Inorg. Chim. Acta, 2_, 463 (1968). 185. W. Hieber, HisHeusiijiger.and 0. Vohler, Chem. Ber., 90, 2425 (1957). 186. R.A. Work III and M.L. Good, Inorg. Chem., 9, 956 (1970). 187. B.F. Sagar, J. Chem. Soc. (B), 428 (1967). 188. W.H. Sharkey and W.E. Mochel, J. Amer. Chem. Soc, 8J_, 3000 (1959). 189. M.V. Lock and B.F. Sagar, J. Chem. Soc. (B), 690 (1966). 190. M.V. Lock and B.F. Sagar, J. Chem. Soc (B), 1047 (1967). 191. F.T.T. Ng., Ph.D. Thesis, University of British Columbia, 1970. 192. K. Buhler, W. Diesch, and A.K. Ghosch, Can. J. Chem., 52_, 1700 (1974). 237 193. J.K. Kochi, B.M. Graybill, and M. Kurz, J . Amer. Chem. Soc, 86, 5257 (1964). 194. R.R. Schrock and J.A. Osborn, J . Amer. Chem. Soc, 93, 2397 (1971). 195. J.Y. Chen and J . Halpern, J . Amer. Chem. Soc, 9_3, 4939. 196. V. Schurig and E. Gil-Av, Chem. Comm., 650 (1971). 197. CH. Langford and H.B. Gray, "Ligand Substitution Processes", W.A. Benjamin, New York, 1965, pbapi8.2. 198. L. Cattalini, Progress in Inorg. Chem., 12^ , 263 (1970). 199. Ref. 128, p. 375. 200. L. Cattalini. A. Orio, R. Ugo, and F. Bonati, Chem. Comm., 48 (1967) . 201. L. Cattalini, R. Ugo, and A. Orio, J . Amer. Chem. Soc, 90_, 4800 (1968) . 202. Ref. 128, p. 639. 203. A.T. Betts and N. Uri, Adv. in Chem. Ser., 76, Amer. Chem. Soc, Washington, D.C, p. 160. 204. J . Roda and F. Hrabak, Die Makromolekulare Chemie, 168, 93 (1973). 205. B.L. Booth and R.N. Haszeldine, Chem. Comm., 1074 (1972). 206. R. Ugo, F. Conti, S. Cenini, R. Mason, and G.B. Robertson, Chem. Comm., 1498 (1968). 207. B.R. Flynn and L. Vaska, J . Amer. Chem. Soc, 95, 5081 (1973). 208. Ref. 128, p. 404. 209. J.E. Lyons.and IJ.O. Turner, J . :0rg,...Chem2^?3Z?,':2881 (^1972). 210. B.M. Hoffman, D.L. Diemente, and F. Basolo, J . Amer. Chem. Soc, 92, 61 (1970). 238 211. G.N. Schrauzer and L.P. Lee, J. Amer. Chem. Soc, 92, 1551 (1970). 212. K.M. Wang and J.H. Lunsford, J. Phys. Chem., 75, 1165 (1971). 213. P.A. Ki l t y , N.C. Rol, and W.M.l.Sachtler, Proc. 5tb Int.^Congress of Catalysis, Palm Beach, 1972, p. 66. 214. J.H.R. Casemier, B.E. Nieuwenhuys, and W.M. Sachtler, J. Catal., 29, 367 (1973). 215. N. Uri, Nature, 177, 1177 (1956). 216. H. Kropf, Ann. Chem., 63_7, 73 (1960). 217. H.L. Roberts5and.-.W.R: .rSymes, U..i.Ghem:.cSoc. (A), ]450^.(0.968). 218. A.G. Sykes and J.A. Weil, Progress in Inorg. Chem., 12^ , 1 (1970). 219. S.N. Ivanova, L.M. Gindin, and"k.Y;a./M-iroriova.,; Rus_s. J;. Inorg. Chem., VZ, 861 (1967). 220. O.V. Fraenkel, Monatsh, 35, 119 (1914). 221. A.W. Addison, private communication. 222. G.L. Rempel, Ph.D. Thesis, University of British Columbia, 1968. 223. P.C. Crouch, G.W.A. Fowles, and RfcA'. .Waltgo.^ sTChem, Soc. (A), 9725(0969). 224. O.N. Efimov,MMl. .Khidekel, V.A. Avilov, P.S. Chekisi-i? 9 O.N. Eremenko, and A.G. Ovcharenko, J. Gen. Chem. USSR, 38, 2581 (1968). 225. R. Saillant and R.A.D. Wentworth, Inorg. Chem., 1606 (1968). 226. G.W.A. Fowles and B.J. Russ, J. Chem. Soc. (A), 517 (1967). 