<%BANNER%>

Oxidation of Ethanol

Permanent Link: http://ufdc.ufl.edu/UFE0042814/00001

Material Information

Title: Oxidation of Ethanol Synthesis and Electrochemical Characterization of Heterobimetallic Catalysts
Physical Description: 1 online resource (151 p.)
Language: english
Creator: CORREIA,MARIE C
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ELECTROCATALYSIS -- ETHANOL -- HETEROBIMETALLIC -- HOMOGENEOUS -- METHANOL -- OXIDATION -- PALLADIUM -- PLATINIUM -- RHODIUM -- RUTHENIUM
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This dissertation describes the synthesis and electrochemical oxidation of ethanol for a series of heterobimetallic complexes. Cyclic voltammetry with previously prepared methanol oxidation catalysts of the type CpRu(PPh3)(mu-Cl)(mu-dppm)PtCl2 (1-28), resulted in large current increases upon addition of ethanol indicating these complexes were also active towards ethanol oxidation. Bulk electrolysis experiments indicated the oxidation of ethanol produced the two-electron oxidation product acetaldehyde and the four-electron oxidation product acetic acid. These were also detected in solution as their respective ethanol condensates 1,1-diethoxyethane and ethyl acetate. Complexes 1-28 were incapable of dissociating the C?C bond of ethanol to produce CO2, the twelve-electron oxidation product. New ethanol soluble complexes cis-(bpy)2Ru(mu-I)(mu-dppm)PdCl2PF6 (30), cis-(bpy)2Ru(mu-I)(mu-dppm)PtCl2PF6 (31), eta5-C5H4CH2CH2NH2Ru(PPh3)(mu-Cl)(mu-dppm)PdCl2 (40), and eta5-C5H4CH2CH2NH2?HClRu(PPh3)(mu-Cl)(mu-dppm)PdCl2 (43) were synthesized and electrochemically studied for ethanol oxidation using cyclic voltammetry and bulk electrolysis. Results indicated complexes 30 and 31 were inefficient complexes for this transformation. Complexes 40 and 43 were easily able to dissolve in ethanol and studies indicated formation of the two- and four-electron oxidation products from ethanol oxidation. The syntheses of novel heterobimetallic complexes bearing N-heterocyclic carbene ligands were attempted as a way to stabilize the complexes during oxidation. The ligands could be synthesized in good yield; however, the synthetic schemes utilized to bind the metal precursor to the NHC ligand have been unsuccessful to date.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by MARIE C CORREIA.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: McElwee-White, Lisa A.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042814:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042814/00001

Material Information

Title: Oxidation of Ethanol Synthesis and Electrochemical Characterization of Heterobimetallic Catalysts
Physical Description: 1 online resource (151 p.)
Language: english
Creator: CORREIA,MARIE C
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ELECTROCATALYSIS -- ETHANOL -- HETEROBIMETALLIC -- HOMOGENEOUS -- METHANOL -- OXIDATION -- PALLADIUM -- PLATINIUM -- RHODIUM -- RUTHENIUM
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This dissertation describes the synthesis and electrochemical oxidation of ethanol for a series of heterobimetallic complexes. Cyclic voltammetry with previously prepared methanol oxidation catalysts of the type CpRu(PPh3)(mu-Cl)(mu-dppm)PtCl2 (1-28), resulted in large current increases upon addition of ethanol indicating these complexes were also active towards ethanol oxidation. Bulk electrolysis experiments indicated the oxidation of ethanol produced the two-electron oxidation product acetaldehyde and the four-electron oxidation product acetic acid. These were also detected in solution as their respective ethanol condensates 1,1-diethoxyethane and ethyl acetate. Complexes 1-28 were incapable of dissociating the C?C bond of ethanol to produce CO2, the twelve-electron oxidation product. New ethanol soluble complexes cis-(bpy)2Ru(mu-I)(mu-dppm)PdCl2PF6 (30), cis-(bpy)2Ru(mu-I)(mu-dppm)PtCl2PF6 (31), eta5-C5H4CH2CH2NH2Ru(PPh3)(mu-Cl)(mu-dppm)PdCl2 (40), and eta5-C5H4CH2CH2NH2?HClRu(PPh3)(mu-Cl)(mu-dppm)PdCl2 (43) were synthesized and electrochemically studied for ethanol oxidation using cyclic voltammetry and bulk electrolysis. Results indicated complexes 30 and 31 were inefficient complexes for this transformation. Complexes 40 and 43 were easily able to dissolve in ethanol and studies indicated formation of the two- and four-electron oxidation products from ethanol oxidation. The syntheses of novel heterobimetallic complexes bearing N-heterocyclic carbene ligands were attempted as a way to stabilize the complexes during oxidation. The ligands could be synthesized in good yield; however, the synthetic schemes utilized to bind the metal precursor to the NHC ligand have been unsuccessful to date.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by MARIE C CORREIA.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: McElwee-White, Lisa A.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042814:00001


This item has the following downloads:


Full Text

PAGE 1

1 OXIDATION OF ETHANOL: SYNTHESIS AND ELECTROCHEMICAL CHARACTERIZATION OF HETEROBIMETALLIC CATALYSTS By MARIE C. CORREIA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Marie C. Correia

PAGE 3

3 To my parents because they believed

PAGE 4

4 ACKNOWLEDGMENTS I thank my Advisor, Professor Lisa McElwee White for providing me with the opportunity to learn about the field of organometallic catalysis and for allowing me to spend my graduate career in her group. I would als o like to thank her for her patience guidance and support throughout the years as well as her tireless efforts when it mattered most I acknowledge my committee members for their time and involvement: Dr. Daniel R. Talham, Dr. Michael J. Scott, Dr. Helena Hagelin Weaver and Dr. Sukwon Hong. I thank the McElwee White group members past and present, for all their help and interesting conversations over the years both of the inane va riety and also about chemistry I am extremely thankful to Dr. Daniel Serra for getting me started and for many helpful discussions about the project. I also acknowledge work done on this project by the following undergraduate REU stude nts: Casie Hilliard Christina Sweeney David Watts and Lucas Parvin. In addition, I acknowledge the staff of the Chemistry department fo r everything They were always patie nt and understanding when faced with my pleas for help and were able to make my difficulties less gloomy with their teasing, random conversations and knowledge ; my research could not have moved forward without them. I am especially thankful for my good f riends ; both new and old. They have provided me with lots of laughter fun, love a nd understanding over the years There are no words to express how grateful I am to have my family; my parents, grandparents, sister and brothers. I am constantly support ed and loved by them. I especially thank my parents who have struggled and sacrificed so much and who

PAGE 5

5 continue to do so in order for me and my siblings to have better lives than they had They have taught us many things through their own actions and beli efs and they continue to be our biggest supporters and never seize to be accepting of us I thank my grandparents for their love, good food and constant belief in us. I am grateful to have my sister and brothers with whom I have fought and laughed and wi th whom I will continue to do so until we all die. I love you all

PAGE 6

6 TABLE OF CONTENTS p age ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIG URES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 13 ABSTRACT ................................ ................................ ................................ ................... 16 CHAPTER 1 LITERATURE REVIEW AND STRATEGIES FOR THE FUTURE .......................... 18 Fuel Cells ................................ ................................ ................................ ................ 18 Oxidation of Al cohols ................................ ................................ .............................. 19 Methanol as Fuel ................................ ................................ .............................. 20 Ethanol as Fuel ................................ ................................ ................................ 22 Types of Catalysts and Catalytic Systems ................................ .............................. 24 Rhodium as a Catalyst ................................ ................................ ..................... 26 2 ELECTROCHEMICAL OXIDATION OF ETHANOL USING HETEROBIMETALLIC CATALYSTS ................................ ................................ ...... 31 Electrochemical Oxidation of Methanol ................................ ................................ ... 32 Electrochemical Oxidation of Ethanol ................................ ................................ ..... 35 Cyclic Voltammetr y with Ethanol ................................ ................................ ...... 35 Bulk Electrolysis with Ethanol ................................ ................................ ........... 38 Bulk electrolysis in dichloroethane using complexes 1 and 4 .................... 39 Bulk electrolysis in ethanol using the alcohol soluble complexes 12 13 and 14 ................................ ................................ ................................ ..... 42 Bulk electrolysis in ethanol using CO complexes 15 20 ............................. 45 Summary ................................ ................................ ................................ ................ 47 3 BIPYRIDYL LIGAND ................................ ....... 49 Background ................................ ................................ ................................ ............. 49 2,2 Bipyridyl Ru/Pd and Ru/Pt Complexes 30 and 31 ................................ ............ 49 Synthesis of Complexes 30 and 31 ................................ ................................ .. 49 Cyclic Voltammograms of Complexes 30 and 31 ................................ ............. 51 Bulk Electrolysis with Complexes 30 and 31 ................................ .................... 53 Attempted Synthesis of Fe Bipyridyl Complexes ................................ .................... 55 Attempt ed Synthesis of Pd Bipyridyl Complexes ................................ .................... 58 Synthesis of Complexes 36 and 37 ................................ ................................ .. 58

PAGE 7

7 Cyclic Voltammograms of Complexes 30 and 31 ................................ ............. 60 Summary ................................ ................................ ................................ ................ 60 4 ETHANOL SOLUBLE COMPLEXES ................................ ................................ ...... 63 Background ................................ ................................ ................................ ............. 63 Functionalized Ethanol Soluble Cyclopentadienyl Ligands ................................ ..... 64 Amino Substituted Cyclopentadienyl Ru/Pd and Ru/Pt Complexes ................. 64 Other attempts at synthesizing complex 40 ................................ ............... 68 Electrochemistry of amino substituted cyclopentadienyl complexes 40 and 43 ................................ ................................ ................................ ..... 70 Attempted Syntheses of Hydroxyl Substituted Cyclopentadienyl Complexes ... 73 Attempted Syntheses of Phosphino Substituted Cyclopentadienyl Complexes ................................ ................................ ................................ .... 79 Summary ................................ ................................ ................................ ................ 82 5 N HETEROCYCLIC CARBENES: STRATEGY TOWARDS MOR E STABLE HETEROBIMETALLIC COMPLEXES ................................ ................................ ..... 84 Background ................................ ................................ ................................ ............. 84 Preliminary Ru/Rh Complex ................................ ................................ .................... 90 Attempted Synthesis of a Ru/Rh Complex bearing an NHC Ligand ....................... 92 Ligand Design and Synthesis ................................ ................................ ........... 92 Attempts to Bind the NHC to Ruthenium ................................ .......................... 96 Attem pt to Bind the NHC to Rhodium ................................ ............................. 101 Electrochemical experiments using the NHC Rh complex 52 .................. 103 Attempt to Bind the NHC to Rhodium and Ruthenium ................................ .... 104 Summary and Perspectives ................................ ................................ .................. 105 6 CONCLUSIONS ................................ ................................ ................................ ... 108 Activity of Ru/Pt Complex 1 for Water Oxidation ................................ ................... 108 Proposed Mechanisms of Alcohol Oxidation with Heterobimetallic Catalysts ....... 108 7 EXPERIMENTAL SECTION ................................ ................................ ................. 118 General Procedures ................................ ................................ .............................. 118 Electrochemical Considerations ................................ ................................ ............ 118 Product Analysis ................................ ................................ ................................ ... 119 Synthetic Procedures ................................ ................................ ............................ 120 APPENDIX: CYCLIC VOLTAMMOGRAMS OF PREVIOUSLY PREPARED HETEROBIMETALLIC COMPLEXES WITH ETHANOL ................................ ....... 134 LIST OF REFERENCES ................................ ................................ ............................. 139 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 151

PAGE 8

8 LIST OF TABLES p age 2 1 Bulk electrolysis data for the oxidation of ethanol by complexes 12, 13 and 14 ................................ ................................ ................................ ....................... 44 2 2 Bulk electrolysis data for the oxidation of ethanol by complexes 15 16 18 and 19 ................................ ................................ ................................ ................ 46 3 1 NMR data for complexes 30 and 31 obtained in CD 3 CN ................................ .... 52 3 2 Formal potentials for complexes 30 and 31 ................................ ........................ 52 3 3 Bulk electrolysis data for the oxidation of ethanol by complexes 30 and 31 ....... 55 4 1 NMR data for complexes 40 and 43 ................................ ................................ ... 67 4 2 Formal potentials for complexes 40 and 43 ................................ ........................ 71 5 1 Attempts to deprotonate the imidazole using a base ................................ .......... 99

PAGE 9

9 LIST OF FIGURES p age 1 1 Schematic representation of a direct alcohol fuel cell ................................ ......... 19 1 2 Methanol oxidation on bulk platinum anode ................................ ....................... 21 1 3 Oxidation of methanol in homogeneous systems ................................ ............... 22 1 4 Ethanol oxidation on bulk platinum anodes ................................ ........................ 23 1 5 Ethanol oxidation in homogeneous systems ................................ ...................... 24 1 6 Scheme showing Rh insertion into a C C bond ................................ ................. 27 1 7 Disso ciation of ethanol on the Rh(111) surface ................................ .................. 28 1 8 Activation of a C C bond ................................ ................................ .................... 29 1 9 Activation of a C C bond at room temperature ................................ ................... 30 2 1 Cyclic voltammograms of 1 ................................ ................................ ................ 33 2 2 First generation catalysts 1 8 ................................ ................................ .............. 36 2 3 Cyclic voltammograms of complex 4 ................................ ................................ .. 36 2 4 Alcohol soluble catalysts 9 14 ................................ ................................ ............ 37 2 5 Ru and Fe carbonyl catalysts 15 20 ................................ ................................ .. 37 2 6 Cyclic voltammograms of complex 19 ................................ ............................... 38 2 7 Ru/Sn catalysts 21 28 ................................ ................................ ........................ 39 2 8 Evolution of the oxidation products from the electrooxidation of ethanol f or the Ru/Pt complex 1 ................................ ................................ ........................... 40 2 9 Evolution of the oxidation products from the electrooxidation of ethanol for the Ru/Pd complex 4 ................................ ................................ .......................... 41 2 10 Evolution of the oxidation products from the electrooxidation of ethanol for the Ru/Pd complex 12 ................................ ................................ ....................... 43 2 1 1 Evolution of the oxidation products from the electrooxidation of ethanol for the Ru/Pd and Ru/Pt complexes 14 and 13 ................................ ....................... 44 3 1 Synthetic scheme to 29 ................................ ................................ ...................... 50

PAGE 10

10 3 2 Preparation of 30 and 31 ................................ ................................ .................... 51 3 3 Alternative synthesis of [ cis (bpy) 2 RuI( 1 dppm)]PF 6 ( 29 ) ................................ .. 51 3 4 Cyclic voltammograms of complex 30 ................................ ............................... 52 3 5 Cyclic volta mmograms of complex 31 ................................ ................................ 53 3 6 Evolution of the oxidation products from the electrooxidation o f ethanol for the Ru/Pd complex 30 ................................ ................................ ........................ 54 3 7 Evolution of the oxidation products from the electrooxidation of ethanol for the Ru/Pt complex 31 ................................ ................................ ......................... 54 3 8 Synthesis of [ cis (bpy) 2 FeCl(PPh 3 )]PF 6 ................................ .............................. 56 3 9 First attempt to synthesize [ cis (bpy) 2 FeCl( 1 dppm)]PF 6 ................................ ... 56 3 10 Second attempt to synthesize [ cis (bpy) 2 FeCl( 1 dppm)]PF 6 ............................. 57 3 11 Attempted synthesis of [bpyFeCl( 1 dppm)]PF 6 ................................ ................. 57 3 12 Synthesis of the Pd bipyridyl starting material ................................ .................... 58 3 13 Attempted syntheses of 36 and 37 ................................ ................................ ..... 59 3 14 Cyclic voltammogram of 36 ................................ ................................ ................ 61 3 15 Cyclic voltammogr am of 37 ................................ ................................ ................ 61 4 1 Synthesis of Ru/Pd 40 ................................ ................................ ........................ 66 4 2 Synthesis of Ru/Pd 43 ................................ ................................ ........................ 67 4 3 Formation of 38 along with attachment of NH 2 to Ru ................................ .......... 68 4 4 Formation of 38 starting from the deprotonated (C 5 H 5 )CH 2 CH 2 NH 2 ligand ........ 69 4 5 Additional attempt at substituting the Cp ring ................................ ..................... 70 4 6 Results obtained from ESI MS data ................................ ................................ ... 71 4 7 Cyclic voltammogram of complex 40 ................................ ................................ .. 72 4 8 Cyclic voltammogram of complex 43 ................................ ................................ .. 72 4 9 Cyclic voltammogram of co mplex 43 ................................ ................................ .. 73 4 10 Synthetic scheme to 44 and 45 ................................ ................................ .......... 74

PAGE 11

11 4 11 Attachment of the dppm ligand as a bidentate chelate ................................ ....... 75 4 12 Synthesis of the (C 5 H 5 )CH 2 CH 2 OH ligand via a different route .......................... 76 4 13 Protection of the 2 cyclopentadienyl ethanol reagent ................................ ......... 77 4 14 Scheme to synthesize 46 and 47 ................................ ................................ ........ 78 4 15 ESI MS data for 47 ................................ ................................ ............................. 79 4 16 Formation of the CpRu(CO) 2 Cl and CpRu(CO) 2 Cl precursors ........................... 80 4 17 Attempted synthesis of (CpPPh 2 )Ru(CO) 2 I ................................ ......................... 80 4 18 Attempted syntheses of (CpPPh 2 )Ru(CO) 2 Cl and (CpPPh 2 )Ru(CO) 2 Cl 3 Pd ....... 81 5 1 Insertion of a carbon atom into an alkene ................................ ........................... 84 5 2 First stable transition metal carbene complexes ................................ ................. 84 5 3 Formatio n of the first stable, free N heterocyclic carbene ................................ .. 85 5 4 Orbital interactions of M NHC bonds ................................ ................................ .. 85 5 5 Examples of well known Ru, Pd and Ir complexes bearing NHC ligands ........... 87 5 6 Examples of NHC complexes ................................ ................................ ............. 87 5 7 Homobimetallic NHC complexes of Ag, and Cu ................................ ................. 88 5 8 Homobimetallic Ru NHC complexes for olefin metathesis reactions .................. 89 5 9 Heterobimetallic NHC complexes ................................ ................................ ....... 89 5 10 Synthesis of Ru/Rh complex 51 ................................ ................................ .......... 90 5 11 Cyclic voltammograms of complex 51 ................................ ................................ 91 5 12 Functionalization of the NHC backbone ................................ ............................. 92 5 13 Design of the NHC ligand ................................ ................................ ................... 93 5 14 Synthesis of the imidazolium ligand ................................ ................................ .... 94 5 15 Wanzlick equilibrium and dimerization of the imidazolium ion during mass spectrometry ................................ ................................ ................................ ....... 96 5 16 Possible bonding modes of Ru and Rh to the NHC ligand and basic MM2 energy minimization of each ................................ ................................ ............... 97

PAGE 12

12 5 17 Attachment of the imidazoli um salt to Rh ................................ ......................... 102 5 18 Fragmentation pattern observed for 52 by LR MS ................................ ............ 102 5 19 Cyclic voltammograms of complex 52 ................................ .............................. 104 5 20 Attempt to b ind the NHC ligand to Rh and Ru consecutively ............................ 105 6 1 Cyclic voltammograms of complex 1 ................................ ................................ 108 6 2 Effect of adding PPh 3 to complex 15 ................................ ................................ 111 6 3 Dissociation of the bridging Cl fr om either Ru or Pt ................................ .......... 112 6 4 Oxidative addition of CH 3 OH to Pt(II) ................................ ............................... 112 6 6 Oxidation of formaldehyde to methyl formate ................................ ................... 114 6 7 Formation of dimethyl ether ................................ ................................ .............. 114 6 8 Oxidation of CH 3 OH on Ru ................................ ................................ ............... 115 6 9 Regeneration of Ru(II) via a Ru(V) intermediate ................................ ............... 115 6 10 Alternate mechanism of CH 3 OH oxidation on Ru ................................ ............. 116

PAGE 13

13 LIST OF ABBREVIATION S AA Acetic a cid a mu Atomic mass unit bimpy Bis(imine)pyridine bpm 2,2' Bipyrimidine bpy 2,2 Bipyridine bpz Bipyrazine CE Current efficiency CI Chemical ionization COD 1,5 Cyclooctadiene COE Cyclooctene Cp Cyclopentadiene CV Cyclic voltammetry Da Dalton DAFC Direct alcohol fuel cell DART Direct analysis in real time d ba Dibenzylideneacetone DCE 1,2 Dichloroethane DEE 1,1 Diethyoxyethane DFT Density functional theory DHP 3,4 Dihydro 2H pyran DIP Direct insertion probe DMF Dimethylformamide DMM Dimethoxymethane DMSO Dimethyl sulfo xide

PAGE 14

14 d ppm 1,1 Bis(diphenylphosphino)methane EA Ethyl acetate ESI Electrospray ionization EtOAc Ethyl acetate EtOH Ethanol FAB Fast atom bombardment FTIR Fourier transform infrared spectroscopy GC Gas chromatography HRMS High resolution mass spectrometry LR MS Low resolution mass spectrometry LSI Liquid secondary ion MeOH Methanol MF Methyl formate MM2 Molecular mechanics 2 MS Mass spectrometry NHC N heterocyclic carbene NHE Normal hydrogen electrode NMR Nuclear magnetic resonance spectroscopy PMEA Polym er Me mbrane Electrode Assembly py Pyridine RHE Reversible hydrogen electrode TBA(BF 4 ) Tetrabutylammonium tetrafluoroborate TBAH Tetrabutylammonium hexafluorophosphate TBAT Tetrabutylammonium trifluoromethanesulfonate THF Tetrahydrofuran

PAGE 15

15 TLC Thin layer chromato graphy TO Turn overs TOF Time of flight Tpy Terpyridine

PAGE 16

16 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy OXIDATION OF ETHANOL: SYNTHESIS AND ELECTROCHEMICAL CHARACTERIZATION OF HETEROBIMETALLIC C ATALYSTS By Marie C. Correia May 2011 Chair: Lisa McElwee White Major: Chemistry This dissertation describes the synthesis and electrochemical oxidation of ethanol for a ser ies of heterobimetallic c omplexes Cyclic voltammetry with p reviously prepared methanol oxidation catalysts of the type CpRu(PPh 3 dppm)PtCl 2 ( 1 28 ) resulted in large current increases upon addition of ethanol indicating these complexes were also active towards ethanol oxidation. Bulk electrolysis experiments indicated the oxidation of ethanol produced the two electron oxidation pro duct acetaldehyde and the four electron oxidation product acetic acid. These were also detected in solution as thei r respective ethanol condensates 1,1 diethoxyethane and ethyl acetate. C omplexes 1 28 were incapable of dissociating the C C bond of ethanol to produce CO 2 the twelve electron oxidation product. New ethanol soluble complexes [ cis (bpy) 2 Ru( I)( dppm)P d Cl 2 ]PF 6 ( 30 ), [ cis (bpy) 2 Ru( I)( dppm)PtCl 2 ]PF 6 ( 31 ), [ 5 C 5 H 4 CH 2 CH 2 NH 2 ]Ru(PPh 3 )( Cl)( dppm)PdCl 2 ( 40 ), and [ 5 C 5 H 4 CH 2 CH 2 NH 2 HCl ]Ru(PPh 3 )( Cl)( dppm)PdCl 2 ( 4 3 ) were synthesized and electrochemically studied for ethanol oxidation using cyclic voltam metry and bulk electrolysis Results indicated complexes 30 and 31 were

PAGE 17

17 inefficient co mplexes for this transformation. Complexes 40 and 43 were easily able to dissolve in ethanol and studies indicated formation of th e two and four electron oxidation pro ducts from ethanol oxidation. The syntheses of n ovel heterobimetallic complexes bearing N heterocyclic carbene ligands were attempted as a way to stabilize the complexes during oxidation. The ligands could be synthesized in good yield ; however, the synt hetic schemes utilized to bind the metal precursor to the NHC ligand have been unsuccessful to date.

PAGE 18

18 CHAPTER 1 LITERATURE REVIEW AN D STRATEGIES FOR THE FUTURE Fuel Cells The fuel cell, discovered in 1839 by the English physicist Sir William Grove, is a n electrochemical conversion device that transforms the energy produced from a chemical reaction into electrical energy. 1,2 first cell consisted of dilute sulfuric acid and two platinum electrodes each encased in closed tubes and surrounded by oxygen and hydrogen, respectively. 3 Al though there are man y different types of fuel cells the basic concepts remain the same Each fuel cell consists of three main components, the cathode, the anode and an electrolyte (Figure 1 1, adapted from Tsiakaras et al. 4 ) At the anode, the fuel (hydrocarbon, hydrogen, natural gas, etc.) is electrochemically oxidized while at the cathode, an oxidant (water, oxygen) is electrochemica lly reduced Electrons, which are produced at the anode, travel around an external circuit to the cathode, while the ions produced from the chemical reactions (H + OH etc.) travel through the electrolyte Typical fuel cells also contain a catalys t which helps these reactions occur and in general, fuel cells should be able to operate continuously as long as the flow of fuel and oxidant into the cell and the flow of products out of the cell are maintained Fuel cells are advantageous because they operate at low temperatures; produce almost no pollution (ideally only CO 2 and H 2 O) and very little heat I n addition the fuel can easily be refilled Research and development on fuel cells has comprised of both engin eering the best cell and also choosing the correct type of fuel The fuel chosen should have several char acteristics that make it viable; some of these are that it should have the ability to be easily stored and transported, it should be safe both for the environment

PAGE 19

19 and for the consumer, it shoul d be readily available and most importantly, should be electrochemically active. 4 Figure 1 1. Schematic representation of a direct alcohol fuel cell ( adapted from Tsiakaras et al. 4 ) Oxidation of Alcohols The price of fuel increases and will continue to increase each year as developing countries demand larger supplies in order to sustain their burgeoning societies. There are many alternative fuel sources that should be explored in order to compensate for the dwindling supply of crude oil such as solar power, wind power and nuclear energy. Alcohols are advantageous sources of fuel because they can be easily transferr ed and stored, are relatively inexpensive, operate at low temperatures and have high theoretical mass energy densities (i.e. the amount of energy stored in a system per unit mass) 2,4 7 There are, however, some di sadvantages for these types of fuel cells ; fuel crossover from the anode to the cathode may be experienced and also the catalysts employed at the anode usually exhibit poor performance. 8 10 These issues can lead t o poisoning of the cathode and can lower the effi ciency of the cell In addition, p resent

PAGE 20

20 Di rect Alcohol Fuel Cells (DAFCs) are inefficient because oxidation of the fuel is both kinetically unfavorable and may occur either slowly or incompletely. When co nsidering systems with heterogeneous catalysts on solid support s one reason for inefficient metal reactivities is the fact that the bulk of the metal is below the surface and therefore does not react. In addition, systems that utilize Polymer Membrane El ectrode Assemblies (PMEAs) result in decreased cell performance because of alcohol crossover from the anode to the cathode through the polymer membrane with increasing time 11 Taking these facts into account, homogeneous catalytic systems should allow for more efficient metal reactivities and also eliminate the use of polymer membranes. Methanol as Fuel Methanol occurs n aturally as a by product from decomposition in landfills and from t until the 17 th century that methanol was first isolated Today, methanol used industrially is produced by the synthesis g as method utilizing both renewable and non renewable fuels In this method, which was first discovered by the German chemists Matth ias and Pier, a mixture of carbon monoxide, carbo n dioxide and hydrogen gases is converted to methanol by means of a catalyst 12 the start of the OPEC Fue l Crisis; since that time methanol has stimulated intense interest as a power source for fuel cells that may be used in portable electronic or mobile devices. 9 Since it is the simplest alcohol the oxidation of methanol should proceed by a reaction mechanism containing fewer inte rmediates than other alcohols In addition, m ethanol has a high theoretical mass energy density (6.1 kWhkg 1 ) and has demonstrated power densities of approximately 200 300 mWcm 2 at 90 C. 5,13 15

PAGE 21

21 Methanol, unfor tunately, has disadvantages because it is not considered a renewable energy source (unless produced from biomass) and is quite toxic Another disadvantage of using methanol in a fuel cell is that although it is capable of produ cing six electrons it underg oes a multistep mechanism on bulk platinum electrodes ( Figure 1 2 ). 16,17 The complete oxidation of metha nol produces six electrons along with CO 2 This mechanism involves several successive dehydrogenations of methanol on the platinum surface resulting in the release of four electron s and surface bound CO. In order for this bound CO to be released from the met al, dissociation and adsorption of water must first occur (two additional electrons released) ; followed by reaction of the adsorbed CO and OH to produce CO 2 and H 2 O (Figure 1 2). Figure 1 2 Methanol oxidation on bulk platinum anode ( adapted from Lege r and Kunimatsu 16,17 ) In a homogeneous catalytic system, methanol first experiences a two electron o xidation to formaldehyde (Figure 1 3) and if water is present, this formaldehyde can be further oxidized to formic acid with an additional loss of two electron s When methanol is present in excess, these products are n o t observed Instead, acid catalyzed condensations occur to yield dimethoxymethane (DMM) the two electron oxidation product and methyl formate (MF) as the observed four electron oxidation product (Figure 1 3) Since these liquid condensates are observable by gas chromatography,

PAGE 22

22 the molar amounts produced can be followed as the reaction progresses in order to determine the ability of each complex to catalyze the oxi dation of methanol Figure 1 3. Oxidation of methanol in homogeneous systems Ethanol as Fuel The first known consumption of ethanol dates to approximately 9000 years ago when it was used as a beverage in l the early 19 th determined by the chemists Lavoisier and de Saussure. 18,19 As early as the mid 19 th century, ethanol was used as a lamp fuel and in the early 20 th century, Henry Ford produced the first ethanol fueled automobile. 20 Like methanol, interest in ethanol was renewed in whe n it was (and still is) commonly blended with gasoline to pro % ethanol and 90 % gasoline (E10) 21 In rece nt years, because of the increased demand in renewable energy sourc es and also because of the advent of more environmentally friendly practices, ethanol is being explored more earnestly than before This is mainly because ethanol (and its oxidation by products) is less toxic than other alcohols (including methanol ) can be produced solely and easily by fermentation of biomass and biowaste (so has net zero carbon emissions) and has a higher theoretical mass energy density than methanol (8.0 kWhkg 1 ). 5,13 15 Although ethanol is onl y able to achieve a power density of

PAGE 23

23 approximately 50 mWcm 2 at 90 C (compared to 200 mWcm 2 at 90 C for methanol), complete oxidation of ethanol w ould theoretically release twelve electrons per molecule In addition, ethanol is the second simplest alco hol type fuel thereby requiring few bond dissociations compared to higher alcohols. 5,7,11 Similar to methanol, ethanol also undergoes a multistep oxidation mechanism (Figure 1 4 mechanism elucidation i n acidic med ia adapted from Lamy et al. ). 2 F igure 1 4. Ethanol oxidation on bulk platinum anodes (adapted from Lamy et al. 2 ) On platinum surfaces (at 0.8 V vs. RHE) ethanol first undergoes dissociative adsorption to acetaldehyde with the release of two electron s. This acetaldehyde i s then readsorbed onto the catalyst and reacts with surface bound hydroxide species to produce acetic acid after a further release of two electron s. It was also determined that at low potentials ( < 0.2 V vs. RHE) 2 the adsorbed acetaldehyde reacts with surface

PAGE 24

24 bound hydroxide species, formed from the activation and dehydrogenation of water by platinum, to complete the oxidation of ethanol and produce CO 2 and methane. In a homogeneous system, this oxidation occurs to first yield acetaldehyde with the release of two e lectron s ( Figure 1 5 ) If water is present, this acetaldehyde may then be oxidized to the four electron oxidation product acetic acid (AA) As experienced with methanol oxidation when excess ethanol is present, condensation of the two and four electron oxidation products acetaldehyde and acetic acid produces 1,1 diethoxyethane ( DEE ) and ethyl acetate (EA) respectively. 2 Figure 1 5 Ethanol oxidation in homogeneous systems In the last decade the majority of ethanol oxidation catalysts were still unable to dis sociate the C C bond thereby mainly resulting in two carbon oxidation products observed (and sometimes very small amounts of CO 2 ). 2,5,6,10 M ost recently, however, a multitude of literature articles demonstrating the complete electrooxidation of ethanol to CO 2 using heterogeneous systems have been published 22 32 Types of Catalysts and Catalytic Systems It is well known that Pt is the most active single site catalyst for alcohol oxidation ; p latinum anodes can allow for successive adsorption and dehydrogenation of both the alcohol and water, followed by surface reactions between the adsorbed species. 16,33 Unfortunately, p latinum has a major disadvantage because it is poisoned by the CO intermediate produced during the reaction. 1,9,34,35 Examples by which this CO poisoning

PAGE 25

25 may be diminished include increasing the working temperature of the cell, employing pulsed voltages during the oxidation and even inserting oxygen atoms onto the platinum surface that could allow for the selective oxidation of CO to CO 2 Expe rimental s tudies have shown however, that t he most reliable way of removing CO from the platinum surface is by adding one or more metals to make either a bi or tri metallic system. 35 Some of the metals th at have been used as catalytic additives include Ru, Sn, Pb, W, Mo, Fe and Rh. 9,34 36 The Pt /Ru combination has been used for some time because it is considerably less sensitive to CO poisoning and also because i t induces oxidation at lower potentials than traditional Pt anodes. 2,5 7,37 The mechanism by which this decrease in overpotential occurs was first proposed by Watanabe and Mootoo in 1975. 38,39 Their findings suggested that Ru was able to adsorb and dehydrogenate water at lower potentials than Pt thereby allowing the electro oxidation to occur sooner In this mechanism, Pt still works to adsorb and dehydrogenate the alcohol to CO but now Ru works to adsorb and activate water to OH or oxygen The adsorbed OH or oxygen o n the Ru sites can then react with the Pt bound CO intermediate s to produce CO 2 ; this in turn allows for faster regeneration and less poisoning of the Pt /Ru catalyst compared to Pt systems. In this bi functional mechanism, the two metals act in a synergis tic fashion where Pt performs as the catalyst and Ru facilitates oxygen transfer. 40 T he Pt/Sn combination when used to oxidize ethanol, w as shown to have increased reaction rates compared to Pt and Pt/Ru. 10,34,35 Similar to the previous Pt /Ru system, Sn is able to adsorb and activate water at lower potentials than Pt thereby allowing for lower overpotentials of the cell and less poisoning of Pt Initial experiments

PAGE 26

26 with S n established that it favored the dissociative adsorption of ethanol which should in turn lead to dissociation of the C C bond at lower overpotentials and with greater selectivity than pure Pt. 41 Other bimetalli c systems studied include Pt /Rh, Pt/Re and Pt/Mo as well as more complicated alloys such as Pt/Ru/Rh and Pt/Ru/Mo/W which are all better than Pt in their ability to oxidize alcohols but not better than the Pt/Ru and Pt/Sn systems. 34 S everal other alloy system s, such as Pd/Pt/Rh Pd/Rh Cr/Os and La/Mn, have all been shown to be efficient catalysts for alcohol oxidation. 28,32,42 44 In addition to alloys, several types of c atalytic systems exist, such as different types of nanoparticles gr aphene nanosheets, and nanocoils 24,30,45 53 as well as catalysts adsorbed onto various other solid support s 26,27,31,54 The majority of catalysts used in alcohol oxidat i on s are heterogeneous in nature and as such, these catalysts only utilize surface reactions, resulting in inefficient metal usage In comparison, homogeneous systems allow for better utilization of the metal. Rhodium as a Catalyst In the past, the gener al acceptance was that C C bond cleavage (caused by the insertion of a transition metal into the bond) could only occur if the bond was in the presence of a carbonyl group or if the cleavage resulted in a decrease in strain or in the formation of an aromat ic system. 12,55 In 1993, however, a b reakthrough was made when the Milstein group observed the first example of a transition metal being inserted into a C C bond (in a homogeneous system) that was neither strained nor activated (Figure 1 6 ). 55 This insertion was observed using a Rh(I) catalyst and was preceded by C H activation (kinetic produc t), followed by elevated temperature and pressure (thermodynamic control) in order to obta in the C C activation product.

PAGE 27

27 Figure 1 6 Scheme showing Rh insertion into a C C bond Their use of a metal hydride complex was intended to provide a significant d riving force through elimination of methane after insertion into the C C bond thereby resulting in an irreversible and thermodynamically stable process. 55 They concluded that their observation was based on two factors: the insertion process needed to be intramolecular and the transition metal needed to be p roximal to the C C bond in which it was to be inserted. In related work in 1993, Brown and Barteau performed surface studies to investigate the activation of C H, C O and C C bonds of oxygenated molecules on the heterogeneous, catalytically active Rh(111) plane. 56 Using this catalyst, they found no evidence of aldehyde formation from primary alcohols and suggested that the decarbonylation pathway for alcohols and aldehydes were divergent because: Aldehydes produced volat ile hydrocarbon products one unit shorter than the parent whereas alcohols did not. CO elimination from ethanol occurred at lower temperatures than CO elimination from acetaldehyde; hence, acetaldehyde formation on the Rh surface is not an intermediate f ormed from ethanol reactions. It was previously shown that dissociation of ethanol to form the ethoxide anion could occur on the Rh(111) surface below temperatures of 150 K, and since ethanol did not form acetaldehyde as an intermediate, a hypothesis was proposed where the ethoxide underwent C oxametallacycle on the Rh surface (Figure 1 7 ). 56 This hypothesis was proven by using

PAGE 28

28 Temperature Program med Desorption (TPD) where the catalyst was exposed to ethanol, acetaldehyde and ethylene oxide in three different experiments. The products observed from ethanol and ethylene oxide were both CO and H 2 while acetaldehyde also produced methane in addition to these. From these observations, it was concluded that the proposed bond activation sequence for ethanol on this type of catalyst should H and finally C C. Figure 1 7 Dissociation of ethanol on the Rh(111) surface Preliminary work in the McElwee White group using previously prepared heterobimetallic catalysts (Ru/Pt or Ru/Pd) to oxidize et hanol indicated the presence of DEE and AA after bulk electrolys e s 57 While it is improbable for monometallic Rh catalysts to form acetaldehyde (or DEE ) as a product from ethanol oxidation, Rh has been shown to oxidize acetaldehyde (as well as ethanol) if it is provided with it. 58,59 As such, a second metal may be paired with Rh in order to facilitate the o xidation of ethanol, either by assist ing with the decarbonylation or dehydrogenation steps or by allowing for lower overpotentials and less poisoning. Some of the catalytic applications for rhodium are hydroformylation of alkenes to aldehydes, dehydrogenat ion of saturated hydrocarbons to alkenes, isomerization of alkenes and also the carbonylation of methanol to acetic acid. 12,60 In addition, a lthough not a traditional industrial use, there are seve ral rhodium catalysts possessing the ability

PAGE 29

29 to cleave C C bonds ; r ecent work in this area either stems from homogeneous Rh(I) cata or from solid heterogeneous catalysts. 61 64 Stemming from their 1993 article, the Milstein group was later able to prove that insertion into a C C bond could also be accomplished without the use of hydrogen as a ligand in the catalytic system (Figure 1 8) Heating the Me PCP substrate to 150 C with [Rh(PEt 3 ) 3 Cl] allowed for quantitative and direct C C activation. 65 Subsequent forays into this design revealed the ability to activate the C C bond at room temperature either by substituti ng bulky tert butyl groups onto the phosphanes or by using the cationic [Rh(COE) 2 (THF) 2 ]BF 4 complex (Figure 1 9 ) 66 69 Figure 1 8. Activation of a C C bond Substantially strong bonds to metals are usually the result of breaking the strongest substrate bonds. 70 72 As such, the new Rh C bond formed should be quite strong and can be thought of as both a driving force of the C C bond activation and a hindrance to the overa ll oxidation of ethanol; if the new Rh C bond is too difficult to break then catalyst poisoning by the adsorbed carbon species may occur. Data listing Rh C dissociation energies vary somewhat depending on the charge of the metal as well as the type of cal culation performed. The dissociation energy of the diatomic Rh + C cation w as determined to be approximately 414 17 kJmol 1 while the Rh C bond is slightly higher at 580 4 kJmol 1 73 While these dissociation energies may seem quite

PAGE 30

30 that the corresponding Pt + C diatomic cation requires 531 5 kJmol 1 73 As a disadvantage, the bond energy of CO adsorbed to Pt is 125 KJmol 1 while for Rh, it is 134 KJmol 1 a fact that may result in slower kinetics of ethanol oxidation for new Pt /Rh complexes wh en compared to previous Ru/Pt complexes. 64,74 Previous studies, however, showed an increase in CO 2 produ ction over acetaldehyde for ethanol oxidation using the Pt/Rh and Pt/Pd / Rh alloys 42,74 This information may ultimately provide valuable clues towards catalysts for ethanol oxidation. It can be hypothesized that because Ru is able to promote lower overpotentials and dehydrogenate water to transfer oxygen to Rh, the Ru/Rh system may be able to oxidize ethanol with less poisoning and enhanced conversion of CO to CO 2 than the Pt/Rh systems. Figure 1 9 Activation of a C C bond at room temperature

PAGE 31

31 CHAPTER 2 ELECTROCHEMICAL OXIDATION OF ETHANOL USING HETEROBIMETALLIC CATALYSTS In a heterogeneous system, the complete six electron oxidation of methanol to CO 2 is complex and involves several stable intermediates including CO and formaldehyde. The CO can poison traditional Pt electrode materials and cause high overpotentials. More ef fective catalysis has been achieved by employing alloys, such as Pt/Ru, where the two types of metal centers can work cooperatively 75,76 Compared to traditional Pt anodes, this combination was found to induce oxi dation at lower potentials and be less susceptible to CO poisoning 2,5 7,37 Watanabe and Mootoo proposed that this decrease in overpotential was the result of the ability of ruthenium to absorb and dehydrogenate w ater thereby allowing the CO intermediate on Pt to be oxidized to CO 2 by a Ru oxo species on the surface 38,39 In addition, Pt still acts to absorb and dehydrogenate the alcohol in Pt/Ru anodes, however, it now do es this at lower potentials. Unfortunately, using bulk Pt/Ru metal anodes for mass produced fuel cells is hard ly practical in terms of cost. Thus, it is appealing to transfer this idea of heterobimetallic catalysis to systems where less of the precious m etals would have to be employed to obtain a similar number of active sites for alcohol electrooxidation A single molecule, homogeneous catalyst can allow for this. Interest in heterobimetallic catalysts stems from the ability of the metal centers to work in tandem whereby each metal may either play a specific role in the mechanism or one metal could influence the reactivity of the other. Consequently, previous work in the McElwee White group focused on he terobimetallic systems using Ru /Pt, Ru /Pd Ru/Au, Fe /Pt, Fe/Pd Fe/Au, Ru/Cu, and Ru/Sn 77 84 The Pt/Ru, Pd/Ru, Pt/Fe and Ru/Cu complexes all form a distorted six membered ring where the two metals are linked

PAGE 32

32 together by a halide and a phosphine bridge. The Ru/Au complexes are joined together in a linear fashion, being linked only through the phosphine bridge while the Ru/Sn complexes have direct Ru Sn bonds. Electrochemical Oxidation of Methanol Cyclic voltammograms of the previously synthesized first generation heterobimetallic complexes of the type CpRu(PPh 3 dppm) Pt X 2 CpRu(PPh 3 dppm) PdX 2 and CpRu(PPh 3 ) X dppm) Au (where X and X can be any combination of Cl or I) are performed in 1,2 dichloroethane which allows for complete solubility of the complex and also allows for straightforward observation of the potentials at which the metals are oxidized. The complexes generally exhibit three oxidation waves ( Figure 2 1 ) in the solvent window. 77,78,81 The f irst, is assigned to the one electron Ru(II/III) oxidation, the second to the two electron Pt(II/IV), Pd(II/IV) or Au(II/III) oxidation while the last is of the Ru(III/IV) oxidation 77 Cyclic voltammograms of the halide bridged complex es indicated differenc es in their oxidation potential s when compared to the mononuclear model compounds. These differences were attributed to electron donation from Ru to the more electron deficient second metal through the halide bridge. The potential of the Ru(II/III) oxida tion depended on the amount of electron donation through the bridging ligands; t his dependence is seen as an increase in the Ru(II/III) potential for the bimetallic complexes when compared to the mononuclear CpRu(PPh 3 1 2 dppm)Cl compounds. In contrast, the potential of the Ru(III/IV) oxidation varied only slightly between complexes. When comparisons were made between the Pt(II/IV) and Pd(II/IV) oxidations for varying complexes, it was found that the complexes with bridging iodides generally exhibited reversible waves while those with bridging chlorides did not. This reversible wave is a

PAGE 33

33 good indication of better stability of the iodide bridged complexe s while in the oxidized state. The linearly bound Ru/Au complexes which are connected only through the bis (diphenylphosphino)methane ligand show almost no difference in their oxidation p otentials when compared to the mononuclear model compounds thereby making the theory of electron donation through the halide bridge plausible. Figure 2 1 Cyclic voltammogram s of 1 in 0.7 M TBAT/DCE; 50 mVs 1 scan rate; glassy carbon working electron; Ag/Ag + reference electrode. Cyclic voltammograms acquired for the monomeric CpRu(PP 3 ) 2 1 dppm) 2 CpRu(PP h 3 1 2 dppm)Cl all demonstrated reversible waves for the Ru(II/III) couple showing good stability of the Ru monomers under these conditions. 77,78,84 In contrast, the monomeric Pt complexes dem onstrated irreversible waves indicating the inability of the Pt complexes to remain stable during the electrochemical experiment. When CVs of the monomers are compared to those of the

PAGE 34

34 heterobimetallic complexes, the reversibility of the Ru(II/III) couple disappears indicating loss of stability of the heterobimetallic complex compared to the Ru monomer. Cyclic voltammograms of these previously synthesized heterobimetallic complexes also display significant current increases after addition of methanol (Fig ure 2 1) 77,80,83 The increase in current occurs at the II/IV oxidation potential of the second metal (Pt, Pd or Au) and i s indicative of the ability of the complex to electrocatalytically oxidize the alcohol Af ter the addition of the substrate, water is added to the cell as a way to provide oxygen for the conversion of CO to CO 2 A further increase in current is usually observed starting at the II/IV oxidatio n potential of the second metal (Figure 2 1) indicati ng that the complexes are more active in wet conditions. In addition to cyclic voltammetry, bulk electrolysis of methanol was also performed using these complexes in order to determine if they could oxidize methanol. The products from these oxidations wer e analyzed using a combination of GC and GC/MS These detection methods indicated that under homogeneous oxidation conditions the oxidation of methanol resulted in the formation of the products from the two electron and four electron pathways; formaldehy de and formic acid respectively. These primary products were detected in solution as their acid catalyzed condensation products formaldehyde dimethyl acetal (dimethoxymethane, DMM) and methyl formate (MF) 77,81,83 ,84 When compared to the monomeric Ru analogues, the Ru/Pt, Ru/Pd and Ru/Au heterobimetallic complexes all yielded substantially larger amounts of the oxi dation products (DMM & MF) and demonstrated better current efficiencies during catalysis. 77,81,83 In addition, bulk electrolysis experiments with methanol either using a mixture of CpRu(PPh 3 ) 2 Cl and Pd(COD)Cl 2 or a mixture of CpRu( 2 dppm)Cl and Pt( 2

PAGE 35

35 dppm)Cl as models for the bimetallic system showed similar pr oduct distributions compared to the Ru monomers, however current efficiencies of these model systems were lower than those for the heterobimetallic complexes. These results demonstrate that the bimetallic structure of the Ru/Pt, Ru/Pd and Ru/Au complexes is significant and aids catalysis. 81,83 Electrochemical Oxidation of Ethanol Early work was focused on the utilization of methanol in fuel cells; however ethanol has many properties which make it an appealing alterna tive to methanol. In particular, ethanol is non toxic, obtainable from biomass, and has a higher mass energy density than methanol (8.0 kWhkg 1 vs. 6.1 kWhkg 1 ) 7,9,34,85,86 Additionally, ethanol fuel cell s are e xcellent model systems for higher hydrocarbon fuel cells due to the need to break the C C bond a problem not encountered with the oxidation of methanol. Cyclic Voltammetry with Ethanol Cyclic voltammograms were performed on the previously synthesized Ru catalysts 1 8 (Figure 2 2) in order to test t heir activity towards ethanol Similar to their electrochemical activity in methanol 82,83 the halide br idged complexes 1 5 and 7 8 and the dppm bridged Ru/Au complex 6 all demonstrated good increases in current after addition of ethanol ( Figure 2 3 Appendix ). 57 Complexes 9 12 (Figure 2 4) which were synthesized to be completely soluble in the substrate (methanol or ethanol) all demonstrated good current increases (Appendix) at the Pt(II/IV) and Pd (II/IV) oxidations after addition of ethanol; similar to that observed with methanol. 79 The bipyridyl substituted complexes 13 and 14 did not show current increases after addition of ethanol to the cell (Appendix).

PAGE 36

36 Figure 2 2. First ge neration c atalysts 1 8 Figure 2 3 Cyclic voltammograms of complex 4 in 0.1 M TBAT/DCE; 50 mVs scan rate; glassy carbon working electrode; Ag/Ag + reference electrode. The Ru carbonyl complexes 15 17 and their Fe analogues 18 20 (Figure 2 5) were prepared to compare the activity of Ru with the cheaper, more readily available, first row transition metal. 80 These complexes were soluble both in methanol and ethanol and demonstrated good current increases at the oxidation potential of the second metal upon ad dition of ethanol (Figure 2 6 Appendix).

PAGE 37

37 Figure 2 4. Alcohol soluble c atalysts 9 14 Figure 2 5. Ru and Fe carbonyl c atalysts 15 20 The p reviously prepared Ru/Sn heterobimetallic complexes 21 28 (Figure 2 7) were first synthesized to test whether t he Lewis acidic metal Sn could allow for better electron acceptance from the electron rich Ru metal center These complexes, when tested for activity towards ethanol (Appendix), indicated the best activity with 21 25 and 27 while 22 and 24 were the wors t (currents decreased after ethanol addition). Changing the electrolyte to TBA(BF 4 ) from TBAT did not allow for better elucidation of

PAGE 38

38 the oxidation waves Overall, the Ru/Sn compounds were generally not as active towards ethanol as the Ru/Pt or Ru/Pd com plexes. Figure 2 6 Cyclic voltammograms of complex 1 9 in 0.1 M TBAT/DCE; 50 mVs scan rate; glassy carbon working electrode; Ag/Ag + reference electrode. Bulk Electrolysis with Ethanol Electrocatalytic oxidation of ethanol w as shown to yield a complica ted mixture of products. Similar to methanol electrooxidation by homogeneous catalysts the electrocatalytic oxidation of ethanol form s the two electron oxidation product, acetaldehyde, and the four electron oxidation product, acetic acid (AA), each of wh ich can then undergo acid catalyzed condensation reactions to form the secondary products 1,1 diethoxyethane (DEE) and ethyl acetate (EA), respectively. 2,5,6,10,87 Since the condensation reactions of ethanol deriv ed products are slower, preliminary data on the electrolysis of ethanol indicated formation of the primary products acetaldehyde and AA as well as their ethanol condensates DEE and EA. 2,5,6,10 Not only does the ox idation

PAGE 39

39 of ethanol occur by a more complicated mechanism, but analysis of the reaction mixtures is also more complex due to the detection of both primary and secondary reaction products. Figure 2 7. Ru/Sn catalysts 21 28 Bulk electrolysis in dichloroet hane using complexes 1 and 4 Bulk electrolys is experiments of ethanol using the complexes 1 and 4 were performed similar to those with methanol, utilizing 0.1 M TBAT/ DCE as t he electrolytic medium with addition of 0.35 M (82 L) ethanol to the cell. These experiments were performed at 1.5 V vs. NHE and resulted in oxidation products similar to those observed with me thanol E thanol underwent a multistep oxidation mechani sm yielding acetaldehyde and AA and because the acid catalyzed condensations of these p roducts are slower, the y could be detected directly along with some amounts of their ethanol condensates

PAGE 40

40 D ata from the chloride bridged Ru/Pt complex 1 indicated formation of the four electron oxidation product AA almost immediately after the start of t he reaction (Figure 2 8 ) 57 The resulting ethanol condensation product, ethyl acetate (EA) appeared after approximately 200 C of charge was passed. This observation is consistent with an acid catalyzed condensation between AA and eth anol that resulted in the formation of EA. Although the two electron oxidation product acetaldehyde is undoubtedly an intermediate in the oxidation of ethanol to AA nei ther acetaldehyde nor its condensate DEE could be detected. Figure 2 8 Evolution of the oxidation products from the electrooxidation of ethanol in 0.1 M TBA(BF 4 )/EtOH at 1.7 V vs. NHE for the Ru/P t complex 1 The iodide bridged Ru/Pd complex 4 meanwhile, was able to electrocatalytically oxidize ethanol to acetaldehyde, the two electr on oxidation product, and AA, the four electron oxidatio n product (Figure 2 9 ) 57 A cetaldehyde was observed early in the electrochemical experiment ; however, its c oncentration did not increase conside rably as

PAGE 41

41 the reaction progressed. In contrast, the evolution of AA did not occur until after 75 C of charge was passed, but the concentration increased almost exponentially. These results are consistent with further oxidation of acetaldehyde to AA as the reaction progresses. Unlike the b ridging chloride complex 1 which continued to function even after 400 C of charge had been passed, the iodide complex 4 seemed to produce a higher concentration of AA during the oxidation but a plateau was reached just before 200 C suggesting that the cat alyst had been deactivated. Figure 2 9 Evolution of the oxidation products from the electrooxidation of ethanol in 0.1 M TBA(BF 4 )/EtOH at 1.7 V vs. NHE for the Ru/Pd complex 4 Attempts to analy ze the oxidation products by FT IR of the h eadspace gases a fter electrolysi s of 1 and 4 showed only the presence of DCE. This observation is understandable since the amount of ethanol added to the cell was small compared to the solvent ; gaseous DCE was therefore more abundant in the cell than any electrolysis

PAGE 42

42 pro ducts and the small amounts of these liquid products would be more likely to stay dissolved in the solution instead of volatilizing Bulk electrolysis in ethanol using the alcohol soluble complex es 12, 13 and 14 Bulk electrolysis of 12 was performed in 0 .1 M TBA(BF 4 )/EtOH at 1.7 V vs. NHE since the protonated amin o substituent allowed for complete solubility of the complex in ethanol GC analysis of the reac tion mixtures during electrolyse s indicated formation of the two electron oxidation/condensation p roduct, DEE along with the four electron oxidation product AA (Figure 2 10 ) 57 No acetaldehyde was observed before the formation of DEE, which is consistent with rapid acid catalyzed condensation in neat ethanol. Although AA was obse rved after 50 C of charge had been passed, its concentration diminished substantially after 75 C. This seems to conflict with the previous observations from complexes 1 and 4 but may be explained by the fact that the GC retention times of EA and ethanol are similar when eluted from a polar GC column ( Alltech EC TM WAX in this case ) and as such, the position of the large solvent peak observed for ethanol in the GC data may simply be overshadowing a much smaller EA peak. This inability to observe EA during electrolysis is in agreement with previous results observed for heterogeneous systems 2 and efforts were made to resolve the issue of overlapping retention times by using a non polar column for subsequent experiments with ethanol soluble complexes. FTIR dat a of the headspace gases obtained during electrolysis with 12 indicated the presence of DEE which corroborated the assignment of the products observed by GC analysis Electrolysis in neat ethanol eliminated the need for a separate solvent and also allowed better quantification of headspace gases since ethanol is less volatile than D CE and does not dominate the FT IR data. Efforts to detect CO or CO 2 in the FTIR

PAGE 43

43 spectra of the head space gases were unsuccessful s uggesting that cleavage of the C C bond by com plex 12 is not viable under these conditions. Figure 2 10 Evolution of the oxidation products from the electrooxidation of ethanol in 0.1 M TBA(BF 4 )/EtOH at 1.7 V vs. NHE for the Ru/Pd complex 12 Bulk electrolyses of complexes 13 and 14 were performed in 0.1 M TBA(BF 4 )/EtOH since the cationic charge of the bipyridyl complexes allowed for good solubility in the polar solvent (Figure 2 11 ). C omplexes 13 and 14 were both able to oxidize ethanol to DEE, the ethanol condensate of acetaldehyde. Although cy clic voltammetry with the Ru/Pt complex 13 ( Appendix ) showed a considerable current increase after addition of ethanol to the cell, bulk electrolysis data after 300 C of charge passed, indicated inconsequential formation of the two electron oxidation produ ct. No four electron product (AA or EA) was observed during electrolysis with this complex. In contrast, the Ru/Pd complex 14 was able to oxidize ethanol more efficiently than 13 and demo nstrated sizable amounts of the acid catalyzed ethanol condensation product

PAGE 44

44 DEE Complex 14 was also unable to produce the four electron oxidation product (AA or EA). Figure 2 11 Evolution of the oxidation products from the electrooxidation of ethanol in 0.1 M TBA(BF 4 )/EtOH at 1.7 V vs. NHE for the Ru/Pd and Ru/Pt complexes 14 and 13 Table 2 1 Bulk electrolysis data for the oxidation of ethanol by c omplexe s 12, 13 and 14 Product Ratios (DEE/AA) Charge (C) 12 13 14 25 50 0.87 75 1.11 100 2.10 125 3.21 150 5.53 C E (%) after 150 C 38.5 No four electron oxidation product formed CE: The ratio of charge needed to produce the observed yields of DEE vs. the total charge passed during the experiment.