227. R.A. Walton, Progress in Inorg. Chem., 1_6, 17 (1972). 228. L.E. Orgel, "An Introduction to Transition Metal Chemistry", 2nd ed., John Wiley and Sons, New York, 1966, p. 47. 239 M.G. Townsend and J.W. Orton, J. Chem. Phys. 45, 4135 (1966). R.P.A. Muniz, N.V. Vugman, and J. Danon, J. Chem. Phys., 5_4, 128 (1971). "Landoldt-Bornstein New Series II/2", K.H. H.ellwege Ed., Springer-Verlag, New York, 1966, p. 1-16. "Handbook of Chemistry and Physics", R.C. Weast Ed., 54th ed., The Chemical Rubber Co., Cleveland, Ohio, 1973, pp. E-125, E-130. A. E. Earnshaw, "Introduction to Magnetochemistry", Academic Press, New York, 1968, p. 101. B. R. James and E i . Ochiai, Spectrosc. Lett., 5_, 287 (1972). G. J. Wessel and D.J.W. Ijdo, Acta Cryst., 10., 466 (1967). C. Brosset, Ark. Kemi Min. Geol., 12A, Nr. 4 (1935). L. Pauling, Chem. and Eng. News, 2970 (1947). H. M. Powell and A.F. Wells, J. Chem. Soc, 1008 (1935). J.L. Hoard and L. Goldstein, J. Chem. Phys., 3_, 199 (1935). G.C. Dobinson, R. Mason, and D.R. Russel, Chem. Comm., 62 (1967). G. Bombieri, R. Graziani, C. Panattoni and L. Volponi, Chem. Comm., 977 (1967). J.A. Ibers and R.G. Snyder, Acta Cryst., ]_5, 923 (.1962). K. Klanderman and L.F. Dahl, Dissert. Abstr., 25_, 653 (1965). A. Bright, J.F. Malone, J.K. Nicholson, J. Powell, and B.L. Shaw, Chem. Comm., 712 (1971). J. Chatt and B.L'. Shaw, J. Chem. Soc, 1718 (1960). J. Chatt and B.L. Shaw, J. Chem. Soc, 285 (1961 ). J. Chatt and A.E. Underhill, J. Chem. Soc. 2088 (1963). F.A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry", 3rd ed., Interscience, New York, 1972, p. 707. A t t a c h m e n t PUBL ICAT IONS B . R . J a m e s a n d G . R o s e n b e r g , " T h e A c t i v a t i o n o f C a r b o n M o n o x i d e b y P l a t i n u m M e t a l C o m p l e x e s i n S o l u t i o n " , p r e s e n t e d a t S y m p o s i u m o n " C o -o r d i n a t i o n a n d A c t i v a t i o n o f S m a l l M o l e c u l e s b y T r a n s i t i o n M e t a l s " . A C S / C I C 1972 B u f f a l o . B . R . James a n d G . R o s e n b e r g , " A c t i v a t i o n o f C a r b o n M o n o x i d e and H y d r o g e n by R h ( I I I ) H a l i d e C o m -p l e x e s i n A q u e o u s A c i d S o l u t i o n " , X V I t h I n t e r n . C o n f . C o o r d . C h e m . , D u b l i n , 1 9 7 4 . P a p e r 4 . 3 3 . B . R . James and G . R o s e n b e r g , "Some A s p e c t s o f R h o d i u m ( I I ) C h e m i s t r y : T h e H e x a c h l o r o b i s ( N N - d i m e t h y l a c e t -amide ) d i r h o d a t e ( I I ) and N o n a c h l o r o d i r h o d a t e ( I I I ) A n i o n s " , I n t e r n . C o n f . on S t a b i l i t y and R e a c t i v i t y o f C o o r d i n a t i o n C o m p o u n d s , B r e s s a n o n e , 197 4 ; t o b e p u b l i s h e d i n C o o r d i n . Chem. R e v . / 

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}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061063/manifest

Comment

Related Items