PAGE 45

45 Bulk electrolysis in ethanol using CO complexes 15 20 The CO containing Ru and F e heterobimetallic complexes 15 20 are soluble in ethanol so experiments were performed in 0.1 M TBA(BF 4 )/EtOH at 1.7 V vs. NHE (Table 2 2 ). P revious GC evaluations using the polar Alltech EC TM WAX column to analyze data after electrolysis with ethanol as the solvent proved complicated because of lack of separation between ethanol and EA. Efforts to resolve this problem indicated a non polar column would allow for better separation of the product mixture. T he products from the following experiments were therefore determined using the non polar J & W Sci entific DB 5 column T he heterobimetallic Ru/Au and Fe/Au complexes 1 7 and 20 decomposed during the electrolysis, as previously observed during methanol oxidation. 88 The Fe/Pd complex 1 9 although stable during the experiment, yielded no oxidation products after bulk electrolysis. The Ru/Pt ( 15 ), Ru/Pd ( 1 6 ) and Fe/Pt ( 1 8 ) complexes all converted ethanol to the two electron oxidation product acetaldehyde, whi ch was detected in the electrolyte solutions as its condensation product DEE. The primary two electron oxidation product acetaldehyde was not observed (similar to results with complex 9 ) nor was the four electron oxidation product AA or its ethanol conde nsate EA. In contrast to the results observed for neat methanol, the Ru/Pd complex 1 6 and the Fe/Pt complex 18 were more active in ethanol resulting in 30.2 % CE, TO = 8 and 17.5 % CE, TO = 5 respectively after 200 C of charge had been passed. The Ru/Pt c omplex 15 was less productive as an ethanol oxidation catalyst; having a CE of 22.9% and 8 TO compared to 63.1% and 12 TO in methanol. Although FTIR data acquired during the electrooxidation of methanol with complexes 15 and 18 indicated the presence of CO 2 in the headspace gas, sub sequent isotopic labeling studies proved this CO 2 was the result of a water gas shift reaction

PAGE 46

46 resulting in decomposition of the catalyst and conversion of the CO ligand to CO 2 88 In light of these results, headspace gas samples were not acquired during the electrooxidation of ethanol. Table 2 2 Bulk electrolysis data for the oxidation of ethanol by complexes 15 16 18 and 19 a,b,c Charge (C) Ru/Pt ( 15 ) Ru/P d ( 1 6 ) Fe/Pt ( 1 8 ) Fe/Pd ( 1 9 ) 20 30.4 59.1 15.1 n/o 40 62.1 95.8 35.8 n/o 60 105.1 134.0 54.5 n/o 80 150.9 171.0 78.2 n/o 100 175.7 193.5 110.5 n/o 150 197.6 249.0 168.3 n/o 200 237.5 312.9 181.4 n/o CE d (%) after 200 C 22.9 30.2 17.5 TO e after 2 00 C 6 8 5 a Electrolys es were performed at 1.7 V vs. NHE in pure ethanol. A 10 mM catalyst concentration was used for each experiment. b Determined by GC with respect to n octane as an internal sta ndard. n/o : no observable product c No observable fou r electron oxidation product was detected during electrolysis. d CE: The ratio of charge needed to produce the observed yields of DEE vs. the total charge passed during the experiment. e TO: The number of moles of DEE produced per mole of catalyst. The R u/Sn complexes were not used to explore the electrochemical oxidation of ethanol since previous methanol oxidation studies in dicated selectivity for the two electron condensation product DMM but very little formation of the four electron oxidation product MF. 84 Although these catalysts are interesting from an organic industrial standpoint because of their ability to easily synthesize DMM, they are not efficient complexes for the complete oxid ation of methanol and ethanol

PAGE 47

47 Summary The previously prepared heterobimetallic complexes 1 28 were investigated as catalysts for the electrocatalytic oxidation of ethanol. Cyclic voltammetry data were consistent with previously observed catalytic activit y for methanol ; when cyclic voltammograms were obtained with ethanol, these complexes all demonstrat ed increases in current typical of the ability of the complex es to be active towards the alcohol Electrooxidation of ethanol by complexes 1 and 4 in DCE resulted in the production of acetaldehyde AA and EA (GC yields) however FTIR analyses showed only DCE Electrocatalytic oxidation of ethanol using the alcohol soluble complex 12 produced DEE, the acetal of the two electron oxidation product acetaldehyde T he four electron oxidation product AA was also observed, however, difficulties with GC elution of the products prevented a definite conclusion about the presence /absence of its condensation product EA FTIR data of the headspace gas produced during ele ctrolysis of 12 indicated presen ce of DEE but n either CO nor CO 2 were observed. Electrooxidation of ethanol with the cationic bipyridyl complexes 13 and 14 could be perfo rmed in ethanol because of increased solubility of these complexes in polar solvents. Both the Ru/Pt complex 13 and the Ru/Pd complex 14 oxidized ethanol to the two electron product acetaldehyde. However, because of the ability of acetaldehyde to undergo rapid acid catalyzed condensation in ethanol, this oxidation product was observed as DEE. No four electron oxidation product (AA or EA) was observed with these catalysts. Electrooxidation with the ethanol soluble carbonyl complexes 15 16 and 18 resulted in the two electron ethanol condensate DEE however no four electron oxidation pro duct (AA or EA) could be observed during the experiments. The headspace gas was not sampled during the experiment due to previous results

PAGE 48

48 indicating the presence of observed CO 2 was due to loss of the CO ligand as the complex decomposed.

PAGE 49

49 CHAPTER 3 COMPL EXES BEARING THE BIPYRIDYL LIGAND Background In the continuing effort towards making complexes that are more easily soluble in methanol and ethanol, work was resumed exploring the cationic compl exes bearing the bipyridyl (bpy) ligand. The bipyri d yl ligand is interesting because it has been shown to allow for g ood stability when coordinated to a low oxidation state metal and as an added advantage the initial Ru(bpy) heterobimetallic complexes are also positively charged which should allow for incr eased solubility in polar solvents. Several ruthenium complexes based on the bipyridyl ligand are well known for their rol es as water oxidation catalysts. E xamples of these include the single site [Ru(tpy)(bpm)(OH 2 )] 2+ [Ru(tpy)(bpz)(OH 2 )] 2+ and [Ru(Mebi mpy)(bpy)(OH 2 )] 2+ complexes 89 91 as well as cis cis [(bpy) 2 (H 2 O)Ru III ORu III (OH 2 )(bpy) 2 ] 4+ or other bimetallic ruthenium or rhodium bipyridyl systems. 92 94 Previous research also indicated that Ru catalysts of the type [Ru(bpy) 2 (py)OH 2 ] 2+ and [Ru(bpy) 2 (py)O] 2+ were able to electrocatalytically oxidize primary and secondary alcohols It was found that these catalysts under went well defined electron transfer pathways via either oxygen atom or hydride io n transfers to result in two reversible one electron oxidations from the Ru(II/III ) and Ru(III/IV) couple s. 95,96 2,2 Bipyridyl Ru/Pd and Ru/Pt Complexes 30 and 31 Synthesis of Complexes 30 and 31 The cis bipyridyl ruthenium dichloride starting material was prepared in good yield according to literature procedure (Figure 3 1) 97,98 After the chloride was obta ined, it

PAGE 50

50 was reacted with sodium iodide in methanol in order produce the iodide derivative of the compound. 99 This iodide version was first reacted with AgPF 6 and the corresponding 1 dppm complex 29 (Figure 3 1 ). 97 Similar to the reactions performed to synthesize the chloride bridged versions 13 and 14 the 1 dppm complex was then reacted with the second metal in acetonitril e at room temperature overnight to obtain the Ru/Pd 30 and Ru/Pt 31 bimetallic complexes shown below ( Figure 3 2 ). Attempts at growing crystals of 30 and 31 for X ray analysis have prov en to be unsuccessful. These attempts were performed using both the diffusion and evaporation methods with different solvents. In many case s the product either never crystallized ( instead forming a sticky precipitate at the bottom of the vials ) or cryst als observed were too small or were twinned Figure 3 1 Synthetic scheme to 29 Synthesis of 29 at room temperature starting from the triphenylphosphine precursor (Figure 3 3) could be performed in a manner analogous to those for complexes 1 12 and 15 20 77,82,83 Unfortunately, t h is synthetic route did not offer any

PAGE 51

51 advantages since the reactions were performed for approximately 14 days and the crude product 29 still needed to be purified on a column in order to separate the 1 dppm product from unreacted starting material and 2 dppm side product. Figure 3 2 Preparation of 30 and 31 Figure 3 3 Alternative synthesis of [ cis (bpy) 2 RuI( 1 dppm)]PF 6 ( 29 ) Cyclic Voltammograms of Complexes 30 and 31 The cycl ic voltammograms of the iodide bridged Ru/Pd and Ru/Pt complexes 30 and 31 demonstrated disappointing results for activity towards ethanol oxidation. Complex 30 showed a small current increase starting from the Pd(II/IV) wave indicating

PAGE 52

52 marginal catalytic activity (Figure 3 4) while, 31 demonstrated almost no catalytic activity (Figure 3 5). Complexes 30 and 31 display the advantage of having the Ru(II/III) oxidation occur at lower potentials than their chloride bridged counterparts ( 13 and 14 ); 1.22 V vs 1.40 V for 30 and 1.23 V vs. 1.33 V for 31 (Table 3 2). Table 3 1 NMR d ata for c omplexes 30 and 31 obtained in CD 3 CN Figure 3 4 Cyclic voltammogram s of complex 30 in 0.1 M TBAT/DCE; 50 mVs 1 scan rate ; glassy carbon working electrode; Ag/Ag + reference electrode. Table 3 2 Formal potentials for c omplexes 30 and 31 Complex Couple E pa (V) Couple E pa ( V) Couple E pa (V) 30 Ru(II/III) 1.22 Pd(II/IV) 1.60 Ru(III/IV) 1.85 31 Ru(II/III) 1.23 Pt(II/IV) 1.63 Ru(III/IV) 1.87 1 31 Cp CH 2 of dppm Ru PPh 2 M PPh 2 30 3.85 (m), 3.32 (m) 38.1 (dd) 17.1 (dd) 31 3.88 (m), 3.40 (m) 36.8 (d d) 7.31 (dd)

PAGE 53

53 Figure 3 5. Cyclic voltammograms of complex 31 in 0.1 M TBAT/DCE; 50 mVs 1 scan rate; glassy carbon working electrode; Ag/Ag + refer ence electrode Bulk Electrolysis with Complexes 30 and 31 Similar to experiments with the chloride bridged complexes 13 and 14 bulk electrolysis with compounds 30 and 31 w ere performed in TBA(BF 4 )/EtOH at 1.7 V vs. NHE. The iodide bridged compounds 30 an d 31 oxidized ethanol to give the two electron condensation product DEE as well as the four electron oxidation product AA (Figures 3 6 and 3 7 Table 3 3 ). This is in contrast to the results obtained with the chloride bridged versions 13 and 14 (Figure 2 11 ); only the two electron condensate DEE was observed with these complexes. Of the iodide bridged complexes the Ru/Pd complex 30 performed better than the Ru/Pt complex 31 This observation parallels the results obtained with the chloride bridged comp lexes; the Ru/Pd complex 14 was more productive than the Ru/Pt complex 13 during the experiment.

PAGE 54

54 Figure 3 6 Evolution of the oxidation products from the electrooxidation of ethanol in 0.1 M TBA(BF 4 )/EtOH at 1.7 V vs. NHE for the Ru/Pd complex 30 F igure 3 7 Evolution of the oxidation products from the electrooxidation of ethanol in 0.1 M TBA(BF 4 )/EtOH at 1.7 V vs. NHE for the Ru/Pt complex 31 FTIR data of the headspace gas obtained during electrolyses of ethanol with these compounds indicated the presence of DEE which corroborates the products observed

PAGE 55

55 by GC. None of the complexes tested were able to oxidize ethanol to CO or CO 2 as observed by the absence of their stretching frequencies in the FTIR. Table 3 3 Bulk e lectrolysis d ata for the o xi dation of e thanol by c omplexes 30 and 31 Product Ratios ( DEE /AA) Charge (C) 30 31 25 50 75 0.24 100 0.34 125 0.35 0.85 150 0.50 0.43 Current efficiencies (%) at 150 C 56.6 50.0 CE: The ratio of charge needed to produce the observed yields of DEE vs. the total charge passed during the experiment. Attempted Synthesis of Fe Bipyridyl Complexes It was previously shown that iron bimetallic complexes have lower overpotentials than their analogous ruthenium complexes 80 i n addition, it is more economical to pur chase iron precursors rather than ruthenium precursors. As such, any electrocatalytic complex made with iron has a potentially more beneficial impact towards industrial and commercial use. In view of these reasons, attempts were made to synthesize a seri es of iron complexes similar to the ruthenium bipyridyl complexes 13 14 30 and 31 The hydrated cis bipyridyl iron dichloride was prepared similarly to its ruthenium analogue as a dark red powder in good yields. T riphenylphosphine (PPh 3 ) was substitut ed onto Fe in order to synthesize 32 by first removing the chloride with TlPF 6 in acetonitrile before reacting the salt with PPh 3 (Fig ure 3 8 )

PAGE 56

56 Figure 3 8 Synthesis of [ cis (bpy) 2 FeCl (PPh 3 ) ]PF 6 This PPh 3 precursor 32 was then reacted with dppm at roo m temperature in various solvents (Figure 3 9) ; however, 31 P NMR data never showed the presence of dppm attached to Fe ( 33 ). Figure 3 9 First attempt to synthesize [ cis (bpy) 2 FeCl ( 1 dppm) ]PF 6 An attempt was made to attach the dppm ligand to the hydrat ed cis bipyridyl iron dichloride in order to synthesize 33 in a manner similar to that performed for the 1 dppm compound 29 (Figure 3 10) This was attempted both at room temperature and by refluxing the reaction mixture, but once again, 31 P NM R data sh owed no reaction. Because of the inability of dppm to attach to the bis substituted iron bipyridyl, the mono substituted iron bipyridyl 34 was synthesized and reacted with dppm at room temperature to investigate whether dppm could be attached to the less hindered site.

PAGE 57

57 After first removing the chloride from the Fe center, dppm was added and the solution 1 2 product 35 was evident, as seen by a singlet at 15.2 ppm in the 31 P NMR (Figure 3 11 ). Figure 3 10 Second attempt to synthesize [ cis (bpy) 2 FeCl ( 1 dppm) ] PF 6 Figure 3 11 Attempted synthesis of [ bpy FeCl ( 1 dppm ) ]PF 6 2 product 35 with (excess) NaI in order to open the ring. It was observed that after stirring for 30 minutes at room temperature, a peak at 22.4 ppm appear ed in the 31 P NMR spectrum; t his peak wa s attributed to free dppm.

PAGE 58

58 2 product had been consumed leaving free dppm and (presumably) the bpyFeI 2 2H 2 O. Attempted Synthesis of Pd Bipyridyl Complexes Synt hesis of Complexes 36 and 37 Attempts were made to substitute the bipyridyl ligand on either the Pt or Pd metal center in the continuing effort to synthesize complexes that are able to oxidize ethanol. This was done primarily to investigate the effect tha t a different ligand may have on the Pt or Pd center. Prior work has shown that metal centers of the type (bpy)PtR 2 (R = H, alkyl), have the ability to bind and oxidize both water and methanol. 100 102 The bipyri dyl substituted [Pd(C 8 H 12 OCH 3 )(bpy)]PF 6 was first synthesized as described by Pietropaolo et al. starting from (COD)PdCl 2 (Figure 3 12) 103 Figure 3 12. Synthesis of the Pd bipyridyl starting material This precursor was first reacted with AgNO 3 in methanol until the precipitation of AgCl, bipyridyl to give the product which can th en be precipitated with NH 4 PF 6 (Figure 3 13) 103 After the Pd salt was obtained, an attempt was made to react it with the previou sly prepared Ru precursors [ 5 C 5 H 4 CH 2 CH 2 NMe 2 HCl]Ru(PPh 3 )( 1 dppm)Cl and ( 5 C 5 H 5 )Ru(PPh 3 )( 1 dppm)Cl in dichloromethane at room temperature overnight to yield 36 and 37 ( Figure 3 1 3 ).

PAGE 59

59 Figure 3 1 3 Attempted synthese s of 36 and 37 Data for the target c omplexes 36 and 37 only partially corresponds to the expected products. The 1 H NMR spectrum (discussed in Chapter 6) displays peaks corresponding to the total number of hydrogens having approximately the same positions as would be expected Also there is no presence of either starting material in the spectrum. The 31 P NMR data however, indicates splitting patterns and shifts that are confusing. As an example, the product obtained after reacting [Pd(C 8 H 12 OCH 3 )(bpy)]PF 6 with [ 5 C 5 H 4 CH 2 CH 2 NMe 2 HCl]Ru(PPh 3 ) ( 1 dppm)Cl has 31 P{ 1 H} NMR resonances at 48.5 ppm (t) (corresponding to Ru P P h, J PP = 36 Hz ) and what seems to be a doublet at 0.42 ppm ( J PP = 36 Hz). (this peak is quite large and may in fact be the presence of 2 singlets very close together). Neither of these peaks 2 dppm product (~ 15 ppm, s) and no starting material is observed. LSI MS data indicates the presence of three ions. The first at 813.1548 amu

PAGE 60

60 respresents the [M] + for CpRu(PPh 3 1 dppm)Cl (after loss of the chloride) whil e the second and third peaks appear at 957.0372 amu and 1095.1415 amu respectively. The peak at 957.0372 amu corresponds to loss of 153.8 amu while the peak at 1095.1415 amu corresponds to loss of 15.7 amu from the expected compound (1110.852 amu). Attem pts at crystallizing this compound for X ray analysis only resulted in precipitation of a stick y residue; a s such, the identities of these compounds are still unknown Cyclic Voltammograms of Complexes 30 and 31 The cyclic volta mmograms of the unidentifi ed Pd bipyridy l compounds 3 6 and 3 7 (Figures 3 1 4 and 3 1 5 ), were measured in order to help with the elucidation of the structure They show good current increases after addition of ethanol in both wet and dry conditions, however, neither CV indicated the presence of the Pd(0/II) oxidation 104,105 and it was assumed that if these complexes do exist as proposed, the Pd(0/II) wave may instead be coincident with the Ru(II/III) wave In order to test this hypothes is, C Vs were performed at scan rate s of 50 mVs 1 20 mVs 1 and 5 mVs 1 ; however the Pd oxidation was still not observed. I t can therefore be concluded that these compounds do not exist as proposed and attempts at growing crystals suitable for X ray structure d etermination proved unsuccessful Summary The heterobimetallic Ru/Pd and Ru/Pt complexes 30 and 31 were synthesized. These complexes exhibit both a bridging phosphine (which helps the complex to remain joined and stable) and a halide (which allows for elec tron transfer from the electron rich Ru to the electron poor Pd). Electrochemical experiments performed with these complexes show oxidation of ethanol to DEE, the two electron ethanol condensate, and AA, the four electron oxidation product.

PAGE 61

61 Figure 3 1 4 Cyclic voltammogram of 36 in 0.1 M TBAT/DCE; 50 mVs 1 scan rate ; glassy carbon working electrode; Ag/Ag + reference electrode Figure 3 1 5 Cyclic voltammogram of 37 in 0.1 M TBAT/DCE; 50 mVs 1 scan rate ; glassy carbon working electrode; Ag/Ag + refere nce electrode In addition to the Ru bipyridyl complexes, the preparation of analogous Fe complexes was explored; however, these compounds could not be made in the manner

PAGE 62

62 attempted. A ttempts at synthesizing complexes with a bipyridyl substituted Pd were al so unsuccessful.

PAGE 63

63 CHAPTER 4 ETHANOL SOLUBLE COMP LEXES Background The complete oxidation of ethanol should first consist of dehydrogenation and C C bond dissociation followed subsequently by CO formation along with adsorption to the metal surface and the n oxidation to CO 2 resulting in the total release of twelve electrons (Figure 1 4) When existing methanol oxidation catalysts (e.g. Ru/Pt) are utilized for ethanol oxidation, the products observed are acetaldehyde (the two electron oxidation product) an d acetic acid ( AA, the four electron oxidation product). If the electrolysis is performed in excess ethanol, acetaldehyde may also react with adsorbed ethanol to form the condensation product 1,1 diethoxyethane (DEE) while acetic acid forms ethyl acetate (EA) 2,5 Previous work in the McElwee White group utilizing a h omogeneous catalytic system indicated the presence of DEE and AA but no EA 57 Although t hese existing catalytic systems may be v aluable for producing industrially useful chemicals, they are unsuitable for use in fuel cells because of their inability to oxidize ethanol completely In order for ethanol to be a viable and environmentally harmless fuel source, its oxidation must be co mplet e and result only in the products CO 2 hydrogen and electrons. Only then will ethanol be considered as a recyclable energy source. Previous reports indicated the presence of many catalysts that are capable only of partial ethanol e lectrooxidation b ecause of their inability to activate or dissociate the C C bond. Density functional theory (DFT) studies have calculated that the energy needed to dissociate the C C bond of ethanol (in the gas phase) is 357 kJmol 1 while the CRC Handbook of Chemistry a nd Physics lists a slightly higher bond dissociation

PAGE 64

64 energy of 364.8 4.2 KJmol 1 (at 298 K). 73,106 The abi lity to break the C C bond through catalysis is, however, kinetically unfavorable because of both the steric hindrance associated with the C C bond and also because of the directional nature of the sp 3 sp 3 hybridized orbitals forming the bond (relative to s s or s sp 3 hybridized bonds for example). 12,107 Consequently the activation energy for this process is higher than the kinetically more favorable C H bond activation. Functionalized Ethanol Soluble Cyclopentadienyl Ligands The cyclope ntadienyl ligand is widely used in organometallic chemistry because it 3 5 ). This ability to change hapticity can allow for an open metal coordination site and incre a sed reactivity of the complex. In addition, ligand functionalization by replacing either one or more of the hydrogens with other substituents can allow for changes in the steric and electronic properties. 108 110 Work in the McElwee White group previously utilized the Cp ligand however, w hen it was determined that the hydrophobic phosphine groups hindered the ability of the complex es to dissolve completely in aqueous methanol additional investigations were made b y synthesizing Cp complexes bearing pendant amines. Since the neutral dimethylethyl amine 9 only provided a limited increase in solubility, the protonated version 12 was also synthesized and investigated (Figure 2 4) 79 Working towards this effort to provide compl ete solubility of the complex in ethanol, attempts were made to substitute the Cp ligand with a primary amine an alcohol and also a phosphine. Amino Su bstituted Cyclopentadienyl Ru/Pd and Ru/Pt Complexes The cyclopentadienyl ethyl amine ligand can be ea sily p repared using literature methods. 111,112 Dicyclopentadiene is first heated to 170 C until the retro Diels Alder

PAGE 65

65 reaction occurs to produce cyclopenta diene which then reacts with molten sodium in the reaction vessel to produce the sodium salt of cyclopentadiene. This salt is then reacted with the hydrochloride salt of 2 chloroethyl amine to produce the ligand which is purified via fractional distillation (Figure 4 1 ) After obtaining the purified ligand, it is immediately reacted with hydrated ruthenium trichloride and triphenylphosphine in ethanol to produce the Ru II starting mat erial 38 This ruthenium precursor is then reacted with bis (diphenylphosphino)methane (dppm) in tetrahydrofuran at room temperature for approximately 14 days or until one of the triphenylphosphine groups is replaced by the dppm ( Figure 4 1 ). The 1 dppm precursor 39 may then be converted to the bimetallic complex 40 via reaction at room temperatur e with either (COD)PdCl 2 or (COD)PtCl 2 It is interesting to note that during the reaction to produce 38 nucleophilic attack by the amine functionality can oc cur on the ruthenium center to replace the chloride group. This occurs when the reaction is either heated to higher temperatures and/or allowed to reflux for longer periods. In this case, the difference can be seen by the downfield shift from 39.9 ppm to 46.6 ppm of the phosphine groups in the 31 P{ 1 H} NMR.

PAGE 66

66 Figure 4 1 Synthesis of Ru/Pd 40 This assumption is corroborated by mass spectrometry which produces a molecular ion peak that is the same for both compounds. The only difference between these tw o compounds is the loss of the chloride and as such, both compounds display equivalent molecular ion peaks. A n attempt was made to break the Ru NH 2 bond by replacing the missing chloride with another more nucleophilic halide (iodide in this case) however, NMR data obtained after 6 days of reacting at room temperature yielded no difference. Another interesting phenomenon occurs when 39 is purified via column chromatography using solvents containing triethylamine as eluant. A 31 P NMR spectrum taken of the second band eluted from the column indicated an additional peak at 25.8 ppm (singlet) which was not present in the crude product mixture before purification. A mass spectrum of this second band indicated a molecular ion peak corresponding to product 39 bu t with triethylamine exchanged for the chloride on the ruthenium center. 1 H NMR data of this compound indicated a downfield shift of the

PAGE 67

67 triethylamine peaks (when compared to an NMR of triethylamine ) which is consistent with substitution In order to incr ease the solubility of the bimetallic complex 40 the protonated version was also prepared (Figure 4 2 ) The starting material [ 5 C 5 H 4 CH 2 CH 2 NH 2 ]Ru(PPh 3 ) 2 Cl ( 38 ) was dissolved in dichloromethane and washed with 1 M hydrochloric acid to give the protonated version of the amine. This protonated version was then reacted with dppm as before to yield 4 2 and the protonated bimetallic complex 43 may be prepared similarly to 40 by reacting 42 with the second metal precursor at room temperature ( Figure 4 2 ). Fi gure 4 2 Synthesis of Ru/Pd 43 Table 4 1 NMR d ata for c omplexes 40 and 43 S pectra were measured at room temperature using CDCl 3 1 31 Cp CH 2 of dppm Ru PPh 3 Ru PPh 2 M PPh 2 40 6.06, 4.53 5.73 5.54 (m) 37.8 (dd) 54.4 (dd) 21.3 (dd) 43 6.30 6.08 (m), 4.56 (m) 37.7 (dd) 53.9 (dd) 21.8 (d)

PAGE 68

68 Other attempts at s ynthesizing complex 40 Working with the amino substituted Cp ligands demonstrated ability of the primary amine to also react with Ru (Figure 4 3). This generally occurs when the reaction is performed at higher temperatures and for longer times, but was al so recently observed to be formed even when the temperature was lowered and the reaction performed using THF instead of EtOH. The side product is problematic because it cannot be removed by recrystallization; the crude product therefore needs to be purifi ed using column chromatography which is intensive and time consuming. Figure 4 3. Formation of 38 along with attachment of NH 2 to Ru Since protect ion of the primary amine would add two unwanted steps to the synthesis, a decision was made to deprotona te the ( C 5 H 4 )CH 2 CH 2 NH 2 ligand using NaH after distillation (Figure 4 4). The ligand was dissolved in THF then transferred to a cold slurry of NaH; after complete transfer, the ice bath was removed and the mixture allowed to react overnight. The solvent was removed from the resulting tan solution and the very fine off white precipitate was washed with hexane to yield the salt in 35 % yield. This reaction can also be performed with good results using n butyl lithium in hexane Once the salt wa s obtained, it under went rapid reaction with the Ru precursor (PPh 3 ) 3 RuCl 2 which can be synthesized according to a procedure by Wilkinson and Stephenson 113 (Figure 4 4). NMR scale reactions had indicated reaction in THF to be too rapid and side products were observed after two hours so CH 2 Cl 2 was chosen as solvent. A cooled solution of the depr otonated ligand was transferred to a cooled

PAGE 69

69 solution of (PPh 3 ) 3 RuCl 2 in CH 2 Cl 2 and the ice bath removed after complete addition. The dark orange mixture was stirred at room temperature for one hour before it was filtered and the solvent removed under redu ced pressure. T he crude product was recrystallized using CH 2 Cl 2 /Hexane to give 38 as a bright yellow powder in 48% yield. Once acquired, 38 was reacted with dppm to produce 39 as before (Figure 4 1) After stirring for 10 days at room temperature, the c rude product was eluted from a silica column starting from a solution of 10% hexane in CH 2 Cl 2 and gradually increasing the polarity to 5% MeOH in CH 2 Cl 2 to produce 39 as a bright yellow solid in 10% yield. Figure 4 4. Formation of 38 starting from the deprotonated (C 5 H 5 )CH 2 CH 2 NH 2 ligand The major hindrance with the reactions discussed above relate to the low yields making it difficult to purify these compounds as well as obtain enough of the product to both continue on to the next reaction and also to be able to amass enough of complex 40 in order to perform bulk electrolysis experiments. A final attempt was made to deprotonate the Cp ligand on CpRu(PPh 3 ) 2 Cl using n b utyl lithium followed by reaction with the ethyl amine ( Figure 4 5). The reaction wa s based on a procedure published by Tyler and coworkers in 1998 and involved dissolution of CpRu(PPh 3 ) 2 Cl in THF and cooling to 78 C followed by addition of the lithium salt. 114 The mixture was allowed to stir for one hour before warming to room temperature to ensure complete deprotonation

PAGE 70

70 of the Cp ligand and the resultant brown solution was recooled to 78 C be fore being transferred to a cold suspension of the amine salt in THF. The brown solution of the deprotonated CpRu(PPh 3 ) 2 Cl changed to yellow upon complete transfer to the amine and eventually became orange when warmed to room temperature. Th e mixture was filtered through C elite, the solvent removed and NMR spectra obtained. The 1 H NMR spectrum showed presence of the ethyl amine substituted on the Cp ligand and the 31 P NMR had several peaks indicating the reaction produced side products most likely stem ming from the reaction with the amine moiety of the bromoethyl amine salt with the deprotonated Cp. Figure 4 5. Additional attempt at substituting the Cp ring The above conclusion is supported by mass spectrometry obtained after the reaction of what w as assumed to be the iodide version of 38 with dppm. Both ESI and DART spectra obtained after the reaction indicated the presence of a peak at 841.1905 amu (Figure 4 6) which could only be explained by loss of two halides followed by attachment of the eth yl moiety to Ru; as such, the amine moiety of the 2 bromoethyl amine hydrochloride was also reacting under the conditions described above. Electrochemistry of a mino s ubstitu ted c yclopentadienyl c omplexes 40 and 43 The cyclic voltammograms of the chloride bridged Ru/Pd complexes 40 and 43 (Figures 4 7 and 4 8 ), both show ed significant current increases after addition of ethanol starting from the Pd(II/IV) oxidation, however, addi tion of water to the cell

PAGE 71

71 caused only a small increase in current. In order t o determine the position of the oxidations, a comparison was made between a cyclic voltammogram scanned at 50 mVs 1 and one scanned at 20 mVs 1 ; the two electron oxidation of Pd is more distinct at slower scan rates and so can be assigned (Fig ure 4 9 ). Wh en compared to previously prepared amines 2 5 30 the oxidation potentials of the Ru(II/III) couple are approximately the same as in the previously prepared tertiary amine complexes and 40 and 41 are also able to easily dissolve in ethanol. Figure 4 6. Results obtained from ESI MS data Table 4 2. Formal p otentials for c omplexes 40 and 43 Complex Couple E pa (V) Couple E pa (V) Couple E pa (V) 40 Ru(II/III) 0.91 Pd(II/IV) 2.07 Ru(III/IV) 2.95 43 Ru(II/III) 1.22 Pd(II/IV) 2.00 Ru(III/IV) 2.99 Bulk elect rolysis data were never acquired for complexes 40 and 43 since they could never be obtained in large enough quantities to acquire both CV and bulk electrolysis data. Attempts at obtaining crystals suitable for X ray diffraction failed, generally resulting in either a sticky residue at the bottom of the vial after solvent evaporation or a hard solid precipitate at the bottom of the vial after the diffusion method.

PAGE 72

72 Figure 4 7 Cyclic voltammogram of complex 40 at 20 mVs 1 in 0.1 M TBAT/DCE; glassy carbo n working electrode; Ag/Ag + reference electrode Figure 4 8 Cyclic voltammogram of complex 43 in TBAT/DCE at 20 mVs 1 in 0.1 M TBAT/DCE; glassy carbon working electrode; Ag/Ag + reference electrode

PAGE 73

73 Figure 4 9 Cyclic voltammogram of complex 43 at scan rates 50 mVs 1 and 20 mVs 1 to determine oxidation positions; 0.1 M TBAT/DCE electrolyte; glassy carbon working electrode; Ag/Ag + reference electrode. Attempted Syntheses of Hydroxyl Substituted Cyclopentadieny l Complexes As a continuation of the amino s ubstituted cyclopentadienyl ligands, the preparation of hydroxyl substituted Cp was also attempted. Although the most common way to synthesize the (C 5 H 5 )CH 2 CH 2 OH ligand is by reaction of ethylene oxide with the cyclopentadien yl ligand 115,116 this reaction was decided against because of the cost of ethylene oxide ($544 per 227 g tank from Sigma Aldrich). The synthesis was performed as for the previously described amino substituted cyclopentadiene adapted from the procedure published by Causey (Figure 4 10) 112 The CpNa was dissolved in THF, cooled to 0 C and a THF solution of 2 chloroethanol was added dropwise resulting immediately in a white precipitate. After refluxing the mixture for six hou rs, the solvent was removed, the solid residue dissolved in water, extracted with diethyl ether and pentane dried with MgSO 4 then concentrated on the rotovap to give a brown orange

PAGE 74

74 very viscous oil. This oil was purified using fractional distillation to produce the ligand (Figure 4 10) in 7.7 % yield. Once the ligand has been synthesized, it can be refluxed with hydrated ruthenium trichloride and triphenylphosphine in ethanol over the course of 2 days (Figure 4 10) At this time, ethanol was removed fro m the resulting yellow brown solution to yield a sticky residue. Several columns were necessary to purify the crude product ; the first being a silica column using 3% MeOH in diethyl ether which remove d unreacted ruthenium trichloride Two s ubsequent alu mina columns employing 3 % MeOH in Et 2 O followed by 3 % MeOH in CH 2 Cl 2 afforded pure 44 as an orange solid in 29% yield The major reason for such intensive purification is because the product 44 does not precipitate from solution since it is highly solub le in ethanol ; this property therefore prevents the complete removal of ethanol under reduced pressure. As such, the initial silica column can be thought of mo re as a filtration since the sticky residue contains such a large amount of ethanol that purific ation is impossible at this point. Figure 4 10. Synthetic scheme to 44 and 45 The additional purification attempts removed unreacted triphenylphosphine and unreacted ligand. However, because they also caused 44 to be dissolved for large periods of ti me, they invariably produced oxidized phosphine as observed in both the 1 H

PAGE 75

75 and 31 P NMR spectra. As before, 44 was reacted with dppm in THF at room temperature for five days to replace one triphenylphosphine ligand on the Ru center. An NMR of the crude re action mixture indicated the anticipated product as well as the oxidized phosphine residue from the starting material ( 44 ), free dppm, free triphenylphosphine, and a minor unknown peak Attempts to recrystallize the crude product using CH 2 Cl 2 /hexane could not remove the oxidized phosphine so it was eluted from a silica column starting with 8% EtOAc in CH 2 Cl 2 followed by 40% EtOAc in CH 2 Cl 2 and finally with addition of a small amount of methanol to remove the yellow band in 45% yield Unfortunately once th is yellow band was removed, a 31 P NMR spectrum indicated a complete change in the splitting pattern of the phosphine peaks; instead of three distinct types of phosphine peaks with appr oximately the same intensity, there were now two distinct phosphine peak s with relative intensities of 1:2 most likely indicating attachment of the dppm ligand to Ru in a 2 fashion (Figure 4 11) Figure 4 11. Attachment of the dppm ligand as a bidentate chelate Additional attempts at resynthesizing 44 following the above procedure were abandoned because distillation of the ligand proved to be difficult; it invariably resulted in the collection of residual solvent while a large amount of dark brown sticky oil always remained in the initial flask The ligand was synthesized (Figure 4 1 2 ) following a procedure described by Partridge and Uskokovic in 1985 whereby a soluti on of methy l 2 bromoacetate dissolved in THF was added dropwise to a solution of CpNa in THF at

PAGE 76

76 78 C. 117 The mixture was stirred over night then the CO 2 /acetone bath removed and the suspension warmed to room temperature T he crude product was then filtered and washed with THF the filtrate collected and the solvent removed to produce the compound methyl 2 (cyclopentadienyl) acetate, in 43% yield Once prepared, this compound is dissolved in THF and added dropwise over two hours to a cold ( 5 C) solution of lithium triethylborohydride in THF. 118 The ice bath was removed after complete addition and the yellow brown solution stirred overnight. The resulting brown solution was diluted with diethyl ether then water added and the organic layer extracted. When tested, both aqueous and organic layers were pH 14 so an aqueous solution of HCl was use d to decrease the pH in order to protonate the hydroxide ion. The organic layer was separated again, dried with MgSO 4 and the solvent removed. An attempt to distill the dark green oily liquid under reduced pressure resulted in only one distillate (approx imately 20 29 C) which corresponded to a mixture of residual solvents. This observation points towards a problem with the distillation of the ligand and not necessarily with the synthetic route used. Figure 4 1 2 Synthesis of the (C 5 H 5 )CH 2 CH 2 OH liga nd via a different route Attempts to deprotonate the Cp ligand of CpRu(PPh 3 ) 2 Cl (similar to Figure 4 5) using either n butyl lithium or ter t butyl lithium followed by reaction with 2 chloroethanol resulted in reformation of CpRu(PPh 3 ) 2 Cl. Efforts were al so made to synthesize the protected version of 45 (Figure 4 10) by first protecting the 2 chloroethanol before reaction with the cyclopentadienyl salt. The

PAGE 77

77 protecting group chosen was 3,4 d ihydro 2H pyran (DHP) because it is unreactive with metal nucleoph iles ( lithium or sodium salts of Cp) The reaction was performed based on a procedure by Salehi 119 where zirconium tetrachloride was first reacted with 2 chloroethanol in CH 2 Cl 2 then the sol ution cooled in an ice bath before dropwise addition of DHP (Figure 4 1 3 ) The reaction was monitored by FT IR and upon completion, quenched with water and extracted with diethyl ether. The organic layer was dried with CaCl 2 then concentrat ed under reduced pressure to give a pa le yellow solution in 72% yield. Once obtained the protected alcohol may then be reacted with the cyclopentadienyl anion either by using previously prepared CpNa or by the in situ deprotonation of Cp with n butyl lithium. After stirring the mixture at 4 C for two days, it was warmed to room temperature, dissolved with water and extracted with ether to yield the protected ligand in 99% as a viscous oil Figure 4 1 3 Protection of the 2 cyclopentadienyl ethanol reagent It should be noted at this poi nt that the ligand may be easily deprotected following 119 with ZrCl 4 in MeOH however if the deprotected ligand is reacted with ruthenium trichloride using the general procedure the problem of co mpound 44 not precipitating from solution will be encountered again. The protected ligand was therefore reacted with ruthenium trichloride and triphenylphosphine in ethanol (Figure 4 1 4 ) and after removal of the solvent a sticky residue remained due to e xcess ligand

PAGE 78

78 used in the reaction. The crude compound was eluted from an alumina column starting from 40% CH 2 Cl 2 in hexane then gradually increasing the polarity to 69% CH 2 Cl 2 Figure 4 14. Scheme to synthesize 46 and 47 Although some unreacted ligan d was co eluted with 46 the decision was made to proceed to the next step in the synthesis by reacting 46 with dppm s ince previous purifications ha d indicated the more ready oxidation of PPh 3 compared to dppm. A THF solution containing 46 and dppm was st irred at room temperature for 10 days before the solvent was removed under reduced pressure to yi eld a bright orange solid (Figure 4 14) The 31 P NMR spectrum of the crude product indicated presence of the product as well as oxidized phosphine and a large amount of unreacted dppm. The crude product was eluted from a silica column starting with 40% CH 2 Cl 2 in hexane and gradient shift until a fourth band eluted with 5% MeOH in CH 2 Cl 2 When NMR analyses were performed, it was determined that the product, 47 should have been present in the fourth fraction however similar to the previous results obtained after elution of 45 from a silica column, this compound showed two peaks in the 31 P NMR instead of the three peaks initially seen in the crude product Th ese peaks were identical to those observed with 45 after purification. A mass spectrum of this compound indicated the presence of molecular ion peaks (Figure 4 1 5 ) corresponding to loss of Cl from 47 (941.2 37 amu) as

PAGE 79

79 well as loss of both Cl and cleavage o f the protecting group (857.1800 amu). These results substantiate the NMR data which points towards the presence of two distinct phosphines (instead of three) that may be the result of loss of the halide ligand and subsequent attachment of the free PPh 2 t o form the 2 compound as shown with 45 (Figure 4 11) Figure 4 1 5 ESI MS data for 47 Attempted Syntheses of Phosphino Substituted Cyclopentadienyl Complexes Working from the premise that less phenyl ligands (in the form of PPh 3 and dppm) around the metal cente rs should allow for greater solubility in ethanol, a ruthenium precursor was proposed bearing a Cp PPh 2 ligand (which sh ould act as a bridging ligand) and no other phosphine ligands. The CpTMS and [RuCl 2 CO 3 ] x polymer may each be prepared easily from litera ture procedures then reacted to produce the CpRu(CO) 2 Cl and CpRu(CO) 2 I precursors (Figure 4 1 6 ) 80,120,121 Once obtained, attempts were made to substitute the Cp ligand using PPh 2 Cl and based on a procedure repor ted by Gladysz et al. in 2005. 122 The Ru precursor was dissol ved in THF, cooled to butyl lithium added dropwise to the bright orange solution which turned red then orange brown (Figure 4 17). The mixture was stirred for an additional six hours then warmed briefly to room temperature and was observed to darken further to red brown before being recooled to

PAGE 80

80 PPh 2 Cl in THF was then added and the reaction warmed slowly to room temperature overnight. A 31 P NMR spectrum of the crude mixture indicated more than 15 peaks and only a small pe ak corresponding to the PPh 2 Cl starting material. The crude product was filtered through layered silica/Celite and eluted with additional THF. The solvent was removed under reduced pressure; the residue dissolved in benzene, filtered through a silica/Cel ite plug to remove a fine beige solid then the benzene filtrate was concentrated, filtered through a glass filter paper into a round bottom flask and layered with pentane. Figure 4 1 6 Formation of the CpRu(CO) 2 Cl and CpRu(CO) 2 Cl precursors Figure 4 1 7 Attempted synthesis of (CpPPh 2 )Ru(CO) 2 I The precipitate was collected and 31 P NMR data once again indicated several phosphine peaks some of which were not present in the crude mixture. Mass spectrometry indicted the presence of an ion corresponding to 48 after loss of a CO

PAGE 81

81 ligand proving that the compound could be synthesized. When the same reaction was performed using tert butyl lithium the same number of side products were observed however instead of precipitation from benzene, the crude mixtu re was eluted from a silica column starting with 40% CH 2 Cl 2 in hexane and gradually increasing polarity to 7 0% CH 2 Cl 2 in hexane. The product 48 was collected in negligible yields as a yellow orange solid and displayed a phosphine signal at 26.6 ppm. LR D IP CI mass spectroscopy indicated fragmentation patterns corresponding to [C 17 H 14 PRuICO+2H] + at 508 Da proving the presence of 48 When the reaction was performed using the (CpPPh 2 )Ru(CO) 2 Cl precursor and tert butyl lithium the compound was observed in a 31 P NMR spectrum of the crude product after THF removal at 22.9 ppm (Figure 4 1 8 ) however, when workup was done as usual by precipitation from a benzene solution, no precipitate formed. The solution was collected, concentrated and an acquired 31 P NMR spectrum indicated several additional phosphine peaks; sign of either decomposition or additional reaction of any residual lithium salt Figure 4 1 8 Attempted synthese s of (CpPPh 2 )Ru(CO) 2 Cl and (CpPPh 2 )Ru(CO) 2 Cl 3 Pd In an attempt to stop further decom position of the compound, it was reacted with (COD)PdCl 2 in CH 2 Cl 2 overnight (Figure 4 1 8 ). After filtration of the solvent and recrystallization of the crude mixture, NMR data indicated recovery of the (COD)PdCl 2 starting material The reaction was repe ated this time by refluxing in the reagents in

PAGE 82

82 DCE The mixture was cooled and filtered to produce a yellow filtrate and residual metal (silver) mirror. Spectra obtained of the bright yellow solid after solvent removal indicated a small change in the 31 P NMR spectrum for this case (disappearance of 31 P peak at 29.5 ppm and appearance of a peak at 33.1 ppm) A sample was submitted for mass s pectrometry Summary Attempts to prepare ethanol soluble co mpounds by substituting the cyc lopentadienyl ligand wi th amino, hydroxyl, and phosphino substituents were no t successful The synthetic schemes are complicated by the purification techniques employed throughout and also by observed decomposition of the amino and hydroxyl substituted ligand s as they are heate d under vacuum for distillation When the amino substituted compound 38 was made, results demonstrated that the amino moiety had the ability to react with the ruthenium center thereby replacing the chloride ligand. In addition, other synthetic attempts t o obtain 38 and 39 resulted in low yield thereby making the attempts not viable preparation techniques The compounds containing the hydroxyl substituted ligands were also difficult; 44 could be synthesized, however, because purification involved several columns, it was collected in small amounts that also contained some oxidized phosphine. The presence of 45 was observed in a 31 P NMR spectrum of the crude mixture, but unfortunately, when the compound was eluted from a silica column, the splitting patter n in the 31 P NMR had changed, most likely demonstrating the loss of the halide ligand and formation of the 2 dppm version of 45 (Figure 4 11) This observation was also demonstrated by the protected version, 47 after it was elu ted from the silica column ( 31 P NMR data obtained after 45 and 47 were eluted from silica were identical). It should be concluded

PAGE 83

83 th at although silica does not generally affect the first generation complexes ( 1 14 ) when used for purification, the slight acidity may in these cases, play a role in the observed dehalogenation I t is uncertain whether purification from alumina will preve nt this from happening. Synthetic procedures to prepare the phosphino substituted compounds 48 and 49 resulted in several side products and the compound seemed to decompose readily. Attempts to react (COD)PdCl 2 with 49 by reflux in DCE indicated a chang e in the 31 P NMR data after the reaction; tests are still ongoing to determine whether a complex containing both Ru and Pd exists in the crude mixture. Although the acquired data point towards the ability to synthesize these complexes, the difficulties o bserved with purification at every stage as well as low yields demonstrate non viable synthetic preparations for these compounds. Since the concept behind using these amino, hydroxyl and phosphino substituted cyclopentadienyl ligands would offer increased solubility of the compounds in ethanol but not necessarily better performance of the catalyst, continued exploration of these ligands with the specific compound design based on the first generation catalysts ( 1 8 ) was abandoned

PAGE 84

84 CHAPTER 5 N HETEROCYCL IC CARBENES: STRATEGY TOWARDS MORE STABLE HETEROBIMETALLIC COMPLEXES Background The existence of carbenes was first reported in the organic literature by Skell and Sandler in 1958 with the insertion of a carbon atom into two doubly bonded carbons as a gene ral way of extending a carbon chain (Figure 5 1). 123 Transition metal carbene species ensued in 1964 when Fis c her e t. al reported the first stable W carbene complex ; complexes of Cr and Mo followed soon after (Figure 5 2) 124,125 Figure 5 1. Insertion of a carbon atom into an alkene Figure 5 2. First stable transition metal carbene complexes In 1968, seminal papers published by fele and Wanzlick reported the first N heterocycl ic carbene (NHC) transition metal complexes synthesized from imidazolium salts and transition metal precursors 126,127 These carbenes were found to be very stable, nucleophilic donor ligands because of electron donation from the N atoms into the empty p orbital of the carbene carbon. 126 Based off these initial studies, Arduengo and coworkers were, in 1991, able to synthesize and obtain the crystal structure of the first free carbene ligand (Fi gure 5 3) and t heir concept for NHC preparation has been used extensively since then. 128 The steric hindrance of the bulky adamantyl groups

PAGE 85

85 effectively protect ed the free NHC from dimerization thereby allowing the ligand both thermodynamic and kinetic stability in th e absence of water and oxygen. 129 Figure 5 3. Formation of the first stable, free N heterocyclic carbene N heterocyclic carbenes continue to be one of the most ve rsatile and ubiquitous types of ligands in organometallic chemistry and are considered to be one of the few, generally useful ligands for catalysis (along with phosphines and cyclopentadienyls) because they are non toxic chemically tunable and diverse and form highly stable metal complexes 130,131 NHC ligands form bonds with metal centers via three orbital contributions (Figure 5 4, adapted from Cavallo et al. ) 129 however, th e stability afforded by these carbene s is mainly because they form strong donor bonds to metals which in turn helps the complex resist decomposition during catalysis. Figure 5 4. Orbital interactions of M NHC bonds

PAGE 86

86 Their usefulness is c ompounded by the fact that they a re easily synthesized; either comp rehensively, startin g from aryl or alkyl amines along with glyoxal and formaldehyde or via alkyl or aryl ation of imidazoles 132,133 In addition, a wide array of substituents may be attached to the azole ring allowing for tunability of the steric and electronic properties of the carbene. These appended functional groups can allow the metal NHC complex to fun ction in a way akin to an enzymatic catalyst where the entire complex works as a whole to perform the reaction not just the met al. The ability for NHCs to function in this way stems from the fact that the N substituents point towards the metal (as opposed to away from the metal as in the case of Cp and PR 3 groups) and as such have the ability to interact with the metal during the reaction. This has allowed organo metal lic NHC complexes bearing functional group substituents to d emonstrate a variety of properties such as molecular recognition, metal binding, cooperative acid/base and redox effects. 131,134 In the last two decades, there have been many organometallic NHC complexes prepared for use in a variety of catalytic transformations. 129,135,136 The most notable and well known of these are the r uthenium olefin metathesis catalysts 129,136 139 the p alladium cross coupling catalysts, 129,135,136,139 141 and the iridium hydrogenation catalysts 135,136,139,142,143 ( Figure 5 5 ). In each of these cases, the traditional phosphine ligands on Ru, Pd and Ir were replaced by NHCs resulting in superior catalytic abilities of the new complex es 129

PAGE 87

87 F igure 5 5 Examples of well known Ru, Pd and Ir complexes bearing NHC ligands Although the previous examples are some of the most common cases of NHC applications in the literature, there are numerous co mplexes that utilize these ligands for catalytic transformations 127,135,136,139,144 These transformations include hydrogenation and hydroformylation reactions by rhodium complexes 145,146 redox reactions with iridium complexes 143,147,148 hydrofunctionalization of unsaturated hydrocarbons using gold complexes, 149,150 oxidation of alcohols u sing palladium, 151,152 cyclo and conjugate addition reactions using copper, 153,154 cyclization reactions with cobalt 155,156 as we ll as an assortment of miscellaneous reactions. 135 (Figure 5 6). Figure 5 6 E xamples of NHC complexes While the majority of complexes are monometallic, there are several cases of bimetallic NHC complexes, the majority of these being complexes that are homobimetallic in nature. 157 165 Examples of these range from the carbene linked silver cyclophane complex (Figure 5 7 ) prepared by Youngs et al. in 2001 157 that has

PAGE 88

88 application as a carbene transfer reagent to the copper complexes by Eastham et al. 158 that exist because of the chelating prope rties of the ligands (Figure 5 7 ). Figure 5 7 Homobimetallic NHC complexes of Ag, and Cu Some of the most useful of these types of homobimetallic NHC complexes are the Ru complexes introd uced by Herrmann et al. for use in ring opening metathesis polymerization (ROMP) and ring closing metathesis polymerization (RCM) reactions (Figure 5 8 ) 159,160 These complexes were generated quantitatively and ea sily by reacting previously prepared Ru alkylidene complexes with [RuCl 2 ( p cymene)] 2 and showed substantially higher catalytic ability than the monometallic precursors. Stemming from this work, Delaude and Demonceau et al. were also able to synthesize bim etallic Ru complexes (Figure 5 8 ) using the Ru( p cymene) dimer and Ru NHC precursors for use as catalysts in RCM, ROMP and atom transfer radical polymerization (ATRP) reactions 160 162 Unlike their homobimetallic counterparts, heterobimetallic NHC complexes are rare; one example (Figure 5 9) was published in 1995 by fele and coworkers, where the two metals are linked via a backbone functionalized NHC ligand. 127,166 When a cyclic voltammo gram was obtained, this complex demonstrated no interaction between

PAGE 89

89 the Os IV and Cr 0 metal centers. R ecently, both Braunstein and Cowie were able to independently synthesize heterobimetallic complexes where the metals are connected throug h an NHC linker (Figure 5 9). 167,168 Following a different tactic, Bertrand and coworkers were also able to synthesize heterobimetallic NHC complexes (Figure 5 9) in high yield where one metal was bound directly t o the carbene while the second metal was bound to a diphenylphosphino substituent on the carbene backbone. 169 Figure 5 8 Homo bimet allic Ru NHC complex es for olefin metathesis reactions Figure 5 9 Heterobimetallic NHC complexes

PAGE 90

90 Although these complexes currently have no specific use, they are unique in the fact that unlike the homobimetallic complexes where the metals are general ly linked through bridging halide ligands and the NHC ligand is typically attached to only one metal, these complexes utilize the NHC as a chelating or bridging ligand and as such promote the idea that these ligands can indeed replace the more common phosp hine in modern organometallic complexes. Preliminary Ru/Rh Complex A preliminary electrochemical test was performed to judge the ability of a Ru/Rh complex to oxidize ethanol. The control Ru/Rh complex 51 was synthesized by reacting the previously prepa red CpRu(PPh 3 1 dppm)Cl precursor with [RhCl(CO) 2 ] 2 in CH 2 Cl 2 at room temperature for one hour (Figure 5 10) The solvent was removed and the crude mixture eluted from a silica column using 10% MeOH in Et 2 O to give a yellow solid in 32% yield. Mass spectrometry and NMR data confirmed presence of the bimetallic Ru/Rh complex 51 Figure 5 10. Synthesis of Ru/Rh complex 51 Cyclic voltammograms were performed to observe the potentials at which each metal oxidized and also to test the activity of the complex towards ethanol (Figure 5 11) Literature citations offer conflicting data for the oxidation of Rh complexes; depending on the type of complex, Rh may either be electrochemically oxidized as a single two electron oxidation (Rh(I/III)) or it may be oxidized as tw o successive one electron oxidations ( Rh(I/II) followed by Rh(II/III)). 170 173 The cyclic voltammogram of the Ru/Rh

PAGE 91

91 complex 51 indicated four oxidation potentials at 0.723, 1.03, 1.29 and 2.23 V vs. NHE (Figure 5 11) Based on previously acquired CVs (Appendix) the waves at 1.03 and 2.23 V vs. NHE were ascribed to the Ru(II/III) and Ru(III/IV) oxidations, respectively. Subsequently, the remaining anodic waves at 0.723 and 1.29 V vs. NHE were attributed to the Rh (I/II) and Rh(II/III) oxidations respectively. Figure 5 11. C yclic voltammograms of complex 51 in 0.1 M TBAT/DCE; 50 mVs 1 scan rate; glassy carbon working electrode; Ag/Ag + reference electrode After addition of ethanol, an increase in current was obs erved at the potential corresponding to the Ru(III/IV) oxidation (Figure 5 11). This current increase helps to confirm the assignment s of the Ru and Rh oxidations since it is known that Ru(III/IV) is active for ethanol oxidation. Although the cyclic volt ammogram indicates that Rh is not active towards ethanol, the Ru/Rh complex has the ability to oxidize ethanol at higher potentials.

PAGE 92

92 Attempted Synthesis of a Ru/Rh Complex bearing an NHC Ligand Ligand Design and Synthesis N heterocyclic carbene ligands m ay be functionalized i n a variety of ways (Figure 5 12 ) whereby the central five membered carbene backbone may either be completely saturated as in the c ase of imidazolin 2 ylidenes, partially unsaturated such as the imidazole 2 ylidenes or attached to an aromatic ring like the benzimidazol 2 ylidenes. In addition, carbons C 4 and C 5 may have a range of substituents from hydrogen or halogen atoms to even bulkier groups such as alk oxides or various alkyl moieties. Most commonly, the nitrogen substituents of the imidazole backbone are varied to present either a symmetrical or an unsymmetrical NHC ( Figure 5 12 ). This range of substituents and degrees of unsaturation to the backbone offer many advantages by allowing NHCs to have varying degrees of basicity as well as electronic and steric tunability within the ligand class 129 Figure 5 12 Functionalization of the NHC backbone The major design strategy employed for the ligand was the ability of the complex to dissolve completely in ethanol while at the same time retaining its stability and

PAGE 93

93 catalytic ability under electrochemical conditions. Consequently, the conventional imidazol 2 ylidene backbone was chosen since the free ligand is approximately 20 kcal mol 1 more stable than the saturated imidazolin 2 ylidene, and is more basic than the benzimidazol 2 ylidene ( Figure, 5 12 ) 127 These feature s should allow for ease of attachment to the metal, less likelihood of ligand dimeri zation and continued solubility in ethanol ( Figure 5 13 ) In order to provide an additional element to aid stability during catalysis the initial ligand design was aimed towards a bidentate chelating ligand whereby a pair of carbenes would be linked to e ach other through a bridge and could each attach to the metal. In order to facilitate this che lation, it was decided that a single nitrogen of the imidazole backbone would be substituted while the other would be used for attachment to the bridge ( Figure 5 13 ) Figure 5 13 Design of the NHC ligand Preparation of the ligan d may be accomplished either by synthesizing the imidazole ligand or by functionalizing the already prepared imidazole with an aryl group. Using a method proposed by Zhang et al. 132 (Figure 5 1 4 ) the imidazole may be synthesized by stirring a para substituted aniline with glyoxal in methanol overnight, adding ammonium chloride and formaldehyde to the mixture and refluxing f or one hour then finally adding phosphoric acid and refluxing overnight. The mixture is then neutralized to pH 9, extracted with diethyl ether and eluted from a silica column using an ethyl acetate/petroleum ether solvent mixture to yield the substituted imidazole in range

PAGE 94

94 from 4 67 % As an alternative, a method reported by Viel in 2005 may be used 133 starting from imidazole and the para substituted iodobenzene and refluxing these in a xylene mixture containing copper (I) triflate, dibenzylideneacetone (dba), 1,10 phenanthroline, and cesium carbonate (F igure 5 1 4 ) until almost complete disappearance of the N H peak is observed by FTIR (usually after 72 96 h). The reaction mixture is diluted with ethyl acetate and filtered through a silica plug with additional ethyl acetate before being concentrated and elut ed from a silica column using 2 % methanol in dichloromethane to yield beige or grey white solids in 17 41%. Figure 5 1 4 Synthesis of the imidazolium ligand Since yields were still relatively low with this reaction (starting material could be reco vered from the column) and xylene did not dissolve the reagents, 1,2 dichlorobenzene was used in attempts to attain better yields. Unfortunately, using dichlorobenzene as solvent complicated the work up by dissolving the metal salts

PAGE 95

95 thereby leading to mor e difficult column elution. Mass spectrometry also indicated the tendency of dichlorobenzene to be non innocent under the reaction conditions; when synthesis of the bromo substituted imidazole was attempted using the Viel method in dichlorobenzene, data s howed t he presence of both the chloro and the bromo substituted phenyl imidazole. After these attempts, benzene was subsequently used as solvent but resulted in only slightly better yields than xylene. Once the para substituted phenyl imidazoles were o btained, they were linked through a methylene bridge using the method proposed by Herrmann in 1999 by reacting methylene bromide with the imidazole in a pressure vessel at 150 C for two days using THF as the solvent (Figure 5 1 4 ) 140,174,175 During this time, a grey or beige solid appears in the initial yellow solution indicating formation of the imidazolium bromide salt which can be filtered, washed with THF and collected in 40 65 % yields. When synthesis of the ethylene bridged imidazolium bromide salt was attempted, no precipitate appeared after the required time and the reaction was assumed to be a failure. The attempt to produce the ortho phenyl bridged imidazolium using 1,2 dibromobenzene also resulted in no salt precipitate even after reacting for 7 days with excess 1,2 dibromobenzene, however, when the solvent was removed and a proton NMR obtained of the residue it was observed t hat the imidazolium proton on C2 had shifted from 7.83 ppm to 8.29 ppm an in dication of attachment to the phenyl bridge. Although proton NMR data obtained in deuterated DMSO for the methylene and ortho phenylene bridged imidazolium ions clearly indicates presence of the imidazolium proton on C2 data from mass spectrometry are mi sleading. N o molecular ion peak is ever observed for the ligand ; instead peaks are observed for the fragmented starting

PAGE 96

96 imidazole and also for the dimerize d imidazole after loss of the C2 protons. This dimerization is known to occur by the Wanzlick equil ibrium 131,176 (Figure 5 1 5 ) b ut generally arises from the free carbene ligand i nstead of the imidazolium salt, as such, mass spectro metry conditions act to promote this dimerization. Figure 5 1 5 Wanzlick equilibrium and dimerization of the imidazolium ion during mass spectro metry Attempt s to Bind the NHC to Ru thenium Since the ligand was synthesized in order to allow for good solubility in ethanol, the initial design involved either attachment of both meta ls to the ligand or attachment of ruthenium to the ligand and attachment of rhodium to ruthenium via a bridge ( Figure 5 16 ; showing attachment of the two metals to the NHC ligand and omitting all other ligands ). This design would allow ruthenium to be bon ded directly to the electron rich, stable carbene moiety while rhodium could be attached either to a separate area of the ligand or to ruthenium via a halide bridge both of which would allow the metals to be in close proximity to each other which is known to aid electrocatalytic oxidation Simple MM2 energy minimization s using the Chem3D Pro 12.0 program w ere performed with the metals attached to the ligand in the ways previously discussed. R esults indicated that attempting to bind rhodium to ruthenium v ia a halide bridge would produce a

PAGE 97

97 complex higher in energy (147.627 kcal mol 1 ) compared to binding rhodium oxidatively to the phenyl substituent of the ligand (33.146 kcal mol 1 ) Consequently, attempts were made to first bind ruthenium to the bidentate carbene in order to test the electrochemical ability of this compound before synthesizing the more stable Ru/Rh complex Figure 5 1 6 Possible bonding modes of Ru and Rh to the NHC ligand and basic MM2 energy minimization of each There are several es tablished methods for attaching an NHC ligand to a transition metal precursor: 144 Reaction of the free NHC with an organometallic precursor Reaction of an electron rich olefin dimer with an organometallic fragment Reaction of an imidazolium salt with a basic t ransition metal salt Transmetallation using Ag(I) NHC precursors Reaction of an azolium salt with metal precursors using basic phase transfer catalytic (PTC) conditions

PAGE 98

98 The imidazolium 1 (4 methoxyphenyl) 1H imidazole, was chosen for the following reactio ns because it could be produced in high yields compared to the other examples. The first few attempts to attach Ru to the imidazolium salt (Table 5 1 entries 1 7 ) were based on published procedure s by Foth, Albrecht Basato, and Herrmann 177 180 which involved r eaction of the imidazolium ion in THF or DMSO with a base followed by the transition metal precursor. When THF was used as the solvent (Table 5 1, entries 1 3) work up involved filtration of the reaction m ixture followed by addition of hexane to the filtrate in order to precipitate the compound. When DMSO was used as the solvent (Table 5 1, entries 4 7), work up involved using CH 2 Cl 2 to dilute the reaction mixture before precipitation with Et 2 O. Entries 1 5 and 7 all produced dark brown precipitates (not in line with the yellow/orange precipitate one would expect from a Ru(II) complex) and determination of metal attachment to the ligand by 1 H NMR spectroscopy was complicated because of the ability of deute rated DMSO to dissolve any salts or unreacted imidazolium present (DMSO d 6 was used as the NMR solvent because it allows for elucidation of the imidazolium proton since proton exchange does not occur in aprotic solvents). Entry 6 produced a pale yellow be ige solid that was insoluble in CH 2 Cl 2 Acetone, MeOH, CH 3 CN, EtOH, EtOAc and hexane but soluble in DMSO and a 1 H NMR spectrum indicated loss of the imidazole proton. When a mass spectrum was obtained, the ions found were consistent with a 174 Da compound indicating the presence of 1 (4 methoxyphenyl) 1H imidazole however, no isotope patte rn containing Ru was observed. A single attempt at transmetallating the ligand from Ag onto Ru (Table 5 1, entry 8) was attempted using the literature procedure reported by Xue et al. in 2010. 181 The

PAGE 99

99 ligand and Ag 2 O were suspended in methanol and stirred at room temperature for 2 h before filtration through a Celite pad to result in a yellow solution. The solvent was evaporated under reduced pressure, the chlorocarbonyl rut henium polymer was added along with CH 2 Cl 2 and the mixture stirred at room temperature overnight. The mixture was then filtered through a Celite pad to remove the silver salt and the solvent removed from the brown filtrate. After TLC analysis, the crude mixture was eluted from a silica column (MeOH:CH 2 Cl 2 1/40 v/v) resulting in a yellow solution that after NMR analysis showed absence of imidazolium protons, however, mass spectrometry data once again indicated ions at 174 Da and no isotope patterns contain ing ruthenium. Table 5 1. Attempts to deprotonate the imidazole using a base Trial Solvent Temperature Time Base / Transmetallating Agent Organometallic Precursor 1 THF r.t. 12 h Cs 2 CO 3 [(COD)RuCl 2 ] x 2 THF r.t. 12 h Cs 2 CO 3 [RuCl 2 (CO) 2 ] x 121 3 THF r.t. 12 h Cs 2 CO 3 [Ru 2 (O 2 CCH 3 ) 4 Cl] 182 4 DMSO r.t. 12 h Cs 2 CO 3 [RuCl 2 (CO) 2 ] x 5 DMSO ~ 50 C 12 h Cs 2 CO 3 [(COD)RuCl 2 ] x 6 DMSO ~ 50 C 12 h Cs 2 CO 3 [RuCl 2 (CO) 2 ] x 7 D MSO ~ 50 C 12 h Cs 2 CO 3 [Ru 2 (O 2 CCH 3 ) 4 Cl] 8 MeOH r.t. 12 h Ag 2 O [RuCl 2 (CO) 2 ] x 9 THF 78 C 12 h KN(Si(CH 3 ) 3 ) 2 [RuCl 2 (CO) 2 ] x 10 IPA 110 C 48 h Na 2 CO 3 RuCl 3 .xH 2 O 11 THF 30 C 12 h KN(Si(CH 3 ) 3 ) 2 [RuCl 2 ( p cymene)] 2 Since reactions using Cs 2 CO 3 to deproto nate the ligand failed, an attempt was made using the strong er base KN(Si(CH 3 ) 3 ) 2 (Table 5 1, entry 9) following a procedure by Veige et al. published in 2009. 183 The ligand was suspended in THF and cooled to

PAGE 100

100 78 C before KN(Si(CH 3 ) 3 ) 2 dissolved in THF was added dropwise. The mixture was then warmed and stirred at room temperature for 10 minutes before it was recooled to 78 C and the Ru precursor dissolved in THF was added. The reaction mixture was stirred at 78 C overnight then filtered and precipitated from hexane t o produce a dirty yellow powder however the 1 H NMR spectrum indicated continued presence of the i midazole peak at 10.75 ppm and mass spectrometry once again showed an ion corresponding to 174 Da but no ruthenium isotope pattern. An additional attempt to attach ruthenium to the ligand (Table 5 1, entry 10) was performed as a variation of a procedure published by Salzer et al. in 2004. 184 Hydrated ruthenium trichloride was heated to reflux with Na 2 CO 3 and the ligand in isopropanol overnight after which the mixture was observed to remain dark brown/black indicating no reaction had occurred. Pyridine was then added to the refluxing mixture as a means of providing ruthenium with a second neutral coordinating liga nd in order to help with the reduction of Ru(III) to Ru(II), however, even after reflux ing for an additional day, the reaction mixture was still observed to be brown /black and after removal of the solvent, it was found that the mixture was only partially s oluble in deuterated DMSO most likely indicating the presence of polymeric ruthenium and Ru(0). Using the procedure described by Veige and coworkers, 183 1 (4 bromophenyl) 1H imidazole was deprotonated with KN(Si(CH 3 ) 3 ) 2 in THF followed by addition of [RuCl 2 ( p cymene)] 2 (Table 5 1, entry 11) After reaction, the brownish red solution was observed to change color to brown as the xylene/CO 2 bath was removed and the solution attained room temperature. Silver hexafluorophosphate was added in the hopes of providing a

PAGE 101

10 1 counter ion that could help prevent halt decomposition of the compound however after removal of the silver salt b y filtration through Celite the resultant filtrate was still brown. A s a final at tempt, 1 (4 bromophenyl) 1H imidazole was reacted with magnesium turnings in diethyl ether in order to form the Grignard reagent This was done with the hopes of binding ru thenium to the phenyl substituent of the imidazolium salt. Unfortunately, even after addition of iodine and refluxing the mixture overnight the magnesium was unconsumed and the Gr ignard reaction never initiated leaving a black residue in the S chlenk flas k Attempt to Bind the NHC to Rhodium Since attachment of the ligand to ruthenium was proving to be difficult, a reaction was performed to bind rhodium based on the procedure described by Veige and coworkers. 183 The methoxy substituted ligand was first deprotonated with KN(Si(CH 3 ) 3 ) 2 in THF ( Figure 5 17 ) then [(COD)RhCl] 2 added and the mixture stirred overnight at 78 C. The reaction yielded a yellow brown solid that wa s eluted from silica using MeOH in DCM at an initial concentration of 1/40 v/v and increasing polarity until the yellow orange compound was eluted in 47 % yield. A 1 H NMR obtained in deuterated chloroform and deuterated DMSO indicated the absence of imidazolium protons as well as peaks corresponding to attachment of the cyclooctadiene ligand Mass spectrometry indicated the presence of 1 (4 methoxyphenyl) 1H imidazole (174 Da) as in previous experiments along with the presence of cyclooctadiene (108 Da) and more interestingly, peaks at 501 and 503 Da corresponding to the isotope pattern of rhodium attached to the defragmented ligand and bromine after loss of the methyl substituents (Figure 5 18).

PAGE 102

102 Figu re 5 1 7 Attachment of the imidazolium salt to Rh Figure 5 18 Fragmentation pattern observed for 5 2 by LR MS Although it i s difficult to determine the correct structure of the rhodium NHC complex from the mass spectrum, previous accounts have indicat ed that short imidazole linkers (e.g. methylene or ethylene ) favo r the non chelated bis Rh product 131,185 This however is slightly incongruous with NMR data of 5 2 which indicates only one COD ligand per NHC ( as opposed to two COD ligands per NHC for the bis Rh bound compound). In addition 1 H NMR data indicate the COD olefin peaks a re unsymmetrical in regards to their attachment thereby displaying peaks at 4.51 and 3.49 ppm each integrating to 2 protons. Unfo rtunately a crystal structure could not be obtained for this compound.

PAGE 103

103 Electrochemical e xperiments using the NHC Rh c omplex 5 2 A cyclic voltammogram was obtained of 5 2 in TBAT/DCE scanning at 50 mVs 1 and o xidation peaks were observed at 0.914 V and 1.1 32 V vs. NHE ( Figure 5 19 ) correspond ing to the two irreversible one electron oxidations of Rh. L iterature reports offer conflicting information about Rh oxidations ; data exists showing either a single two electron oxidation for the Rh(I/III) couple, or, two single electron oxidations for the Rh(I/II) and Rh(II/III) couples, however, the CV of 5 2 indicates two single electron oxidations for Rh(I/II) and Rh( II/III). 170,171 Successive scans of the complex indicated the merging of the oxidation peaks (Figure 5 1 9 insert ) as well as a large decrease in the observed current even while stirring the solution between each scan. These observations point towards decomposition of the complex under CV conditions and large de positions of this new (oxidized) compound on the electrode. Although the electrode was cleaned before addition of ethanol to the cell, no increase in current was observed at the voltage corresponding to the Rh oxidations indicating that Rh bound to the NH C ligand is not active towards ethanol and as such is unable to oxidize ethanol on its own. While it is known that Ru is able to oxidize ethanol at the Ru(III/IV) potential, this potential tends to be relatively higher than the non catalyst assisted oxida tion of ethanol and as such electrooxidation of ethanol with a Ru/Rh system will not be viable unless this Ru(III/IV) potential is decreased in the bimetallic complex

PAGE 104

104 Figure 5 1 9 Cyclic voltammograms of complex 5 2 in 0.1 M TBAT/DCE; 50 mVs 1 scan r ate; glassy carbon working electrode; Ag/Ag + reference electrode Insert showing deposition of oxidized 5 2 on working electrode during successive scans. Attempt to Bind the NHC to Rhodium and Ruthenium Since results pointed towards Rh attachment to the N HC, a reaction was performed to specifically bind Rh to the ligand followed by reaction of this intermediate 168 procedure 183 The imidazolium was suspended in THF, cooled to 78 C then dissolved KN(Si(CH 3 ) 3 ) 2 w as added and the mixture stirred for an additional 30 minutes before warming to room temperature ( Figure 5 20 ) After the mixture had attained room temperature, it was recooled a solution of [(COD)RhCl] 2 was slowly added and the reaction allowed to stir overnight at 78 C until eventually attaining room temperature. The brown mixture was cooled a third time and an additional equivalent of + Complex 52 Complex 52 scan 2 Complex 52 scan 4

PAGE 105

105 KN(Si(CH 3 ) 3 ) 2 was added in order to repeat the previous steps, this time adding the [RuCl 2 ( p cymene)] 2 dimer (Figure 5 20 ). Figure 5 20 Attempt to bind the NHC ligand to Rh and Ru consecutively After completion, the crude mixture was filtered to produce a fine beige precipitate corresponding to the dimerized NHC according to the Wanzlick mechanism ( Figure 5 15 ) an d a dark red brown filtrate. The solvent was removed from the filtrate and a 1 H NMR spectrum acquired indicating loss of the imidazolium protons Mass spectrometry indicated a molecular ion peak corresponding to [ ( C 21 H 22 N 4 O 2 )+RhC 8 H 10 ] + indicative of the ligand and COD attached to Rh. Additional isotopic patterns indicated the presence of ruthenium however because of the complicated fragmentation patterns observed for these compounds, X ray crystallographic data will be necessary to elucidate the struc ture of the complex formed. These efforts are underway. Summary and Perspectives There are many literature examples of NHC ligands bound to ruthenium, however, these complexes contain monodentate NHC ligands; i n addition, there are several reported insta nces of chelating bidentate NHC ligands bound to metals such as Pd, Ir

PAGE 106

106 and Rh. Unfortunately, although it seems possible to bind Ru to a chelating bidentate NHC ligand, attempts were unsuccessful using the procedures discussed. It is plausible that t he f ailure of these reactions stemmed from a number of reasons including the design of the chelating imidazolium the ease of dimerization of the free carbene or even the absence of the correct stabilizing ligands around the ruthenium center. In each case, sy nthetic reactions to bind Ru to the imidazolium ligand usually resulted in formation of a black/brown solid which indicated the absence of Ru(II), 1 H NMR spectra were complicated due to the presence of salts in the mixture and mass spectra showed fragmenta tion of the ligand but no ruthenium ionization patterns indicating attachment of the ligand to the metal The ability of rhodium to bind to chelating bidentate NHC ligands is known and the reaction of rhodium with the imidazolium ligand resulted in attac hment of the metal to the carbene to produce compound 5 2 Although lit erature reports point towards a bis Rh type of attachment to the ligand, a 1 H NMR spectrum obtained for the isolated compound showed only one cyclooctadiene ligand per NHC ligand pointi ng towards only one Rh atom bound to the NHC. A mass spectrum indicated the presence of Rh, COD and the NHC ligand in the sample however, because no crystal s were obtained for X ray analysis, a definite structure of the compound was elusive. A cyclic vol tammogram of the compound indicated two oxidations that may correspond to Rh(I/II) and Rh(II/III) at 0.914 and 1.132 V vs. NHE respectively however when ethanol was added to the cell, no increase in current was observed at these potentials indicating tha t the Rh bound NHC c ompound obtained from this reaction is incapable of oxidizing ethanol.

PAGE 107

107 The attempt to bind both rhodium and ruthenium to the ligand resulted in a dark red brown compound, which displayed no imidazolium peaks in the 1 H NMR and which in dicated the presence of the [(C 21 H 22 N 4 O 2 )+RhC 8 H 10 ] + ion as well as the presence of ruthenium isotopic patterns in the mass spectrum. Attempts at obtaining a crystal structure of this compound are underway. Since it seems difficult to attach Ru directly to the NHC ligand and because the Rh NHC compounds synthesized above need to be eluted from columns in order to purify them, the recently reported complexes by Bertrand et al. (Figure 5 9) are an interesting possibility for future work in this area. The a dvantages of their complexes are high yields and also ease of transmetallating the metal precursor (Pd or Au) onto the carbene. In addition the ir complexes are substituted with PPh 2 which has previously been shown to bind to Ru rapidly so should allow fo r attachment of both metals in a relatively easy fashion. Based on the preliminary electrochemical data obtained with these complexes it is still uncertain whether th is type of Ru/Rh bimetallic system may catalyze the electrooxidation of ethanol. Since it has been shown that the Rh NHC is incapable of oxidizing ethanol and since it is known that Ru(III/IV) can oxidize methanol and ethanol at much higher potentials the Ru/Rh complex would need to allow for lower Ru(III/IV) potentials in order for this t ype of bimetallic system to be viable. Additionally, synthesis of the Rh/Pt or Rh/Pd bimetallic systems may be attempted as well as trimetallic Ru/Rh/Pt or Ru/Rh/Pd type systems.

PAGE 108

108 CHAPTER 6 CONCLUSION S Activity of Ru/Pt Complex 1 for Water Oxidation S ince t he first generation complexes 1 20 are all active for methanol and ethanol oxidation (both highly polar substrates) a cyclic voltammogram was performed using Ru/Pt complex 1 to test for activity towards water oxidation (Figure 6 1) A fter addition of 10 L of water to the system, no increase in current was observed indicating that complexes 1 20 are not active catalysts for water oxidation and that the current increases observed during bulk electrolyses experiments are due solely to alcohol oxidation Figure 6 1. C yclic voltammograms of complex 1 in 0.1 M TBAT/DCE; 50 mVs 1 scan rate; glassy carbon working electrode; Ag/Ag + reference electrode Proposed Mechanisms of Alcohol Oxidation with Heterobimetallic Catalysts T he oxidation s of methanol and ethan ol in heterogeneous systems have been studied thoroughly enough in the past two decades to propose the mechanism s by

PAGE 109

109 which they undergo their respective oxidations. In spite of this not much is known about electrochemical alcohol oxidation in homogeneous systems and as such a mechanistic pathway for alcohol oxidation has never been proposed for homogeneous heterobimetallic complexes. Over the years, several electrochemical experiments performed in the McElwee White group have been aimed towards elucidati ng a me chanism. The following is an account of these experiments as well as the proposed mechanisms of electrocatalytic alcohol oxidation based on these experiments and preceding literature discussion T he cyclopentadienyl ligand has remained constant f or all of the previously prepared complexes 1 28 (Figures 2 2 2 4, 2 5 and 2 7 ) both because it readily coordinates to Ru and can provide stability and also because it possesses the ability to ring slip and move between an 5 and 3 coordination to Ru. Th is change in hapticity can effectively open a coordination site on Ru for either the alcohol or water to bind. In order to test whether the Cp ligand changes hapticity during electrolysis, a complex bearing the indenyl ligand in place of the cyclopentadie n yl ligand was synthesized. The indenyl ligand has the ability to aromatize the six membered ring as it moves from 5 to 3 coordination; it should therefore undergo the change in hapticity at a faster rate than the Cp ligand resulting in higher current efficiency and more turnovers during alcohol oxidation if the mechanism involves ring slip to open a coordination si te on Ru. The ( 5 indenyl)Ru(PPh 3 )( I)( dppm)PdCl 2 complex ( 7 Figure 2 2 Appendix) was synthesized in excellent yield by reacting ( 5 indenyl)Ru(PPh 3 )( 1 dppm)I with (COD)PdCl 2 in CH 2 Cl 2 at room temperature. Cyclic voltammograms indicated good curren t increases at the Pd(II/IV) and Ru(III/IV) oxidations after addition of both

PAGE 110

110 methanol and ethanol to the system while bulk electrolysis data from the oxidation of methanol demonstrated a current efficiency of 27% compared to 42% for the corr esponding Cp c ontaining complex. These results demonstrate that the change in ligand from Cp to indenyl does not enhance methanol oxidation and as such, change in hapticity of the Cp ligand is not a viable mechanistic step during the oxidation of alcohols using the pre viously prepared heterobi metallic complexes Another ligand used extensively in the heterobimetallic catalyst design is the triphenylphosphine (PPh 3 ) ligand. Phosphines are typically employed in catalysis because of their ability to dissociate from the metal center thereby opening a coordination site f or initiation of catalysis Previous studies in the McElwee White group have explored the effect of triphenylphosphine addition during cyclic voltammetry and bulk electrolysis with methanol When added to systems containing Cp(CO)Ru( I)( dppm)PtI 2 ( 15 ) r esults indicated the ability of the phos phine to stabilize the complex thereby increasing the total observed current during cyclic voltammetry (Figure 6 2a) as well as the evolution of oxidized products duri ng bulk electrolysis (Figure 6 2 b ) and the subsequent increase in current efficiency of the catalyst Since previous experiments have indicated that these c omplexes lose the CO ligand during catalysis resulting in degradation of the complex, 88 the addition of PPh 3 and the resulting increase in current efficiency prove that presence of the PPh 3 ligand improves catalytic behavior As such, the complexes do not gain an open coordination site by loss of the phosphine ligand during catalysis Bulk electrolysis experiments performed with meth anol demonstrated that the bridged Ru/Pd bimetallic complexes Cp(PPh 3 )Ru( Cl)( dppm)PdCl 2 ( 2 ) and

PAGE 111

111 Cp(PPh 3 )Ru( I)( dppm)PdCl 2 ( 4 ) both attain higher current efficienci es (25% and 42%, respectively) after 100 C of charge passed than either the equimolar mixture of CpRu(PPh 3 ) 2 Cl and Pd(COD)Cl 2 (18%) or the monometallic CpRu(PPh 3 ) 2 Cl (12%) complex. 77,83 These results indica ted that both metals are necessary and also that the electrooxidation is more efficient in the bridged structure of the catalyst than the equimolar model system. It can further be assumed that the phosphine bridge acts to keep the metals in close proximit y during catalysis. Figure 6 2. Effect of adding P Ph 3 to complex 15 (a) Cyclic voltammograms of complex 15 in 0.1 M TBAT/DCE; 50 mVs 1 scan rate; glassy carbon working electrode; Ag/Ag + reference electrode. (b) Product evolution from bulk electroly sis of MeOH using complex 15 before and after PPh 3 addition to the system ; 1.9 V vs. NHE; 0.7 M TBAT/DCE; 0.35 M MeOH added. Since neither change in hapticity of the Cp ring nor loss of a phosphine ligand was shown to occur during the electrocatalytic exp eriments, the only remaining alternative for acquiring an open coordination site is dissociation of the bridging Cl ligand. It can be hypothesized that t his dissociation may occur at ei ther metal (Figure 6 3) with no change in the oxidation state of the m etal. If dissociation occurs at Pt as shown (Figure 6 3), a mechanistic pathway can be proposed based on studies by Goldberg et al. which explore both the oxidative addition of C H bonds to Pt(II) and the reductive

PAGE 112

112 elimination of methoxid e and hydroxide s pecies from Pt (IV) either by C H, C C C O or hydride elimination reactions. 101,186 190 Further more a mechanistic pathway may also be proposed if dissociation occurs at the Ru center; this can be based on studies by Meyer et al. using Ru bipyridyl complexe s for water oxidation. 89,91,92,191 197 Figure 6 3. Dissociation of the bridging Cl from either Ru or Pt Using methanol as the substrate and the Pt site as the catalyst, the alcohol may be oxidized first by coo rdination then O H oxidative addition to Pt(II) (Figure 6 4). At this point, the Pt Cl bond forming the bridge may either break to form a trigonal bipyramidal intermediate or the reaction may proceed with the octahedrally coordinated Pt(IV); both have bee n shown to occur to varying degrees (Figure 6 4) 186 188,198 Figure 6 4. Oxidative addition of CH 3 OH to Pt(II) As a non productive step in the pathway, C O reductive elimination may occur to reform CH 3 OH howe ver, as was shown in several examples by Milstein and Ozerov, hydride elimination is more likely to occur in solutions containing methanol 199 201 This

PAGE 113

113 elimination yield s formaldehyde; the two electron oxidation product from methanol oxidation (Figure 6 5) and t he Pt(II) c atalyst may then be regenerated by loss of gaseous hydrogen. Figure 6 5. hydride elimination and formation of formaldehyde Once formed, formaldehyde may then be further oxidized by nucleophilic attack of water present in the solution to produce form ic acid, the fo ur electron oxidation product, or it may undergo rapid acid catalyzed condensation with methanol to produce the observed two electron oxidation product, dimethoxymethane (DMM) In addition, it may react again with Pt by migratory insert ion into the Pt OCH 3 bond to produce the Pt OCH 2 OCH 3 intermediate which can also undergo hydride elimination to produce methyl formate (MF), the observed four electron oxidation product (Figure 6 6)

PAGE 114

114 Figure 6 6. Oxidation of formaldehyde to methyl formate As an alternative step, methanol may also react nucleophilically with Pt OCH 3 to p roduce P t OH and Me 2 O (Figure 6 7); however, the presence of Me 2 O has not been confirmed in solution (either by GC or FTIR analysis) Figure 6 7. Formation of dimethyl ether Although oxidation of the alcohol is typically discussed as occurring on Pt, Ru has also been known to perform the reaction ; this may occur if the Cl bridge dissociates from Ru to open a coordination site for the substrate to bind (Figure 6 3). The proposed mechanism by which this occurs involves i nitial coordination of the substr ate followed by oxidation of Ru(II) to Ru(III) with loss of a single electron and proton (Figure 6 8) 89,91,92,191 193 This intermediate can then undergo further oxidation to form the Ru(IV) oxo intermediate with an additional electron and CH 3 + released, however, this may be unlikely to occur in solutions containing high methanol concentration. In these solutions, methanol may act as a nucleophile and attack the Rh OCH 3 intermediate once again producing Me 2 O as we ll as an electron and proton (Figure 6 8) although this has never been proven and if possible, formation of Me 2 O may be a terminal step during the oxidation reaction In addition, similar to oxidation on Pt, hydride

PAGE 115

115 elimination may also occur to release formaldehyde fr om the Ru center, leaving a Ru H species however there is no precedence for this in the literature Figure 6 8. Oxidation of CH 3 OH on Ru In their water oxidation studies, Meyer et al. suggest a further oxidation of Ru(IV) to Ru(V) as a step to regenerate the catalyst. 89,92,191 193 This may also be proposed with methanol oxidation when water is present in homogeneous systems (Figure 6 9). Although there have never been studies to prove the exis tence of a Ru(V) species during homogeneous methanol oxidation, it is known from hetero geneous systems that Ru can act to dehydrogenate water and produce oxygen or a hydroxide species. Figure 6 9. Regeneration of Ru(II) via a Ru(V) intermediate Additi onal studies recently published by Meyer et al. in 2011 have also indicated that both the Ru(IV) and the Ru(V) oxo species have the ability to oxidize benzyl alcohol to produce benzaldehyde with the Ru(V) oxo species being significantly more reactive

PAGE 116

116 for t his reaction. 197 Although the Meyer study incorporates a phosphonate buffer to help with this transformation, it can be hypothesized that in electrolytic solutions where methanol is the solvent, methanol may also act as both a n acid and a base. Using the Ru(IV) oxo species to demonstrate this reaction, it may either be formed via the initial oxidation of methanol (Figure 6 8) or through the two electron oxidation of water as reported by Meyer (Figure 6 10) The Ru(IV) oxo spe cies may then form a transition structure (Figure 6 10) with both methanol and methoxide (or hydroxide) which should undergo rapid transformation to produce formaldehyde, a Ru (II) OH species and methanol (or water). The Ru(II) OH species may then be elect rochemically oxidized to produce the Ru(III) hydroxide species and reenter the catalytic cycle. Figure 6 10. Alternate mechanism of CH 3 OH oxidation on Ru Previous experiments in the McElwee White group offer conflicting rationales for proving or dispr oving the Pt v ersus Ru mechanism s ; when the analogous Fe and Ru carbonyl complexes ( 15 20 ) are used during methanol oxidation, the Fe complexes are observed to be inadequate compared to the Ru complexes. This observation may point towards oxidation of the alcohol on the Ru center, however, it should also be noted that these complexes have been proven to lose the CO ligand during catalysis so this

PAGE 117

117 observation could also be from the ease of CO loss from Fe versus Ru. In addition, the first generation Ru/Pt and Ru/Pd complexes ( 1 5 ) have all been shown to oxidize methanol with relatively equal current efficiencies; the Pt complexes being slightly more active while the Pd complexes are more robust. This conclusion also seems to point towards oxidation of the alcohol on Ru, however, this hypothesis is inconclusive since both Pt and Pd a re used in catalysi s for C H activation reactions and as such, during electrocatalytic methanol oxidations, there should be no large differences in reactivity between the two met als In conclusion, although several mechanisms for alcohol oxidation using the heterobimetallic catalysts may be hypothesized, additional studies need to be performed before a final mechanism can be determine d.

PAGE 118

118 CHAPTER 7 EXPERIMENTAL SECTION General Procedures Reactions were performed under argon atmosphere utilizing standard Schlenk line techniques. Tetrahydrofuran (THF) and diethyl ether (Et 2 O) were dried by distillation from sodium/benzophenone ketyl. Methanol (MeOH) and ethanol (EtOH) were doub le distilled using magnesium shavings and iodine. Acetone was distilled from powdered sodium iodide. Dichloromethane (CH 2 Cl 2 ) and hexane were dried on an MB raun solvent purification system using an activated alumina column and copper catalyst. Anhydrous acetonitrile (CH 3 CN) dimethylformamide (DMF) and dichloroethane (DCE) were purchased from Sigma Aldrich and used without further purification. NMR solvents were purchased from Cambridge Isotope Laboratories, Inc. and degassed via freeze pump thaw cycle s. All solvents w ere stored over type 3 or 4 molecular sieves. 1 H and 31 P NMR spectra were obtained on a Varian Mercury 300 MHz spectrometer and referenced either to tetramethylsilane or the solvent ( 1 H), and to 85 % H 3 PO 4 or PF 6 ( 31 P). 202,203 Mass spectrometric analyses were carried out either on a Agilent 6210 TOF MS or on a ThermoScientific Trace GC DSQ mass spectromet er Heterobimetallic complexes 1 28 were prepared as previously described. 77 82,84 All other starting materials were purchased in reagent grade purity and used without further purification. Electrochemi cal Consi derations Electrochemical experiments were performed at room temperature in a glove box using an EG&G PAR mode l 263A potentiostat/galvanostat and a three compartment H cell separated by a medium porosity sintered glass frit. All potentials are reported

PAGE 119

119 ve rsus NHE and referenced to Ag/Ag + The E o values for the ferrocenium/ferrocene couple in DCE and EtOH were +0.72 V and +0.68 V vs. NHE, respectively. 204 C yclic voltammograms were recorded at a scan rate of 50 mVs 1 in 3.5 mL of 0.1 M tetrabutylammonium trifluoromethanesulf onate (TBAT) / DCE using a BASi 3 mm diameter glassy carbon working electrode. When oxidation waves could not be observed using TBAT as the electrolyte, tetrabutylammonium tetrafluoroborate (TBA(BF 4 )) or tetrabutylammonium hexafluorophosphate (TBAH ) was used instead. The reference electrode consisted of a silver wire immersed in an acetonitrile solution containing 0.01 M AgNO 3 and 0.1 M TBAT (or TBA(BF 4 )) These were contained in a 75 mm glass tube fitted at the bottom with a porous Vycor tip. A platin um flag was used as the counter electrode. Constant potential (bulk) electrolysis was performed similarly using solutions of 0.1 M (TBA(BF 4 )) dissolved in either DCE or EtOH and a cylindrical Duocel reticulated vitreous carbon electro de ( 60 ppi, 10 mm dia meter) was substituted in place of the glassy carbon electrode. Product Analysis Analyses of the e lectrolysis products were analyzed on a Shimadzu GC 17A gas chromatograph For experiments using DCE as the electrolytic solvent, either a 15 m 0.45 mm All tech EC TM WAX (1.0 m) column or a 30 m 0.32 mm Alltech EC TM WAX (0.25 m) column on fused silica was used. When EtOH was the solvent, a 15 m x 0.53 mm DB 5 fused silica column (J & W Scientific, 1.5 m film) was utilized instead The column s were atta ched to the injection port with a neutral 5 m 0.32 mm AT TM WAX deactivated guard column. During bulk electrolysis, a known amount of n octane was

PAGE 120

120 used as an internal standard and identification was confirmed by comparison of the retention times of authe ntic samples to the oxidation products. FT IR analyse s of the headspace gases produced during electrolyse s were performed using a Perkin Elmer Spectrum One FT IR Spectrometer. A modified airfree glass cell bound by two NaCl plates was used to contain the he adspace gas taken from the cell Identification of products was confirmed by comparison to authentic samples Synthe tic Procedures cis (bpy) 2 Ru Cl 2 2H 2 O. 97,98 Dimethylformamide (100 mL) was added to a Parr vessel containing RuCl 3 2 O (2.64 g, dipyridyl (3.15 g, 20.2 mmol) and the dark yellow brown solution was refluxed overnight. The resulting dark red black solution was cooled to room temperature, acetone added and the product precipitated in a fr eezer. The product was filtered, washed with water (2 10 mL) and Et 2 O ( 5 20 mL) and the b lack crystals dried (3.94 g, 75 %). 1 H NMR ( DMSO d 6 ): 9.98 ( d, 2H ), 8.64 ( d, 2H ), 8.49 ( d, 2H), 8.07 ( d, 2H ), 7.77 (t, 2H ), 7.68 (dd, 2H) 7.51 (d, 2H), 7.10 (t, 2H). cis (bpy) 2 RuI 2 2H 2 O 99 Methanol (100 mL) w as added to a Parr vessel charged with cis (bpy) 2 Ru Cl 2 2H 2 O (1.49 g, 2.86 mmol) and NaI (4.41 g, 29.4 mmol) and the purple black solution refluxed until a brown black solution appeared (48 h). After cooling to room temperature, a black precipitate appeare d which was filtered and washed several times with cold methanol to yield 1.89 g (94 %) of cis (bpy) 2 RuI 2 2H 2 O 1 H NMR ( DMSO d 6 ): 10.4 ( d, 2H), 8.66 ( d, 2H ), 8.50 (d, 2H), 8.07 (d, 2H), 7.74 (m, 4H), 7.55 (d, 2H), 7.18 (t, 2H).

PAGE 121

121 [ cis (bpy) 2 Ru I( 1 dppm)]PF 6 ( 29 ). AgPF 6 (0.613 g, 2.43 mmol) was added to a Schlenk flask containing cis (bpy) 2 RuI 2 2H 2 O (1.71 g, 2.43 mmol) and dissolved with acetone (50 mL). The dark purple brown solution was allowed to stir at room temperature until precipitation of AgI (~ 3 h) at which point the solution was cannula filtered into a second S chlenk flask containing dppm (0.936 g, 2.43 mmol) and the resu lting solution was allowed to stir at ambient temperature (48 h). The reaction was monitored by TLC, which indicated a mixture of unreacted dppm along with both the 1 dppm product and the 2 dppm side product. The solvent was removed and the crude produ ct purified on an alumina column using CH 3 CN/toluene (2:3 v/v). The eluant was removed on a rotovap then the red orange product was re dissolved in acetone, precipitated with Et 2 O, filtered, washed with Et 2 O and allowed t o dry (0.937 g, 36 %). 1 H NMR (ace tone d 6 bpy), 8.24 6.34 (m, 31H, bpy + Ph), 4.07 (dd, 1H, PPh 2 C H 2 PPh 2 ), 2.44 (dd, 1H, PPh 2 CH 2 PPh 2 ). 31 P{ 1 H} NMR (acetone d 6 37.9 (d, Ru P Ph 2 CH 2 PPh 2 ), 28.4 (d, RuPPh 2 CH 2 P Ph 2 ) 145 ( sept et P F 6 ). HRMS (FAB) calcd for C 45 H 38 P 2 N 4 IRu 925.0667 [M] + found 925.0684. [ cis (bpy) 2 Ru( I)( dppm)PdCl 2 ]PF 6 ( 30 ). [ cis (bpy) 2 RuI( 1 dppm)] + PF 6 (0.199 g, 0.186 mmol) and Pd(COD)Cl 2 (0.055 g, 0.19 mmol) were dissolved in acetonitrile (30 mL) and allow ed to stir at room temperature overnight. The solvent was removed under reduced pressure and the product recrystallized using acetone/Et 2 O, filtered and washed with Et 2 O t o give 0.184 g (79 %) of powdery rust colored solid. 1 H NMR (CD 3 8.35 6.17 (m, 30H, bpy + Ph), 3.85 (m, 1H, PPh 2 C H 2 PPh 2 ), 3.32 (m, 1H,

PAGE 122

122 PPh 2 CH 2 PPh 2 ). 31 P{ 1 H} NMR (CD 3 38.1 (dd, Ru P Ph 2 CH 2 PPh 2 Pd), 17.1 (dd, RuPPh 2 CH 2 P Ph 2 Pd), 145 (sep tet, P F 6 ). HRMS (FAB) calcd for C 45 H 38 P 2 N 4 IRuPdCl 2 1102.9077 [M] + found 1102.9022. [ cis (bpy) 2 Ru( I)( dppm)PtCl 2 ]PF 6 ( 31 ). [ cis (bpy) 2 RuI( 1 dppm)] + PF 6 (0.214 g, 0.200 mmol) and Pt(COD)Cl 2 (0.075 g, 0.20 mmol) were dissolved in acetonitrile (40 mL) and allowed to stir at room temperature overnight. The solvent was removed under reduced pressure and the product recrystallized using CH 2 Cl 2 /Et 2 O, filtered and washed with Et 2 O to give 0.221 g ( 82 %) of powdery orange solid. 1 H NMR (CD 3 1H, bpy), 9.32 (d, 1H, bpy), 8.63 (t, 1H, bpy), 8.55 (d, 1H, bpy), 8.4 6.33 (m, 30H, bpy + Ph), 5.96 (m, 2H, PPh 3 ), 3.88 (m, 1H, PPh 2 C H 2 PPh 2 ), 3.40 (m, 1H, PPh 2 C H 2 PPh 2 ). 31 P{ 1 H} NMR (CD 3 P Ph 2 CH 2 PPh 2 Pt), 7.31 (dd, RuPPh 2 C H 2 P Ph 2 Pt), 145 (septet, P F 6 ). HRMS (FAB) calcd for C 45 H 38 P 2 N 4 IRuPtCl 2 1190.9679 [M] + found 1190.9379. cis (bpy) 2 FeCl 2 2H 2 O. Dimethylformamide (50 mL) was added to a schlenk flask containing FeCl 3 (1.00 g, 6.19 mmol), LiCl (0.019 g, 0.44 dipyridyl (1.96 g, 12.5 mmol) and the dark yellow brown solution refluxed overnight. The resulting dark red black solution was cooled to room temperature, and the product precipitated using diethyl ether. The product was filtered, washed with water (2 10 mL) and Et 2 O (5 20 mL) and the red powder dried (0.619 g, 21 %). 1 H NMR (DMSO d 6 8.22 (br s, 2H), 7.52 (br s, 2H), 7.39 (br s, 2H). HRMS (FAB) calcd for C 20 H 16 N 4 Cl 2 Fe 184.0357 [M] 2+ found 184.0352. [ cis (bpy) 2 FeCl(PPh 3 )]PF 6 ( 32 ). A pressure vessel charged with cis (bpy) 2 FeCl 2 2H 2 O (0.09 g, 0.189 mmol), TlPF 6 (0 .252 g, 0.721 mmol) and approximately

PAGE 123

123 40 mL acetonitrile was allowed to stir at room temperature overnight until a dark red solution was observed along with precipitation of the TlCl salt. The mixture was then filtered and the filtrate reacted with PPh 3 ( 0.257 g, 0.978 mmol) overnight at room temperature. The solvent was removed under reduced pressure and the crude product re dissolved in a small amount of acetone then precipitated with ether. The red powder was then washed with ether and hexane to remov e excess unreacted PPh 3 (0.132 g, 90% ). 1 H NMR ( DMSO d 6 8.87 7.28 (m, 31H, bpy + Ph) 31 P{ 1 H} NMR (DMSO d 6 ): 25.7 (s Ru P Ph 3 ), 145 ( septet P F 6 ). bpy FeCl 2 2H 2 O Anhydrous FeCl 2 (1.00 g, 7.89 mmol) was placed in a Schlenk flask along with 30 mL et hanol and the mixture warmed to 60 C to dissolve the iron bipyridyl (0.411 g, 2.63 mmol) dissolved in 10 mL ethanol was added dropwise to the solution and the resulting red orange mixture allowed to react for an additional 10 minutes. The solvent was then filtered off and the red orange powder washed with a mixture of ethanol (4 15 mL) and hydrochloric acid (4 1 drop) to yield 0.634 g (8 5 %) of rust orange powder. 1 H NMR (CH 3 CN d 3 8.14 (br s, 1H), 7.4 4 (br s, 2H). HRMS (FAB) calcd for C 10 H 8 N 2 Cl 2 Fe 246.9720 [M] + found 246.9734. [Pd(C 8 H 12 OCH 3 )(bpy)]PF 6 103 To a Schlenk flask co ntaining a bright yellow suspension of Pd(COD)Cl 2 (0.297 g, 1.04 mmol) in methanol (60 mL), AgNO 3 (0.362 g, 2.13 mmol) was added and the mixture allowed to stir until precipitation of AgCl (~ 1h). The ivory solution was cannula filtered into a second Schl dipyridyl (0.169 g, 1.08 mmol) and after dissolution, excess NH 4 PF 6 was added to precipitate the product. The precipitate was filtered, washed several times with

PAGE 124

124 methanol and dried to yield 0.225 g ( 40 %) of powdery white [Pd(C 8 H 1 2 OCH 3 )(bpy)] + PF 6 1 H NMR (DMSO d 6 (d, 2 H, bpy), 8.54 (d, 2 H, bpy), 8.36 (t, 2 H, bpy), 7.87 (t, 2 H, bpy), 6.13 (m, 1H, CH=CH ), 5.85 (m, 1H, CH=CH ), 3.65 (m, 1H, C H OCH 3 ), 3.34 (s, 3H, CHO C H 3 ), 2.99 (m, 1H, Pd C H ), 2.74 1.64 (m, 8H, C H 2 ) 31 P{ 1 H} N MR (DMSO d 6 ): 145 ( septet, PF 6 ). Cp Ru(PPh 3 ) ( Cl)( dppm)Pd(bpy) ( 36 ). To a Schlenk flask containing CpRu(PPh 3 )( 1 dppm)Cl (0.106 g, 0.125 mmol) and [Pd(C 8 H 12 OCH 3 )(bpy)] + PF 6 (0.070 g, 0.128 mmol) was added dichloromethane (40 mL). The yellow solution was allowed to stir at room temperature overnight before the solvent was removed in vacuo The crude product was recrystallized using CH 2 Cl 2 /Et 2 O, then filtered and washed with Et 2 O to yield 36 as a beige powder (0.104 g, 75 %). 1 H NMR (DMSO d 6 (d, 1H, bpy), 8.59 (d, 1H, bpy), 8.37 ( t 1H, bpy), 7.82 (t, 1H bpy ), 7.55 7.15 (m, 35H, P Ph ), 6.78 (br s 2H, bpy), 5.94 (br s, 2H, bpy), 5.58 ( m, 1H, PPh 2 C H 2 PPh 2 ), 4.98 (s, 5H, C 5 H 5 ) 3. 89 (s, 1 H, PPh 2 C H 2 PPh 2 ). 31 P{ 1 H} NMR (DMSO d 6 P P h ), 0.42 (d ). [ 5 C 5 H 4 CH 2 CH 2 N (CH 3 ) 2 HCl ]Ru(PPh 3 ) ( Cl)( dppm)Pd(bpy) ( 37 ). CH 2 Cl 2 (20 mL) was used to dissolve [ 5 C 5 H 4 CH 2 CH 2 N(CH 3 ) 2 HCl]Ru(PPh 3 )( 1 dppm)Cl (0.112 g, 0.117 mmol) and [Pd(C 8 H 12 OCH 3 )(bpy)] + PF 6 (0.063 g, 0.12 mmol) in a schlenk flask. The y ellow orange solution was stirred at room temperature overnight then the solvent removed under reduced pressure. The crude product was recrystallized with CH 2 Cl 2 /Et 2 O, filtered and washed with Et 2 O to yield beige powd er (0.091 g, 65 %). 1 H NMR (DMSO d 6 ): 7.11 (m, 43H, PPh), 6.78 (m, 1H, bpy), 6.18 (m, 2H, bpy), 5.08 (s, 2H, C 5 H 4 ), 4.91 (m, 1H, PPh 2 C H 2 PPh 2 ), 4.86 (s, 2H, C 5 H 4 ) 4.16 (s, 1H, PPh 2 C H 2 PPh 2 ), 3.01 (m, 2H,

PAGE 125

125 C 5 H 4 CH 2 C H 2 NH 2 ), 2.50 (s 6H, C 5 H 4 CH 2 CH 2 N Me 2 ), 1.85 (m, 2H, C 5 H 4 C H 2 CH 2 NH 2 ) 31 P{ 1 H} NMR (DMSO d 6 Ru P Ph ), 0.32 (d ). C 5 H 5 Na. 111 Excess dicyclopentadiene (700 mL) was added to freshl y cut sodium (29.4 g, 1.28 mol) under an argon atmosphere. The mixture was subsequently heated at 170 C for 6 h or until all the sodium had been consumed and no further hydrogen evolution was observed. The reaction mixture was then filtered and the whit e precipitate washed several times with hexane and dried (111.58 g, 99 %). 1 H NMR (CD 3 CN 5.60 (s, 5H). C 5 H 5 CH 2 CH 2 NH 2 112 A solution of sodium cyclopentadienide (53.2 g, 0.602 mol) in THF (100 mL) was added dropwise to a Schlenk flask containing a stirred suspension of 2 chloroethylamine hydrochloride ( 32.1 g, 0.2 77 mol) in THF (1 00 mL) at 0 C. The pale pink mixture was allowed to attain room temperature and then refluxed for 4 h, resulting in a pale yellow solution containing a whit e precipitate. The solvent was removed at room temperature under reduced pressure, the precipitate dissolved in water (200 mL) and extracted with Et 2 O (3 50 mL) and pentane (3 50 mL). The organic layer was dried (MgSO 4 ), filtered and concentrated on a rota ry evaporator at ambient temperature to give a viscous yellow liquid Purification was achieved by fractional distillation under vacuum (0.01 mmHg, 30 C) to yield the mixture of three isomers as a viscou s colorless liquid (5.33 g, 18 %). 1 H NMR (CD Cl 3 6.03, 5.89, 5.44 (m, CH=CH of all 3 isomers), 3.15 (m, C 5 H 5 CH 2 C H 2 NH 2 ), 2.92 and 2.84 (d, J = 2 Hz, C 5 H 4 C H 2 ) 2.82 2.43 (m, C 5 H 5 CH 2 C H 2 NH 2 ), 2.19 2.08 (m, C 5 H 5 C H 2 CH 2 NH 2 of all three isomers), 1.62 1.41 (m, C 5 H 5 CH 2 CH 2 N H 2 of all three isomers ).

PAGE 126

126 [ 5 C 5 H 4 CH 2 CH 2 NH 2 ]Ru(PPh 3 ) 2 Cl ( 38 ). To an ethanolic solution (100 mL) containing hydrated ruthenium trichloride (2.77 g, 13.4 mmol) and triphenylphosphine (14.0 g, 53.5 mmol) was added freshly distilled 2 cyclopentadienylethylamine in ethanol (20 mL). The brown mixture was refluxed for 48 h, until a dark yellow brown solution formed. The solvent was removed and the solid purified via several crystallizations using methylene chloride/diethyl ether to yield a bright oran ge powdery product (1.05 g, 10 %) 1 H NMR (CDCl 3 5 H 4 CH 2 CH 2 N H 2 ), 7.81 7.09 (m, 30H, P( C 6 H 5 ) 3 ), 4.16 (br s, 2H, C 5 H 4 ), 3.33 (br s, 2H, C 5 H 4 ), 3.30 3.21 (m, 2H, C 5 H 4 CH 2 C H 2 NH 2 ), 2.86 2.74 (m, 2H, C 5 H 4 C H 2 CH 2 NH 2 ) 31 P{ 1 H} NMR (CDCl 3 (s, Ru P Ph 3 ). HRMS (FAB) calcd for C 43 H 40 P 2 NClRu 734.1686 [M] + found 734.1669. [ 5 C 5 H 4 CH 2 CH 2 NH 2 ]Ru(PPh 3 ) ( 1 dppm) Cl ( 39 ). To a 500 mL Schlenk flask charged with [ 5 C 5 H 4 CH 2 CH 2 NH 2 ]Ru(PPh 3 ) 2 Cl (0.450 g, 0.584 mmol) and bis(diphenylphosphino) methane (0. 450 g, 1.17 mmol) was added THF ( 60 mL) and the resulting orange mixture allowed to stir at ambient temperature for 1 4 days before the solvent was removed under reduced pressure The crude product was purified on a silica column using NEt 3 :EtOH:Et 2 O (2.1:5.4:92.5 v/v/v) as eluant. T he r esulting bright yellow solid was dried overnight to yield 0. 334 g ( 64 % ) of 39 1 H NMR (CDCl 3 (br s, 2H, C 5 H 4 CH 2 CH 2 N H 2 ), 7.88 6.82 (m, 35H, P( C 6 H 5 ) 3 + ( C 6 H 5 ) 2 PCH 2 P( C 6 H 5 ) 2 ), 4.61 (br s, 2H, C 5 H 4 ), 4.17 (br s, 2H, C 5 H 4 ), 3.71 3.61 (m, 2H, Ph 2 P C H 2 PPh 2 ), 3.40 3.24 (m, 2H, C 5 H 4 CH 2 C H 2 NH 2 ), 2.92 2.75 (m, 2H, C 5 H 4 C H 2 CH 2 NH 2 ). 31 P{ 1 H} NMR (CDCl 3 43.5 (dd, Ru P Ph 2 CH 2 PPh 2 ), 37.5 (dd, Ru P Ph 3 ), 27.5 (dd, RuPPh 2 CH 2 P Ph 2 ). HRMS (FAB) calcd for C 50 H 47 P 3 NClRu 856.1973 [M] + found 856.1905.

PAGE 127

127 [ 5 C 5 H 4 CH 2 CH 2 NH 2 ]Ru( PPh 3 ) ( Cl)( dppm) PdCl 2 ( 40 ). In a 100 mL S chlenk flask, [ 5 C 5 H 4 CH 2 CH 2 NH 2 ]Ru(PPh 3 )( 1 dppm)Cl (0.152 g, 0.171 mmol) and Pd(COD)Cl 2 (0.065 g, 0.173 mmol) were dissolved in methanol (30 mL). The dark red solution was allowed to stir at room temperature f or 1 h at which point, a fine salt like precipitate was observed. This precipitate was filtered off and the solvent removed from the filtrate. The crude product was recrystallized from dichloromethane/hexane to yield 0.077 g (42%) of red orange solid. 1 H NMR (CDCl 3 50 6.67 (m, 35H, P( C 6 H 5 ) 3 + ( C 6 H 5 ) 2 PCH 2 P( C 6 H 5 ) 2 ), 6.06 (br s, 2H, C 5 H 4 ), 5.73 5.54 ( m 2 H, Ph 2 P C H 2 PPh 2 ), 4.53 (br s, 2H, C 5 H 4 ), 3.84 2.97 (m, 2H, C 5 H 4 CH 2 C H 2 NH 2 ), 2.85 2.21 (m, 2H, C 5 H 4 C H 2 CH 2 NH 2 ). 31 P{ 1 H} NMR (CDCl 3 ): 54.3 (dd, Ru P Ph 2 C H 2 PPh 2 ), 37.6 (dd, Ru P Ph 3 ), 21.4 (dd, RuPPh 2 CH 2 P Ph 2 ). HRMS (FAB) calcd for C 50 H 47 NP 3 Cl 3 RuPd 1034.0384 [M] + found 1034.0337. [ 5 C 5 H 4 CH 2 CH 2 NH 2 3 ) 2 Cl ( 41 ). A solution of 38 (0.562 g, 0.730 mmol) in methylene chloride (50 mL) was washed twice wi th 1 M HCl (2 20 mL) and dried with MgSO 4 The solvent was then removed and the crude product eluted from a silica column using N E t 3 /EtOH/Et 2 O (4.9/0.1/5.0 v/v/v) as eluant The solvent was removed to yield 0.249 g (42%) of dark yellow product 1 H NMR ( CDCl 3 2H, C 5 H 4 CH 2 CH 2 N H 2 ), 7.80 6.89 (m, 30H, P( C 6 H 5 ) 3 ), 4.40 (br s, 2H, C 5 H 4 ), 4.10 (br s, 2H, C 5 H 4 ), 3.35 3.28 (m, 2H, C 5 H 4 CH 2 C H 2 NH 2 ), 2.95 2.87 (m, 2H, C 5 H 4 C H 2 CH 2 NH 2 ). 31 P{ 1 H} NMR (CDCl 3 P Ph 3 ). [ 5 C 5 H 4 CH 2 CH 2 NH 2 3 )( 1 dppm)Cl ( 42 ). To a 200 mL S chlenk flask charged with [ 5 C 5 H 4 CH 2 CH 2 NH 2 3 ) 2 Cl (0.249 g, 0.323 m mol) and bis(diphenylphosphino) methane (0.258 g, 0.670 mmol) was added THF (40 mL) and the

PAGE 128

128 resulting orange mixture was allowed to stir at ambient tem perature for 16 days before the solvent was removed under reduced pressure and the crude product purified on a silica column using a mixture of NEt 3 :EtOH:Et 2 O (2.1:5.4:92.5 v/v/v) as eluant The resulting yellow solid was dried overnight to yield 0.255 g (85 %) of 42 1 H NMR (CDCl 3 ): 7.79 6.80 (m, 35H, P( C 6 H 5 ) 3 + ( C 6 H 5 ) 2 PCH 2 P( C 6 H 5 ) 2 ), 4.26 (br s, 2H, C 5 H 4 ), 4.12 (br s, 2H, C 5 H 4 ), 3.89 (t, 1H, Ph 2 P C H 2 PPh 2 ), 3.84 (t, 1H, Ph 2 P C H 2 PPh 2 ), 3. 75 3.6 5 (m, 2H, C 5 H 4 CH 2 C H 2 NH 2 ), 2.90 2.77 (m, 2H, C 5 H 4 C H 2 CH 2 NH 2 ). 31 P{ 1 H} NMR (CDCl 3 ): 0 (dd, Ru P Ph 2 CH 2 PPh 2 ), 38. 7 (dd, Ru P Ph 3 ), 27.2 (dd, RuPPh 2 CH 2 P Ph 2 ). [ 5 C 5 H 4 CH 2 CH 2 NH 2 3 )( Cl)( dppm)PdCl 2 ( 43 ). The reaction was performed as for 40 using [ 5 C 5 H 4 CH 2 CH 2 NH 2 3 )( 1 dppm)Cl (0.255 g, 0.275 mmol) and Pd(COD)Cl 2 (0.079 g, 0.27 5 mmol). The red orange solution (dissolved in 20 mL CH 2 Cl 2 ) was allowed to stir at room temperature for 24 hours then the solvent was removed and the crude product was recrystallized from dichloromethane/hexan e to yield 0.095 g (31 %) of rust orange solid 1 H NMR (CDCl 3 ): 8.17 6.75 (m, 35H, P( C 6 H 5 ) 3 + ( C 6 H 5 ) 2 PCH 2 P( C 6 H 5 ) 2 ), 6.30 (br s, 4H, C 5 H 4 ), 6.08 (m, 1H, Ph 2 P C H 2 PPh 2 ), 4.56 (m, 1H, Ph 2 P C H 2 PPh 2 ), 2. 96 2.84 (m, 2H, C 5 H 4 CH 2 C H 2 NH 2 ), 2.63 2. 55 (m, 2H, C 5 H 4 C H 2 CH 2 NH 2 ). 31 P{ 1 H} NMR (CDCl 3 Ru P Ph 2 CH 2 PPh 2 ), 37.7 (dd, Ru P Ph 3 ), 2 1 8 (dd, RuPPh 2 CH 2 P Ph 2 ). C 5 H 5 CH 2 CH 2 OH 112 CpNa (54.8 g, 0.622 mol) was dissolved in THF (pale dusky pink solution) and cooled to 0 C. 2 chloroethanol (25.2 g, 0.314 mol, 21 mL) was dissolved in a second Schlenk flask containing THF and the solution added dropwise to the CpNa solution, immediately resulting in a pink solution with white precipitate. The ice bath was removed a nd the mixture refluxed for 6 h to produce a bright yellow

PAGE 129

129 solution with a fine white precipita te present. T he mixture was then cooled to room temperature, the solvent removed the s olid dissolved in water and extracted with Et 2 O (5 x 75 mL ) then pentane (3 x 75 mL ). Th e organic extracts were combined, dried with MgSO 4 and concentrated on the roto vap to give a brown orange very viscous oil (11.51 g). This was then purified using fractional distillation (0.01 mmHg, 33 C) to yield 2.66 g ( 8 %) of the compound. 1 H NMR (CDCl 3 4 7 6. 0 4 (m, CH=CH), 3. 79 (m, C 5 H 5 CH 2 C H 2 OH ), 2.9 9 ( q C H 2 of C 5 H 5 ), 2.94 (q, C H 2 of C 5 H 5 ) 2. 71 2. 62 (m, C 5 H 5 C H 2 CH 2 OH ), 1.89 (m, C 5 H 5 CH 2 CH 2 O H ). [ 5 C 5 H 4 CH 2 CH 2 OH ]Ru(PPh 3 ) 2 Cl ( 44 ). Hydrated RuCl 3 (2.6 1 g, 12.6 mmol) was refluxed with four equivalents of PPh 3 (13.2 g, 50.3 mmol) and four equivalents of C 5 H 5 CH 2 CH 2 OH (8.3 1 g, 75.4 mmol) over the course of two days. The reaction was monitored by 31 P NMR which indicated presence of the product (40.5 ppm) as well as oxidized phosphine and other unknown impu rities. At this time, the ethanol was removed from the resulting yellow brown solution and the sticky residue analyzed by TLC. The first attempt to purify the mixture on a silica column using 3% MeOH in Et 2 O resulted in the removal of unreacted RuCl 3 but no additional separation. Two subsequent alumina columns were performed, first using 3% MeOH in Et 2 O and then 3% MeOH in CH 2 Cl 2 Complex 44 could be obt ained as an orange solid in 29 % yield (2.81 g). 1 H NMR ( CDCl 3 7.71 7. 10 (m 30H, P( C 6 H 5 ) 3 ), 4. 10 (br s, 2H, C 5 H 4 ), 3. 87 ( q 2H, C 5 H 4 CH 2 C H 2 OH ), 3.33 (br s, 2H, C 5 H 4 ), 2.58 ( t 2H, C 5 H 4 C H 2 CH 2 O H) 2.54 ( t 1 H, C 5 H 4 CH 2 C H 2 O H ) 31 P{ 1 H} NMR (CDCl 3 40.5 (s, Ru P Ph 3 ). [ 5 C 5 H 4 CH 2 CH 2 OH ]Ru(PPh 3 )( 1 dppm)Cl ( 45 ). Compound 44 (2.81 g, 3.65 mmol) was reacte d with two equivalents of dppm (2.81 g, 7.30 mmol) in THF (red

PAGE 130

130 orange solution) at room temperature for 5 days. After a 31 P NMR spectrum was obtained of the crude mixture the solvent was removed and the crude product eluted from a silica column starting w ith 8% EtOAc in CH 2 Cl 2 followed by 40% EtOAc in CH 2 Cl 2 and finally with addition of a small amount of methanol to remove the yellow band in 45% yield (1.47 g). For the crude sample: 1 H NMR (CDCl 3 7.76 6.93 (m, 35H, P( C 6 H 5 ) 3 + ( C 6 H 5 ) 2 PCH 2 P( C 6 H 5 ) 2 ), 4. 3 7, 4.19 (br s, 2 H, C 5 H 4 ), 3,87 (t 2H, C 5 H 4 CH 2 C H 2 OH), 3.79 (m, 2H, Ph 2 P C H 2 PPh 2 ), 3.62, 2.65 (br s, 1 H, C 5 H 4 ), 2,60 (m, 2H, C 5 H 4 C H 2 CH 2 OH ) 1.26 ( dd, 2H, Ph 2 P C H 2 PPh 2 ) 31 P{ 1 H} NMR (CDCl 3 44.1 ( dd, Ru P Ph 2 CH 2 PPh 2 ), 38.4 (dd, Ru P Ph 3 ), 27. 4 (dd, RuPPh 2 CH 2 P Ph 2 ). For the purified sample after column elution: 31 P{ 1 H} NMR (CDCl 3 49.4 ( t Ru P Ph 3 ), 3.0 (d, Ru P Ph 2 CH 2 P Ph 2 ). [ 5 C 5 H 4 CH 2 CH 2 OC 5 H 9 O ]Ru(PPh 3 ) 2 Cl ( 46 ). Hydrated RuCl 3 (1.56 g, 7.50 mmol) wa s refluxed with four equivalents of PPh 3 (7.93 g, 30.3 mmol) and thirteen equivalents of C 5 H 5 CH 2 CH 2 O C 5 H 9 O ( 20.3141 g, 105 mmol) over the course of two days. The reaction was monitored by TLC then t he ethanol was removed and the sticky yellow brown residue purified. The first attempt to purify the mixture on alumina using 40 % CH 2 Cl 2 in hexane resulted in the removal of unreacted RuCl 3 but no additional separation. An additional alumina column w as performed, starting from 40% CH 2 Cl 2 in hexane then graduall y increasing polarity to 69 % CH 2 Cl 2 in hexane Compound 4 6 could be obtained as a sticky viscous liquid in very large yields (13.5 g) due to persisting unreacted ligand 31 P{ 1 H} NMR (CDCl 3 40.5 (s, Ru P Ph 3 ). [ 5 C 5 H 4 CH 2 CH 2 OC 5 H 9 O ]Ru(PPh 3 )( 1 dppm)Cl ( 4 7 ). Compound 4 6 ( 13.5 g, 15.7 mmol) was reacted with two equivalents of dppm ( 12.3 g, 32 mmol) in THF ( yellow

PAGE 131

131 brown solution) at room temperature for 10 days. The reaction was monitored by TLC then the solvent was removed and the crude product eluted fr om a silica column starting with 40% CH 2 Cl 2 in hexane then gradually increasing polarity to 5% MeOH in CH 2 Cl 2 to give a pale yellow solid. For the crude sample: 31 P{ 1 H} NMR (CDCl 3 44.1 (dd, Ru P Ph 2 CH 2 PPh 2 ), 38.4 (dd, Ru P Ph 3 ), 27. 5 (dd, RuPPh 2 CH 2 P Ph 2 ). For the purified sample after column elution: 31 P{ 1 H} NMR (CDCl 3 49.4 ( t Ru P Ph 3 ), 3. 1 (d, Ru P Ph 2 CH 2 P Ph 2 ). ( 5 C 5 H 5 ) Ru (PPh 3 )( Cl)( dppm)Rh ( CO ) Cl ( 51 ). The CpRu(PPh 3 1 d ppm)Cl precursor (0.300 g, 0.354 mmol) was reacted with [RhCl(CO) 2 ] 2 (0.070 g, 0.180 mmol) in CH 2 Cl 2 at room temperature for one hour. The solvent was then removed under reduced pressure and the crude product eluted from a silica column using 10% MeOH in Et 2 O to yield a yellow solid (0.114 g, 32 %). 1 H NMR (CDCl 3 8.04 6.14 (m, 35H, P( C 6 H 5 ) 3 + ( C 6 H 5 ) 2 PCH 2 P( C 6 H 5 ) 2 ), 4.60 (s, 5 H, C 5 H 5 ), 3.07 2.90 (m, 2H, Ph 2 PC H 2 PPh 2 ) 31 P{ 1 H} NMR (CDCl 3 45.8 (dd, Ru P Ph 2 CH 2 PPh 2 ), 37. 9 (dd, Ru P Ph 3 ), 32.2 (dd, RuPPh 2 CH 2 P Ph 2 ). HRMS ( ESI TOF) calcd for C 49 H 4 2 OCl P 3 RuRh 979.0241 [ M] + found 979.0210 C 3 H 3 N 2 C 6 H 4 OCH 3 A Schlenk flask was charged with 1H imidazole (3.08 g, 45.3 mmol), 1 iodo 4 methoxybenzene (6.83 g, 28.6 mmol ), copper (I) trif late (0.785 g, 1.52 mmol ) d ibenzylideneacetone (0.377 g, 1.61 mmol ) 1,10 ph enanthroline (5 .33 g, 29.6 mmol ) and benzene. The orange brown mixture was heated to reflux and the reaction monitored by TLC on silica with 2% MeOH in CH 2 Cl 2 After 5 days, the reaction was stopped, the mixture diluted with EtOAc then filtered through a silica plug us ing additional EtOAc to elute the yellow solution. The solution was concentrated under

PAGE 132

132 reduced pressure and the crude product eluted from a silica column using 2% MeOH in CH 2 Cl 2 The product was collected as a very pale yellow solution and the solvent wa s removed to yield a pale yellow solid (4.58 g, 67%). 1 H NMR (DMSO d 6 7.76 (br.s, 1H NC H N of C 3 H 3 N 2 ), 7.30 (dt, 2H, C 3 H 3 N 2 C 6 H 4 OCH 3 ), 7.19 (dt, 2H, C H =C H of C 3 H 3 N 2 ), 6.98 (dt, 2H, C 3 H 3 N 2 C 6 H 4 OCH 3 ), 3.84 (s, 6H, OC H 3 ) CH 2 [(C 3 H 3 N 2 C 6 H 4 OCH 3 ) 2 ] 140 The starting mater ial, 1 (4 methoxyphenyl) 1H imidazole (2.47 g, 16.4 mmol) was dissolved in THF then CH 2 Br 2 added and the solution heated in a pressure vessel at 150 C overnight. The mixture was then cooled to room temperature filtered and washed with THF to yield an off white solid (1.86 g, 63%) 1 H NMR (DMSO d 6 10.23 (t, 2H NC H N of C 3 H 3 N 2 ), 8.36 (d, 4H, C H =C H of C 3 H 3 N 2 ), 7.76, 7.24 (dt, 8H, CH 2 [(C 3 H 3 N 2 C 6 H 4 OCH 3 ) 2 ] ) 6.88 (br.s, 2H, C H 2 [(C 3 H 3 N 2 C 6 H 4 OCH 3 ) 2 ] ), 3.85 (s, 6H, OC H 3 ) CH 2 [(C 3 H 2 N 2 C 6 H 4 OCH 3 ) 2 ]Rh(COD)Br (52 ). The imidazolium salt 1,1' methylenebis(3 (4 methoxyphenyl) 1H imidazol 3 ium) bromide (0.204 g, 0.390 mmol) was suspended in THF and cooled to 78 C then K N(Si (CH 3 ) 3 ) 2 (0.156 g, 0.782 mmol) dissolved in THF was cannula transferred to the cooled suspension. The CO 2 /acetone bath was removed ; the mi xture warmed slowly to room temperature and allowed to react at room temperature for one hour. The deprotonated imidazolium was then recooled to 78 C and a cold ( 5 C) THF solution of [(COD)RhCl 2 ] (0.098 g, 0.198 mmol) transferred dropwise. The reacti on mixture was allowed to stir in the CO 2 /acetone bath overnight then filtered into hexane to precipitate the crude product. The yellow brown crude product was filtered and washed with diethyl ether then eluted from a silica column using CH 2 Cl 2 :MeOH (40/1 v/v) with increasing polarity to elute the

PAGE 133

133 compound as a bright yellow orange solution (0.112 g, 47%). 1 H NMR ( DMSO d 6 8.34 (d, 2H CH 2 [(C 3 H 2 N 2 C 6 H 4 OCH 3 ) 2 ] ), 7.89 (d, 1H C H 2 [(C 3 H 2 N 2 C 6 H 4 OCH 3 ) 2 ] ), 7.54 (dt, 4H C H 2 [(C 3 H 2 N 2 C 6 H 4 OCH 3 ) 2 ] ), 7.06 (dt, 4H C H 2 [(C 3 H 2 N 2 C 6 H 4 OCH 3 ) 2 ] ), 7. 00 (d, 1H C H 2 [(C 3 H 2 N 2 C 6 H 4 OCH 3 ) 2 ] ), 6.31 (d, 1H C H 2 [(C 3 H 2 N 2 C 6 H 4 OCH 3 ) 2 ] ) 4.51 (b r s, 2H, COD ), 3.91 (s, 6 H, COD ), C H 2 [(C 3 H 2 N 2 C 6 H 4 OC H 3 ) 2 ] 3.50 (b r s, 2H, COD ), 2.07 (m, 2H, COD ), 1.89, 1.85 (m, 6 H, COD ). LR MS (DIP CI) found for C 29 H 32 N 4 O 2 BrRh 651 [M] + 501

PAGE 134

134 APPENDIX CYCLIC VOLTAMMOGRAMS OF PREVIOUSLY PREPAR ED HETEROBIMETALLIC COMPLEXES WITH ET HANOL Cyclic vol tammograms of previously synthesized Ru/Pt, Ru/Pd, Ru/Au, Fe/Pt, Fe/Pd, Fe/Au and Ru/Sn complexes 1 3 5 1 8 and 20 28 in 0.1 M TBAT/DCE or TBAH/DCE at a 50 mVs 1 scan rate with a glassy carbon working electrode, Ag/Ag + reference electrode a nd Pt flag counter electrode.

PAGE 135

135

PAGE 136

136

PAGE 137

137

PAGE 138

138

PAGE 139

139 LIST OF REFERENCES (1) Leger, J. M.; Lamy, C. Ber. Bunsen Ges Phys Chem Chem Phys 1990 94 1021. (2) Vigier, F., Rousseau, S., Co utanceau, C., Leger, J. M., Lamy, C. Top ics in Catal. 2006 40 111. (3) Ketelar, J. A. A. Fuel Cell Systems ; Plenum Press: New York, 1993. (4) Song, S.; Maragou, V.; Tsiakaras, P. J Fuel Cell Sci Tech 2007 4 203. (5) Rousseau, S.; Coutanceau, C.; Lamy, C.; Leger, J. M. J Power Sources 2006 158 18. (6) Leger, J. M.; Rousseau, S.; Coutanceau, C.; Hahn, F.; Lamy, C. Electrochim. Acta 2005 50 5118. (7) Chetty, R.; Scott, K. Electrochim. Acta 2007 52 4073. (8) Arico, A. S.; Srinicasan, S.; Ant onucci, V. Fuel Cells 2001 1 133. (9) Zhou, W. J.; Zhou, B.; Li, W. Z.; Zhou, Z. H.; Song, S. Q.; Sun, G. Q.; Xin, Q.; Douvartzides, S.; Goula, M.; Tsiakaras, P. J. Power Sources 2004 126 16. (10) Lamy, C.; Belgsir, E. M.; Leger, J. M. J. Appl. Elect rochem. 2001 31 799. (11) Song, S.; Maragou, V.; Tsaikaras, P. J. Fuel Cell Sci. Technol. 2007 4 203. (12) Elschenbroich, C. Organometallics ; 3rd ed.; Wiley VCH: Weinheim, 2006. (13) Ravikumar, M. K.; Shukla, A. K. J. Electrochem. Soc. 1996 143 26 01. (14) Ren, X.; Wilson, M. S.; Gottesfeld, S. J. Electrochem. Soc. 1996 143 L12. (15) Shukla, A. K.; Jackson, C. L.; Scott, K.; Raman, R. K. Electrochim. Acta 2002 47 3401. (16) Leger, J. M. J Appl. Electrochem 2001 31 767. (17) Beden, B.; La my, C.; Bewick, A.; Kunimatsu, K. J Electroanal Chem 1981 121 343. (18) Alcohol. Encyclopaedia Brittanica, 1911 http://www.1911encyclopedia.org/Alcohol [Online Early Access]. Published Online: 2 006. http://www.1911encyclopedia.org/Alcohol (accessed August 2008).

PAGE 140

140 (19) Roach, J. 9,000 Year Old Beer Re Created From Chinese Recipe. National Geographic News http://news.nationalgeographic.com/news/index.html [Online Early Access]. Published Online: 2005. http://news.nationalgeographic.com/ne ws/2005/07/0718_050718_ancientbeer.html (accessed August 2008). (20) Kovarik, W. Henry Ford, Charles F. Kettering and the Fuel of the Future. Automotive History Review, http://www.radford.e du/~wkovarik/papers/fuel.html [Online Early Access]. Published Online: 1998. Reproduced on the Web at http://www.radford.edu/~wkovarik/papers/fuel.html (accessed August 2008). (21) DiPardo J. Outlook for Biomass Ethanol Production and Demand; Energy Information Administration; http://tonto.eia.doe.gov/FTPROOT/features/biomass.pdf [Online Early Access]. Published Online: 2008. (accessed August 2008). (22) Shen, S. Y.; Zhao, T. S.; Xu, J. B. Int J Hydrogen Energy 2010 35 12911. (23) Abd El Latif, A. A.; Mostafa, E.; Huxter, S.; Attard, G.; Baltruschat, H. Electrochim Acta 2010 55 7951. (24) Khanderi, J.; Contiu, C.; Engstler, J.; Hoffmann, R. C.; Schneider, J. J.; Drochner, A.; Vogel, H. Nanoscale 2011 3 1102. (25) Zhao, Y.; Yang, X.; Zhan, L.; Ou, S.; Tian, J. J. Mater. Chem. 2011 21 4257. (26) Konopka, D. A.; Li, M.; Artyushkova, K.; Marinkovic, N.; Sasaki K.; Adzic, R.; Ward, T. L.; Atanassov, P. J. Phys. Chem. C 2011 115 3043. (27) Liu, C. W.; Chang, Y. W.; Wei, Y. C.; Wang, K. W. Electrochim Acta 2011 56 2574. (28) Souza Garcia, J.; Herrero, E.; Feliu, J. M. ChemPhysChem 2010 11 1391. (29) Koi do, K.; Ishida, Y.; Kumabe, K.; Matsumoto, K.; Hasegawa, T. J. Supercrit. Fluids 2010 55 246. (30) Higuchi, E.; Miyata, K.; Takase, T.; Inoue, H. J. Power Sources 2011 196 1730. (31) Thotiyl, M. M. O.; Kumar, T. R.; Sampath, S. J. Phys. Chem. C 2010 114 17934. (32) Najjar, H.; Batis, H. Appl. Catal., A 2010 383 192. (33) Bagotzky, V. S.; Vassilyev, Y. B. Electrochim. Acta 1967 12 1323.

PAGE 141

141 (34) Antolini, E. J. Power Sources 2007 170 1. (35) Bagotzky, V. S.; Osetrova, N. V.; Skundinz, A. M. Rus sian J. Electrochem. 2003 39 919. (36) Cao, L.; Sun, G.; Li, H.; Xin, Q. Electrochem Commun 2007 9 2541. (37) Bagchi, J.; Bhattacharya, K. J. Power Sources 2007 163 661. (38) Watanabe, M.; Motoo, S. J. Electroanal. Chem. 1975 60 267. (39) Wat anabe, M.; Motoo, S. J. Electroanal. Chem 1975 60 276. (40) Iwasita, T. Electrochim. Acta 2002 47 3663. (41) Colmati, F.; Antolini, E.; Gonzalez, E. R. J. Power Sources 2006 157 98. (42) J Solid State El ectrochem 2005 9 1. (43) J Solid State Electrochem 2007 11 339. (44) Wheatley, N.; Kalck, P. Chem Rev 1999 99 3379. (45) Spinac, E. V.; Neto, A. O.; Linardi, M. J. Power Sources 2004 129 121. (46) Lee, K. S .; Park, I. S.; Park, H. Y.; Jeon, T. Y.; Cho, Y. H.; Sung, Y. E. J. Electrochem. Soc. 2009 156 B1150. (47) Wu, B.; Hu, D.; Kuang, Y.; Liu, B.; Zhang, X.; Chen, J. Angew. Chem., Intl. Ed. 2009 48 4751. (48) Wang, R.; Li, H.; Feng, H.; Wang, H.; Lei, Z. J. Power Sources 2010 195 1099. (49) Asthagiri, A.; Sholl, D. S. Surf. Sci. 2005 581 66. (50) Liang, Y. Q.; Cui, Z. D.; Zhu, S. L.; Liu, Y.; Yang, X. J. J. Catal. 2011 278 276. (51) Lazaro, M. J.; Celorrio, V.; Calvillo, L.; Pastor, E.; Moliner R. J. Power Sources 2011 196 4236. (52) Wen, Z.; Yang, S.; Liang, Y.; He, W.; Tong, H.; Hao, L.; Zhang, X.; Song, Q. Electrochim Acta 2010 56 139. (53) Boennemann, H.; Khelashvili, G. Appl. Organomet. Chem. 2010 24 257.

PAGE 142

142 (54) Moghieb, A.; Correi a, M. C.; McElwee White, L. Inorg. Chim. Acta 2011 In Press, Accepted Manuscript (55) Gozin, M.; Weisman, A.; Ben David, Y.; Milstein, D. Nature 1993 364 699. (56) Gokagac, G.; Kennedy, B. J.; Cashion, J. D.; Brown, L. J. J. Chem. Soc., Faraday Trans 1993 89 151. (57) Correia, M. C.; Moghieb, A.; Goforth, S. K.; McElwee White, L. ECS Trans. 2009 19 13. (58) Brown, N. F.; Barteau, M. A. Activation of C H, C C, and C O Bonds of Oxygenates on Rh(111) ; American Chemical Society: New York, 1993; Vol 517. (59) Houtman, C. J.; Barteau, M. A. J. Catal. 1991 130 528. (60) Dickson, R. S. Homogeneous Catalysis with Compounds of Rhodium and Iridium ; D. Reidel Publishing Company: Dordrecht, 1985. (61) Kawaguchi, T.; Rachi, Y.; Sugimoto, W.; Murakami, Y .; Takasu, Y. J. Appl. Electrochem. 2006 36 1117. (62) Lima, F. H. B.; Gonzalez, E. R. Electrochim. Acta 2008 53 2963. (63) Gupta, S. S.; Datta, J. J. Electroanal. Chem. 2006 594 65. (64) Bergamaski, K.; Gonzalez, E. R.; Nart, F. C. Electrochim. A cta 2008 53 4396. (65) Liou, S. Y.; Gozin, M.; Milstein, D. J. Am. Chem. Soc. 1995 117 9774. (66) Rybtchinski, B.; Vigalok, A.; Ben David, Y.; Milstein, D. J. Am. Chem. Soc. 1996 118 12406. (67) Osborn, J. A.; Shrock, R. R. J. Am. Chem. Soc. 1971 93 3089. (68) Shrock, R. R.; Osborn, J. A. J. Am. Chem. Soc. 1970 93 2397. (69) Hiyam, S.; Ben David, Y.; Shimon, L. J. W.; Milstein, D. Organometallics 2006 25 2292. (70) Nolan, S. P.; Hoff, C. D.; Stoutland, P. O.; Newman, L. J.; Buchanan, J. M. ; Bergman, R. G.; Yang, G. K.; Peters, K. S. J. Am. Chem. Soc. 1987 109 3143. (71) Stoutland, P. O.; Bergman, R. G.; Nolan, S. P.; Hoff, C. D. Polyhedron 1988 7 1429.

PAGE 143

143 (72) Bryndza, H. E.; Fong, L. K.; Paciello, R. A.; Tam, W.; Bercaw, J. E. J. Am. Ch em. Soc. 1987 109 1444. (73) Lide, D. R. CRC Handbook of Chemistry and Physics ; 88th ed., 2007 2008. (74) de Souza, J. P. I.; Iwasita, T.; Nart, F. C.; Vielstich, W. J. Appl. Electrochem. 1999 30 43. (75) Wasmus, S.; Vielstich, W. J. Appl. Electroch em. 1993 23 120. (76) Hamnett, A.; Weeks, S. A.; Kennedy, B. J.; Troughton, G.; Christensen, P. A. Ber. Bunsen Ges. 1990 94 1014. (77) Matare, G.; Tess, M. E.; Abboud, K. A.; Yang, Y.; McElwee White, L. Organometallics 2002 21 711. (78) Orth, S. D .; Terry, M. R.; Abboud, K. A.; Dodson, B.; McElwee White, L. Inorg. Chem. 1996 35 916. (79) Serra, D.; McElwee White, L. Inorgan Chim Acta 2008 361 3237. (80) Serra, D.; Abboud, K. A.; Hilliard, C. R.; McElwee White, L. Organometallics 2007 26 3 085. (81) Tess, M. E.; Hill, P. L.; Torraca, K. E.; Kerr, M. E.; Abboud, K. A.; McElwee White, L. Inorg. Chem. 2000 39 3942. (82) Yang, Y.; Abboud, K. A.; McElwee White, L. Dalton Trans. 2003 4288. (83) Yang, Y.; McElwee White, L. Dalton Trans 2004 2352. (84) Anthony, C. R.; McElwee White, L. J Molec Catal A: Chem 2005 227 113. (85) Nie, M.; Tang, H.; Wei, Z.; Jiang, S. P.; Shen, P. K. Electrochem. Commun. 2007 9 2375. (86) Song, S.; Tsiakaras, P. Appl. Catal. B Environ. 2006 63 187. ( 87) Kadirgan, F.; Beyhan, S.; Atilan, T. Int. J. Hydrogen Energy 2009 34 4312. (88) Serra, D.; Correia, M. C.; McElwee White, L. Submitted for publication 2010 (89) Chen, Z.; Concepcion, J. J.; Meyer, T. J. Dalton Trans 2011 40 3789.

PAGE 144

144 (90) Chen, Z. ; Concepcion, J. J.; Jurss, J. W.; Meyer, T. J. J. Am. Chem. Soc. 2009 131 15580. (91) Concepcion, J. J.; Jurss, J. W.; Templeton, J. L.; Meyer, T. J. J Am Chem Soc 2008 130 16462. (92) Concepcion, J. J.; Jurss, J. W.; Brennaman, M. K.; Hoertz, P G.; Patrocinio, A. O. T.; Murakami, I. N. Y.; Templeton, J. L.; Meyer, T. J. Acc. Chem. Res. 2009 42 1954. (93) Xu, Y.; Fischer, A.; Duan, L.; Tong, L.; Gabrielsson, E.; Aakermark, B.; Sun, L. Angew. Chem., Int. Ed. 2010 49 8934. (94) Tanaka, S.; M asaoka, S.; Yamauchi, K.; Annaka, M.; Sakai, K. Dalton Trans. 2010 39 11218. (95) Binstead, R. A.; Moyer, B. A.; Samuels, G. J.; Meyer, T. J. J. Am. Chem. Soc. 1981 103 2897. (96) Moyer, B. A.; Sipe, B. K.; Meyer, T. J. Inorg. Chem. 1981 20 1475. (97) Sullivan, B. P.; Salmon, D. J.; Meyer, T. J. Inorg. Chem. 1978 17 3334. (98) Barnett, N. W.; Bos, R.; Brand, H.; Jones, P.; Lim, K. F.; Purcell, S. D.; Russell, R. A. Analyst 2002 127 455. (99) Hesek, D.; Inoue, Y.; Everitt, S. R. L.; Ishida, H. ; Kunieda, M.; Drew, M. G. B. Dalton Trans 1999 3701. (100) Chen, J.; Zhang, W.; Dong, Z.; Swiegers, G. F.; Wallace, G. G. Organometallics 2007 (101) Rostovtsev, V. V.; Labinger, J. A.; Bercaw, J. E.; Lasseter, T. L.; Goldberg, K. I. Organometallics 1 998 17 4530. (102) Monaghan, P. K.; Puddephatt, R. J. Inorg. Chim. Acta 1982 65 L52. (103) Rotondo, E.; Tresoldi, G.; Faraone, F.; Pietropaolo, R. Inorg. Chem. 1976 15 2102. (104) Kang, W. D.; Wei, Y. C.; Liu, C. W.; Wang, K. W. Electrochem Commu n 2011 13 162. (105) Wang, Y.; Wang, X.; Li, C. M. Appl Catal B: Environ 2010 99 229. (106) Alcala, R.; Mavrikakis, M.; Dumesic, J. A. J. Catal. 2003 218 178. (107) Rybtchinski, B.; Milstein, D. Angew. Chem., Int. Ed. Engl. 1999 38 870.

PAGE 145

145 (10 8) Jutzi, P.; Siemeling, U. J. Organomet. Chem. 1995 500 175. (109) Philippopoulos, A. I.; Bau, R.; Poilblanc, R.; Hadjiliadis, N. Inorg. Chem. 1998 37 4822. (110) Schut, D. M.; Weakley, T. J. R.; Tyler, D. R. New J. Chem. 1996 20 113. (111) Panda T. K.; Gamer, M. T.; Roesky, P. W. Organometallics 2003 22 877. (112) Causey, P. W.; Baird, M. C.; Cole, S. P. C. Organometallics 2004 23 4486. (113) Stephenson, T. A.; Wilkinson, G. J Inorg Nucl Chem 1966 28 945. (114) Nieckarz, G. F.; Litt y, J. J.; Tyler, D. R. J Organomet Chem 1998 554 19. (115) Coolbaugh, T. S.; Coots, R. J.; Santarsiero, B. D.; Grubbs, R. H. Inorg Chim Acta 1985 98 99. (116) Rubner, G.; Bensdorf, K.; Wellner, A.; Bergemann, S.; Ott, I.; Gust, R. European J Me d Chem 2010 45 5157. (117) Partridge, J. J.; Chadha, N. K.; Uskokovic, M. R. Org. Synth. 1985 63 44. (118) Brown, H. C.; Kim, S. C.; Krishnamurthy, S. J. Org. Chem. 1980 45 1. (119) Rezai, N.; Meybodi, F. A.; Salehi, P. Synth. Commun. 2000 30 1799. (120) Miyazawa, A.; Kase, T.; Soga, K. Macromolecules 2000 33 2796. (121) Colton, R.; Farthing, R. H. Aus J Chem 1967 20 1283. (122) Eichenseher, S.; Delacroix, O.; Kromm, K.; Hampel, F.; Gladysz, J. A. Organometallics 2005 24 245. (123) Skell, P. S.; Sandler, S. R. J Am Chem Soc 1958 80 2024. (124) Fischer, E. O.; Maasbol, A. Angew Chem 1964 76 645. (125) Fischer, E. O. Pure Appl Chem 1972 30 353. (126) Chem Rev 2009 109 3408. (127) Herrmann, W. A.; Kcher, C. Angew Chem Int Ed 1997 36 2162.

PAGE 146

146 (128) Arduengo III, A. J.; Harlow, R. L.; Kline, M. J Am Chem Soc 1991 113 361. (129) Jacobsen, H.; Correa, A.; Poater, A.; Costabile, C.; Cavallo, L. Coor Chem Rev 2009 253 687. (130) O., N.; Viciu, M. S. Ann Rep Prog Chem Section B: Org Chem 2010 106 243. (131) Crabtree, R. H. J Organomet Chem 2005 690 5451. (132) Liu, J.; Chen, J.; Zhao, J.; Zhao, Y.; Li, L.; Zhang, H. Synthesis 2003 17 26 61. (133) Bellina, F.; Cauteruccio, S.; Mannina, L.; Rossi, R.; Viel, S. J Org Chem 2005 70 3997. (134) Rivera, G.; Crabtree, R. H. J Molec Catal A: Chem 2004 222 59. (135) Chem Rev 2009 109 3612. (136) Hahn, F. E.; Jahnke, M. C. Angew Chem Int Ed 2008 47 3122. (137) Scholl, M.; Trnka, T. M.; Morgan, J. P.; Grubbs, R. H. Tetrahedron Lett 1999 40 2247. (138) Samec, J. S. M.; Keitz, B. K.; Grubbs, R. H. J Organomet Chem 2010 695 1831. (139) Herrmann, W. A. Angew Chem Int Ed 2002 41 1290. (140) Herrmann, W. A.; Schwarz, J.; Gardiner, M. G. Organometallics 1999 18 4082. (141) Jokic, N. B.; Straubinger, C. S.; Li Min Goh, S.; Herdtweck, E.; Herrmann, W. A.; Khn, F. E. In org Chim Acta 2010 363 4181. (142) Lee, H. M.; Jiang, T.; Stevens, E. D.; Nolan, S. P. Organometallics 2001 20 1255. (143) Hillier, A. C.; Lee, H. M.; Stevens, E. D.; Nolan, S. P. Organometallics 2001 20 4246. (144) Lin, I. J. B.; Vasam, C. S. C oor Chem Rev 2007 251 642. (145) Allen, D. P.; Crudden, C. M.; Calhoun, L. A.; Wang, R.; Decken, A. J Organomet Chem 2005 690 5736. (146) Praetorius, J. M.; Crudden, C. M. Dalton Trans 2008 4079.

PAGE 147

147 (147) Corbern, R.; Peris, E. Organometallics 2008 27 1954. (148) Pontes da Costa, A.; Viciano, M.; Sana, M.; Merino, S.; Tejeda, J.; Peris, E.; Royo, B. Organometallics 2008 27 1305. (149) Schneider, S. K.; Herrmann, W. A.; Herdtweck, E. Zeit anorg allgemeine Chem 2003 629 2363. (150) K inder, R. E.; Zhang, Z.; Widenhoefer, R. A. Org Lett 2008 10 3157. (151) Jensen, D. R.; Schultz, M. J.; Mueller, J. A.; Sigman, M. S. Angew Chem Int Ed 2003 42 3810. (152) Konnick, M. M.; Gandhi, B. A.; Guzei, I. A.; Stahl, S. S. Angew Chem I nt Ed 2006 45 2904. (153) May, T. L.; Brown, M. K.; Hoveyda, A. H. Angew Chem Int Ed 2008 47 7358. (154) Dez Gonzlez, S.; Nolan, S. P. Angew Chem Int Ed 2008 47 8881. (155) Saino, N.; Kogure, D.; Okamoto, S. Org Lett 2005 7 3065. (156) Geny, A.; Gaudrel, S.; Slowinski, F.; Amatore, M.; Chouraqui, G.; Malacria, M.; Aubert, C.; Gandon, V. Adv Synth Catal 2009 351 271. (157) Garrison, J. C.; Simons, R. S.; Talley, J. M.; Wesdemiotis, C.; Tessier, C. A.; Youngs, W. J. Organometa llics 2001 20 1276. (158) Tulloch, A. A. D.; Danopoulos, A. A.; Kleinhenz, S.; Light, M. E.; Hursthouse, M. B.; Eastham, G. Organometallics 2001 20 2027. (159) Weskamp, T.; Kohl, F. J.; Herrmann, W. A. J Organomet Chem 1999 582 362. (160) Sauva ge, X.; Demonceau, A.; Delaude, L. In Metathesis Chemistry ; Imamoglu, Y., Dragutan, V., Karabulut, S., Eds.; Springer Netherlands: 2007; Vol. 243, p 91. (161) Sauvage, X.; Borguet, Y.; Noels, A. F.; Delaude, L.; Demonceau, A. Adv Synth Catal 2007 349 255. (162) Borguet, Y.; Sauvage, X.; Demonceau, A.; Delaude, L. In Green Metathesis Chemistry ; Dragutan, V., Demonceau, A., Dragutan, I., Finkelshtein, E. S., Eds.; Springer Netherlands: 2010, p 89. (163) Sauvage, X.; Borguet, Y.; Zaragoza, G.; Demoncea u, A.; Delaude, L. Adv Synth Catal 2009 351 441.

PAGE 148

148 (164) Ogata, K.; Nagaya, T.; Fukuzawa, S. i. J Organomet Chem 2010 695 1675. (165) Leung, C. H.; Incarvito, C. D.; Crabtree, R. H. Organometallics 2006 25 6099. (166) Herrmann, W. A.; Roesky, P. W.; Elison, M.; Artus, G.; Oefele, K. Organometallics 1995 14 1085. (167) Raynal, M.; Cazin, C. S. J.; Vallee, C.; Olivier Bourbigou, H.; Braunstein, P. Dalton Trans 2009 3824. (168) Zamora, M. T.; Ferguson, M. J.; McDonald, R.; Cowie, M. Dalton T rans 2009 7269. (169) Mendoza Espinosa, D.; Donnadieu, B.; Bertrand, G. Chem Asian J 2011 6 1099. (170) Yamazaki, S. i.; Yamada, Y.; Yasuda, K. Inorg Chem 2004 43 7263. (171) Ding, K.; Miller, D. L.; Young, V. G.; Lu, C. C. Inorg Chem 2011 50 2545. (172) Tortorelli, L. J.; Woods, C.; McPhail, A. T. Inorg. Chem. 1990 29 2726. (173) Connelly, N. G.; Loyns, A. C.; Fernandez, M. J.; Modrego, J.; Oro, L. A. J. Chem. Soc., Dalton Trans. 1989 683. (174) Gardiner, M. G.; Herrmann, W. A.; Re isinger, C. P.; Schwarz, J.; Spiegler, M. J Organomet Chem 1999 572 239. (175) Scherg, T.; Schneider, S. K.; Frey, G. D.; Schwarz, J.; Herdtweck, E.; Herrmann, W. A. Synlett 2006 2006 2894. (176) Bhm, V. P. W.; Herrmann, W. A. Angew Chem Int E d 2000 39 4036. (177) Hahn, F. E.; Foth, M. J Organomet Chem 1999 585 241. (178) Heckenroth, M.; Neels, A.; Stoeckli Evans, H.; Albrecht, M. Inorg Chim Acta 2006 359 1929. (179) Tubaro, C.; Biffis, A.; Gonzato, C.; Zecca, M.; Basato, M. J M olec Catal A: Chem 2006 248 93. (180) A. Herrmann, W.; Reisinger, C. P.; Spiegler, M. J Organomet Chem 1998 557 93. (181) Cheng, Y.; Lu, X. Y.; Xu, H. J.; Li, Y. Z.; Chen, X. T.; Xue, Z. L. Inorg Chim Acta 2010 363 430.

PAGE 149

149 (182) Stephenson, T A.; Wilkinson, G. J Inorg Nucl Chem 1966 28 2285. (183) Jeletic, M. S.; Jan, M. T.; Ghiviriga, I.; Abboud, K. A.; Veige, A. S. Dalton Trans 2009 2764. (184) Salzer, A.; Bauer, A.; Geyser, S.; Podewils, F.; Turpin, G. C.; Ernst, R. D. Inorg Syn th 2004 34 59. (185) Wells, K. D.; Ferguson, M. J.; McDonald, R.; Cowie, M. Organometallics 2008 27 691. (186) Wick, D. D.; Goldberg, K. I. J. Am. Chem. Soc. 1997 119 10235. (187) Bartlett, K. L.; Goldberg, K. I.; Borden, W. T. Organometallics 20 01 20 2669. (188) Jensen, M. P.; Wick, D. D.; Reinartz, S.; White, P. S.; Templeton, J. L.; Goldberg, K. I. J. Am. Chem. Soc. 2003 125 8614. (189) Smythe, N. A.; Grice, K. A.; Williams, B. S.; Goldberg, K. I. Organometallics 2009 28 277. (190) Cru mpton Bregel, D. M.; Goldberg, K. I. J. Am. Chem. Soc. 2003 125 9442. (191) Chen, Z.; Concepcion, J. J.; Hu, X.; Yang, W.; Hoertz, P. G.; Meyer, T. J. Proc. Natl. Acad. Sci. U. S. A. 2010 107 7225. (192) Concepcion, J. J.; Jurss, J. W.; Norris, M. R. ; Chen, Z.; Templeton, J. L.; Meyer, T. J. Inorg. Chem. 2010 49 1277. (193) Concepcion, J. J.; Tsai, M. K.; Muckerman, J. T.; Meyer, T. J. J. Am. Chem. Soc. 2010 132 1545. (194) Jurss, J. W.; Concepcion, J. C.; Norris, M. R.; Templeton, J. L.; Meyer, T. J. Inorg. Chem. 2010 49 3980. (195) Chen, Z.; Concepcion, J. J.; Luo, H.; Hull, J. F.; Paul, A.; Meyer, T. J. J. Am. Chem. Soc. 2010 132 17670. (196) Concepcion, J. J.; Jurss, J. W.; Hoertz, P. G.; Meyer, T. J. Angew. Chem., Int. Ed. 2009 48 94 73. (197) Paul, A.; Hull, J. F.; Norris, M. R.; Chen, Z.; Ess, D. H.; Concepcion, J. J.; Meyer, T. J. Inorg. Chem. 2011 50 1167. (198) Procelewska, J.; Zahl, A.; Liehr, G.; Van, E. R.; Smythe, N. A.; Williams, B. S.; Goldberg, K. I. Inorg. Chem. 2005 44 7732.

PAGE 150

150 (199) Blum, O.; Milstein, D. J. Am. Chem. Soc. 1995 117 4582. (200) Blum, O.; Milstein, D. J. Organomet. Chem. 2000 593 594 479. (201) Fafard, C. M.; Ozerov, O. V. Inorg. Chim. Acta 2007 360 286. (202) Dixon, K. R. Phosphorus to Bismuth ; Plenum Press: New York, 1987. (203) Mann, B. E. The Common Nuclei ; Academic Press: London, 1978. (204) Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay, P. A.; Masters, A. F.; Phillips, L. J. Phys. Chem. B 1999 103 6713.

PAGE 151

151 BIOGRAPHIC AL SKETCH Marie Christina Correia was born in Georgetown, Guyana; the first of five children to parents Desmond and Indranie Correia. She attended St. Agne s Primary School followed by w successfully completing her fifth form C aribbean Examination C ouncil Exams, she went on to study for her General Certificate of Education A dvanced Level Exams at Queen s Coll ege. After high school, Marie worked for a year at The Bank of Nova Scotia in Georgetown before leaving Guyana in January 2001 to attend the College of Charleston in Charleston, South Carolina During her time at CofC, she worked in the lab of Prof. Fred erick J. Heldrich performing multistep organic synthesis. In December 2004, she graduated with Bachelors of Science in Biochemis try and Chemistry and spent the next six months working for Dr. G. Frederick Hutter in the Specialty Chemicals Division at Mead Westvaco Corporation (now MWV Specialty Chemicals) where she learnt to synthesize polymers. In the fall of 2005 Marie began her graduate career at the University of Florida and in the summer of 2006, she started her PhD. research in organometallic chemis try in the g roup of Prof. Lisa McElwee White She graduated in the spring of 2011 with a Doctor of Philosophy in Chemistry and subsequently wen t on to postdoctoral studies with Prof. Marcel Schlaf at the University of Guelph in Ontario Canada.