HOMOGENEOUS CATALYTIC OXIDATION OF AMMONIA UNDER MILD CONDITIONS By Faezeh Habib Zadeh A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Chemistry - Doctor of Philosophy 2019 ii ABSTRACT HOMOGENEOUS CATALYTIC OXIDATION OF AMMONIA UNDER MILD CONDITIONS By Faezeh Habib Zadeh Electrochemical ammonia splitting as a route for extract ing hydrogen from ammonia, is plagued by relatively high oxidation overpotentials . Such requirements for an extra energy input are often caused by the sluggish kinetics associated to the multi - electron, multi - proton nature of the oxidation of ammonia to dinit rogen (N 2 ) . A molecular catalyst added homogeneously to the solutions of ammoni a , is a novel approach to assist the kinetics of the oxidation process and consequently lower the required overpotentials. In this dissertation, ruthenium (II) polypyridyl ammin e complexes are employed as homogeneous catalysts for electro - oxidation of ammonia in tetrahydrofuran (THF) at room temperature and ambient pressure. Oxidation of ammonia was conducted in the presence of the catalyst at a constant applied anodic potential and evolution of N 2 and H 2 gases as the products of the electrolysis was confirmed by gas chromatography. The possibility of a heterogenous mechanism for the catalysis was dismissed after control rinse tests and microscopic examination of the working elect rode showed that no active ruthenium depositions were formed on the electrode surface. Authentic samples of ruthenium (III) polypyridyl ammine and ruthenium (II) polypyridyl hydrazine complexes were synthesized and characterized, and their role in the ca talytic cycle were investigated. Isotopic N - labeling experiments, revealed that both the coordinated ammine and the free ammonia in the solution participate in the N 2 formation , enlighten ed the path for generation of a hydrazine intermediate. The hydrazine complex was observed in low temperature 1 H NMR experiments, in which the Ru(III) am m ine complex was treated with NH 3 . Interestingly, reactions iii between the Ru(III) ammine complex and ammonia resulted in regeneration of the Ru(II) ammine complex . Same observations was made when a non - coordinating base was used instead of ammonia, suggesting that in the presence of a proton acceptor a redox disproportionation would afford the Ru(II) ammine from Ru(III) ammine. Our further electrochemical, spectro photochemical and mass spectroscopy results that are discussed in this dissertation, enabled us to construct a proposed catalytic cycle for oxidation of ammonia in THF using the ruthenium (II) ammine catalyst . In this dissertation the first example of a mo lecular transition metal is introduced that can catalyze ammonia electro - oxidation, opening up the gate for further developments and progresses in the field of electrochemical ammonia oxidation. iv ACKNOWLEDGEMENTS Firstly, I would like to express my sincere gratitude to my advisor Prof. Thomas Hamann for the continuous support of my Ph.D. study and related research, for his patience, motivation, and immense knowledge. His guidance helped me in all the time of research and writing of this thesis. Besides my advisor, I would like to thank the rest of my thesis committee: Prof. Arron Odom, Prof. Mitch Smith, and Prof. Dana Spence, for their insightful comments and encouragement . My special grati tude goes out to Prof. Smith and Dr. S. Miller for their invaluable contribution in my project, and also for our very helpful discussion s which incented me to widen my research from various perspectives. I would like to special ly mention my friends Mrs. Pa risa Shadabipour, Miss. Yujue Wang, Mrs. Mersedeh Sanipey and Miss. Nona Ehya ei and my former and present colleagues in Hamann lab. It was fantastic to have the opportunity to work with you and sharing the laboratory with you during last five years. You ma de graduate school fun! I am also thankful to Michigan State University and the U.S. Department of Energy for helping me and providing the funding for the work , and Chemistry faculty and staff for having their doo rs open and being supportive and helpful. v TABLE OF CONTENTS .. . viii .. ix .. xv KEY TO ... xvi INTRODUCTION ................................ ................................ ................................ ... 1 1.1. The Essence of Carbon Neutrality ................................ ................................ ....................... 2 1.2. Carbon - Neutral Fuels ................................ ................................ ................................ ........... 4 1.3. Energy - Related Applications of Ammonia ................................ ................................ .......... 6 1.3.1. Using NH 3 in Internal Combustion Engines ................................ ................................ .. 6 1.3.2. Ammonia - Fueled Solid Oxide Fuel Cells (SOFC) ................................ ........................ 6 1.3.3. Ammonia as a Hydrogen Carrier ................................ ................................ ................... 6 1.4. Ammonia Splitting ................................ ................................ ................................ ............... 9 1.4.1. Thermal Cracking of Gaseous NH 3 ................................ ................................ ............... 9 1.4.2. Ammonolysis of Alkali Metal Hydrides ................................ ................................ ..... 10 1.4.3. Electrolysis of Ammonia ................................ ................................ ............................. 10 1.5. Transition Metal Complexes as Catalysts for Ammonia Oxidation ................................ .. 13 1.6. This work ................................ ................................ ................................ ............................ 17 APPENDIX ................................ ................................ ................................ ................................ ... 19 REFERENCES ................................ ................................ ................................ ............................. 22 EXPERIMENTAL DETAIL S ................................ ................................ ............... 27 2.1. G eneral Materials and Methods ................................ ................................ ......................... 28 2.2. Preparation of Dry Liquid Ammonia ................................ ................................ ................. 29 2.3. Preparation of Saturated 15 NH 3 Solutions in Tetrahydrofuran (THF) ................................ 30 2.4. Electrochemistry ................................ ................................ ................................ ................. 31 2.4.1. Cyclic Voltammetry (CV) ................................ ................................ ........................... 31 2.4.2. D etermination of the Onset Potential ................................ ................................ .......... 33 2.4.3. Hydrodynamic Voltammetry ................................ ................................ ....................... 33 2.4.4. Controlled Potential Electrolysis (CPE) ................................ ................................ ...... 34 2.5. Gas Chromatography (GC) ................................ ................................ ................................ 36 2.5.1. GC Calibration ................................ ................................ ................................ ............. 37 2.6. GC - MS Experiment ................................ ................................ ................................ ............ 38 2.7. X - Ray Spectroscopy ................................ ................................ ................................ ........... 39 REFERENCES ................................ ................................ ................................ ............................. 40 CATALYTIC ACTIVITY OF [RU(TRPY)(DMABPY) NH 3 ] 2+ IN THF ............. 42 3.1. Synthesis of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1a ) ................................ ................................ ..... 43 3.1.1. Synthesis of (2,2´:6´,2´´ - terpyridyl) trichloro ruthenium (III), [Ru(trpy)Cl 3 ]. ............ 43 vi 3.1.2. (2,2´:6´,2´´ - terpyridyl) (2,2´ - bipyridyl) chloro ruthenium (II) chloride, [Ru(trpy)(bpy)Cl]Cl. ................................ ................................ ................................ ............. 43 3.1.3. (2,2´:6´,2´´ - Terpyridyl)(2,2´ - bipyridyl)ruthenium(II) ammine dihexafluoro phosphate, [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 , ( 1a ). ................................ ................................ ........................... 44 3.2. Catalytic Activity of 1a in THF ................................ ................................ ......................... 44 3.3. Synthesis of [Ru(trpy)(dmabpy)NH 3 ](PF 6 ) 2 ( 2a ) ................................ ............................... 49 3.3.1. (2,2´:6´,2´´ - Terpyridyl)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl)chloro ruthenium(II) chloride, [Ru(trpy)(bdmabpy)Cl]Cl. ................................ ............................... 49 3.3.2. (2,2´:6´,2´´ - Terpyridyl)( 4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl)ruthenium(II) ammine dihexafluoro phosphate, [Ru(trpy)(dmabpy)NH 3 ](PF 6 ) 2 , ( 2a ). ............................... 50 3.4. Catalytic Activity of 2a in THF ................................ ................................ ......................... 51 3.4.1. Assessment of the Catalytic Current by Cyclic Voltammetry ................................ ..... 51 3.4.2. Quantification of the Products of the Electrolysis Using Gas Chromatography ......... 53 3.5. Homogeneous vs. Heterogeneous Catalysis ................................ ................................ ....... 56 3.6. Isotopic Labeling Experiments ................................ ................................ ........................... 59 3.6.1. Synthesis of [Ru(trpy)(dmabpy) 15 NH 3 ](PF 6 ) 2 ( 15 N - 2a ) ................................ .............. 60 3.6.2. Electrolysis of 15 NH 3 using 2a and 15 N - 2a ................................ ................................ .. 60 3.7. Conclusions ................................ ................................ ................................ ........................ 61 APPENDIX ................................ ................................ ................................ ................................ ... 63 REFERENCE ................................ ................................ ................................ ................................ 83 THE RU(III) INTERME DIATE ................................ ................................ ........... 85 4. 1. Synthesis of [Ru(trpy)(dmabpy)NH 3 ](PF 6 ) 3 , ( 2b ) ................................ .............................. 86 4.1.1. Electrochemical Synthesis ................................ ................................ ........................... 86 4.1.2. Chemical Synthesis ................................ ................................ ................................ ..... 87 4.2. Characterization of 2b ................................ ................................ ................................ ........ 88 4.2.1. Proton NMR ................................ ................................ ................................ ................ 88 4.2.2. Elemental Analysis ................................ ................................ ................................ ...... 88 4.2.3. Magnetic Susceptibility ................................ ................................ ............................... 89 4.2.4. Electrochemistry ................................ ................................ ................................ .......... 91 4.3. Deprotonation of [Ru(trpy)(dmabpy)NH 3 ] +3 ................................ ................................ ..... 93 4.4. Regeneration of 2a from 2b ................................ ................................ ............................... 98 4.5. R eaction of [Ru(trpy)(dmabpy)NH 3 ] +3 With Bases ................................ ......................... 101 4.5.1. Electrochemistry ................................ ................................ ................................ ........ 101 4.5.2. 1 H NMR Studies ................................ ................................ ................................ ........ 104 4.5.3. UV - Vis Spectrophotometric Titrations ................................ ................................ ..... 106 4.6. Rotating Ring Disk Electrode (RRDE) Experiments ................................ ....................... 111 4.6.1. Titration with DBU: Scanning E D , Constant E R ................................ ........................ 113 4.6.2. Titration with DBU: Scanning E R , Constant E D ................................ ........................ 118 4.7. Conclusions ................................ ................................ ................................ ...................... 122 APPENDIX ................................ ................................ ................................ ................................ . 123 REFERENCES ................................ ................................ ................................ ........................... 142 THE [RU(TRPY)(D MABPY)N 2 H 4 ] 2+ INTERMEDIATE ................................ . 144 5.1. Synthesis of an Authentic Sample of [Ru(trpy)(bpy)N 2 H 4 ] 2+ Complex ( 2e ) ................... 146 5.2. Ligand Displacement in 2e / THF / NH 3 ................................ ................................ .......... 147 vii 5.3. Cyclic Voltammetry Studies ................................ ................................ ............................ 148 5.4. Variable Temperature (VT) 1 H NMR Experiments 6 ................................ ........................ 155 5.5. Spectrophotochemical Studies ................................ ................................ ......................... 159 5.6. Conclusions ................................ ................................ ................................ ...................... 162 APPENDIX ................................ ................................ ................................ ................................ . 163 REFERENCES ................................ ................................ ................................ ........................... 166 PRELIMINARY ELE CTROCHEMICAL STUDIES OF OTHER RUTHENIUM POLYPYRIDYL AMINE CA TALYSTS ................................ ................................ ................... 169 6.1. [Ru(Me 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 , ( 3a ). ................................ ................................ ......... 170 6.1.1. Synthesis ................................ ................................ ................................ .................... 170 6.1.1.1. (4',4,4'' - trimethyl - 2,2':6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) chloro ruthenium(II) chloride, [Ru(Me 3 trpy)(dmabpy)Cl]Cl. ............................ 170 6.1.1.2. (4',4,4'' - trimethyl - 2,2':6',2'' - terpyridine) (4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) ruthenium (II) ammine dihexafluorophosphate, [Ru(Me3trpy)(bdmabpy)NH 3 ](PF 6 ) 2 , ( 3a ). ................................ ................................ ......... 171 6.1.2. Ammonia Oxidation Using 3a as the Catalyst ................................ .......................... 171 6.2. [Ru( t Bu 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 , ( 4a ). ................................ ................................ ......... 174 6.2.1. Synthesis ................................ ................................ ................................ .................... 174 6.2.1.1. (4',4,4'' - tri - tert - butyl - 2,2':6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) chloro ruthenium (II) chloride, [Ru( t Bu 3 trpy)(dmabpy)Cl]Cl. ........................... 174 6.2.1.2. (4',4,4'' - tri - tert - butyl - 2,2':6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) ruthenium (II) ammine dihexafluorophosphate, ( 4a ). ................................ ......... 174 6.2.2. Ammonia Oxidation Using 4a as the Catalyst ................................ .......................... 175 6.3. [Ru(dmaptrpy)(dmabpy)NH 3 ](PF 6 ) 2 , ( 5a ). ................................ ................................ ...... 178 6.3.1. Synthesis ................................ ................................ ................................ .................... 178 6.3.1.1. (4 - N,N - d imethylaminophenyl) - 2,2',6',2'' - terpyridine, dmaptrpy. ....................... 178 6.3.1.2. ((4 - N,N - dimethylaminophenyl) - 2,2',6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) chloro ruthenium (II) chloride, [Ru(dmaptrpy)(dmabpy)Cl]Cl. ................................ ................................ ............................ 179 6.3.1.3. ((4 - N,N - dimethylaminophenyl) - 2,2',6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) ruthenium(II) ammine dihexafluorophosphate, ( 5a ). ....... 179 6.3.2. Ammonia Oxidation Using 5a as the Catalyst ................................ .......................... 180 6.4. Conclusions ................................ ................................ ................................ ...................... 182 APPENDIX ................................ ................................ ................................ ................................ . 184 REFERENCES ................................ ................................ ................................ ........................... 190 CHAPTER 7 CONCLUDING REMARKS A ND FUTURE DIRECTIONS ........................... 192 viii LIST OF TABLES Table 2.1 Comparison of energy densities between hydrogen and ammonia as fuels. Gasoline is also shown as the reference. ................................ ................................ ................................ ............ 5 Table 2.1 Experimenta l data used for GC signal calibration for moles of injected N 2 and H 2 . Red and blue colored numbers are used to construct the corresponding calibration lines in Fig. 2.5.2. ................................ ................................ ................................ ................................ ....................... 37 ix LIST OF FIGURES Figure 1.1.1 Relative amounts (in percent) of U.S. greenhouse gas emissions in 2016 , shows the large contribution of CO 2 . Figure reproduced from ref. [2]. ................................ .......................... 2 Figure 1.1.2 The U.S. energy consumption in 2017 from various sources shows the significance reliance on the fossil fuels. Figure reproduced from ref. [2]. ................................ ......................... 3 Figure 1.3.1 The ideal carbon - neutral energy cycle of ammonia as a hydron carrier. .................. 8 Figure 1.4.1 An ammonia cracker by Sam Gas Projects Ptv. Ltd. (India) which is maintained at 850 o C using an electric furnace. The purity of the produced hydrogen is 99.5% (0.5% nitrogen contamination). Visit: https://www.psa - nitrogen.com ................................ ................................ .. 10 Figure 1.6.1 Ruthenium catalysts 1 a - 5 a used in this study. ................................ ........................ 18 Figure 2.2.1 The apparatus used for the storage of dry liquid ammonia under ambient pressure. ................................ ................................ ................................ ................................ ....................... 30 Figure 2.4.1 Diagram of the three - electrode electrochemical cell used for CV experiments. The side arms are inert gas inlet and outlets. CE: Counter Electrode, W E: Working Electrode and RE: Reference Electrode. ................................ ................................ ................................ ..................... 31 Figure 2.4.2 CVs (three cycles) of ferrocene in THF and NM using the Ag/AgNO 3 ref erence electrode. Top: 3.0 10 - 3 M ferrocene in THF, and Bottom: 1.6 10 - 3 M ferrocene in NM. CVs were obtained at 0.1 V s - 1 scan rate. E p,a and E p,c stand for anodic and cathodic peak potentials, respectively. The E 1/2 is defined as the midpoint between E p,a and E p,c and is calculated as their arithmetic average. ................................ ................................ ................................ ........................ 32 Figure 2.4.3 Determination of the onset potential for oxidation of ammonia at the surface of the glassy carbon electrode in THF. The potential of the cross - point between the baseline and the oxidation current is conside red as the onset potential. ................................ ................................ . 33 Figure 2.4.4 Diagram of the three - electrode electrochemical cell used in RDE and RRDE experiments. The side arms are inert gas inlet and outlets. CE: Counter Electrode, WE: Working Electrode and RE: Reference Electrode. ................................ ................................ ....................... 34 Figure 2.4.5 The cell used in some controlled potential electrolysis experiments. An empty cell is presented here for better visibility. The counter electrode is Pt mesh, and work ing is glassy carbon plate. The inset shows the position of the sampling port. ................................ ............................. 35 Figure 2.5.1 Full TCD Chromatograms obtained for two headspace injections. Injection A (black): Headspace of a cell containing THF and the catalyst. Injection B (red): Headspace of the cell after 2 h of electrolysis in the presence of NH 3 and the catalyst. At point 1, a GC column valve switches to isolate N 2 , H 2 , and residual O 2 in the molsieve column, while heavier volatiles, THF, and NH 3 , elute through the PLOT/U column to the detector during time window 2. At time point 3, the column isolation valve resets and diato mic gases elute through the molsieve column to the thermal conductivity detector as seen in time window 4. ................................ ................................ .......... 36 x Figure 2.5.2 Gas Chromatography calibration lines obtained for N 2 (top, blue) and H 2 (bottom, red) based on data in Table 2.1. ................................ ................................ ................................ .... 38 Figure 3.2.1 Top: Scan rate dependence of the current in 2.13 × 10 - 3 M [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1a ) in THF shown for seven scan rates. Bottom: Plots of anodic and cathodic peak currents obtained from CVs on the top versus square r oot of the scan rate. From the slope of the anodic branch a diffusion coefficient of D ox = 4.06 10 - 6 cm 2 s - 1 is calculated for 1a . ............................ 46 Figure 3.2.2 Cyclic voltammograms for a solution containing ferrocene in THF with (red) and without (blue) NH 3 (0.34M). The onset of the direct ammonia oxidation was measured as +0.25 V vs. ferrocene. In solutions that do not contain ferrocene, the oxidation of ammonia appears at the same potential using an Ag/AgNO 3 reference electrode that is separately calibrated with Fc +/0 in THF. ................................ ................................ ................................ ................................ .............. 47 Figure 3.2.3 Top: Cyclic voltammograms for THF solutions of 2.13 × 10 - 3 M catalyst 1 a (blue), 0.34 M NH 3 added to the solution of 2.13 ×10 - 3 M catalyst 1a (red) and direct NH 3 oxidation in THF in the absence of 1a (dotted green). Scan rate 0.1 V s - 1 . Bottom: Normalized catalytic (i / ) currents when NH 3 (0.34 M) is added to a solution of 2.13×10 - 3 M 1a obt ained for mentioned scan rates. The magnitude and the onset of the normalized currents are being improved as the scan rate is decreasing. ................................ ................................ ................................ ................................ . 48 Figure 3.4.1 Cyclic voltammograms for THF solutions of 2.5 × 10 - 3 M 2 a (black), 2.5 × 10 - 3 M 2 a and 0.34 M NH 3 (red), uncatalyzed NH 3 oxidation in THF, 0.34 M (green) and the electrolyte background (gray). Scan rate 0.1 Vs - 1 . This figure shows that upon the presence of the catalyst 2a , a catalytic current is appeared at lower overpotentials relative to direct oxidation of ammonia in THF . ................................ ................................ ................................ ................................ .............. 52 Figure 3.4.2 Dependence of catalytic current on the concentration of ammonia when [ 2 a ] = 2.5 × 10 - 3 M, [NH 3 ] = (a) 0 M, (b) 0.008 M, (c) 0.04 M, (d) 0.07 M, (e) 0.17 M and (f) 0.34 M. Inset: Peak currents (at 0.90 V for catalytic currents) versus concentration of ammonia. The first data point (gray) is the anodic peak current for the catalyst in the absence of ammonia and is shown f or comparison. Scan rate 100 mV s - 1 . ................................ ................................ ............................... 53 Figure 3.4.3 Current passing during three steps of controlled potential electrolysis (each 3600 s) while the solution was being stirred under argon. [ 2a ] = 2.7 × 10 - 3 M, [NH 3 ] = 0.34 M, 0.1 M [NH 4 ](PF 6 ) in dry THF. Numbers show the amount of charge (q, in coulombs) pa ssed in each step, calculated based on the integration of the area under each curve (q = i × t). ............................... 54 Figure 3.4.4 Gas chromatograms obta ined after injection of 100 L of the electrolysis headspace before applying potential and after 60, 120 and 180 min of electrolysis. The peaks associated with N 2 and H 2 are growing in after each step. The residual O 2 in the headspace sample is attributed to the leaks. For full chromatogram See Appendix, Figure A3.4.4.1. ................................ .............. 55 Figure 3.5.1 XPS spectra of the glassy carbon electrode after the BE. Top: full spectrum. The source of silicon and oxygen is electrode contamination with silicon grease during the removal of the electrode from the electrolysis cell. The grease was heavily applied aro und the joints to prevent leaks. Bottom: the scan around 285 eV where the characteristic peak for ruthenium is expected to appear if depositions have had taken place. ................................ ................................ .................. 57 xi Figure 3.5.2 Rinse test results. Left: The CV recorded for a solution of NH 3 (sat d)/THF in which the rinsed glassy electrode was immersed. Scan rate 0.1 Vs - 1 . Right: Current vs. times for three electrolysis steps of elect rolysis. The inset is the gas chromatogram of the cell headspace injected to GC before the electrolysis and after the third hour of electrolysis. No H 2 is generated and the N 2 :O 2 ratio matches with leaks from the air. ................................ ................................ ................. 58 Figure 3.5.3 EDS analysis of the selected area (red circle in the inset) on the glassy carbon electrode at the end of a failed electrolysis experiment. The insets are the SEM images of the electrode with two different magnifications. These results show that a rutheniu m containing material has deposited on the surface, possibly acting as an active catalytic surface. ................. 59 Figure 4.1.1 The diagram of the three - electrode electrolysis cell used for chemical oxidation of 2 a to 2b . ................................ ................................ ................................ ................................ ............. 87 Figure 4.2.1 The 1 H NMR resonance shift observed for th e standard ferrocene after addition of paramagnetic 2b . ................................ ................................ ................................ ........................... 90 Figure 4.2.2 Top: CV of 4.5 10 - 3 M 2 b (green) compared to the CV of M 4.12 10 - 3 M 2 a (orange) in NM. Both complexes appear at the same E 1/2 , Bottom: Extended potential window CVs of 2.10 10 - 3 2b in NM shown for three scan rates. Here a second oxidation process is seen at 0.95 V vs. Fc +/0 (labeled as II). ................................ ................................ ................................ ................. 92 Figure 4.3.1 Total ion spectrum obtained for a solution containing 1.16 × 10 7 M 2 b and 1.072 × 10 7 M DBU in NM. ................................ ................................ ................................ ............................. 96 Figure 4.3.2 Assignments to four selected m/z values III, V, VI, and VII. The spectra in red are the experimentally obtained spectra and the inset shows the simulated spectra. .......................... 97 Figure 4.4.1 The structure of the mentioned Ru(IV)oxo species. ................................ .............. 101 Figure 4.5.1 CVs of a solution of 5 .36 10 - 3 M 2 b in NM before (green) and after (orange) addition of 1.34 10 - 2 M DBU. Scan rate 0.1 Vs - 1 . After the reaction with the base, the solution contains 2a and three new redox active species labeled as I, II and III. ................................ ........................ 102 Figure 4.5.2 CVs of a solution containing 4.30 10 - 3 M [Ru(trpy)(dmabp y)NH 3 ] + 3 , 2 b , in NM before (green) and after (red) addition of NH 3 (saturated, 0.93 M, determined by NMR measurements). The black curve shows the non - catalytic oxidation of NH 3 in NM. ................. 103 Figure 4.5.3 Vs of 4.30 10 - 3 M [Ru(trpy)(dmabpy)NH 3 ] + 3 in NM without NH 3 (green), and after addition of NH 3 (4.60 10 - 3 M, 1.06 equiv.) (brown).CV obtained for 2.80 10 - 3 M [Ru(trpy)(dmabpy)N 2 H 4 ](PF 6 ) 2 in NM in the absence of NH 3 . ................................ .................. 104 Figure 4.5.4 1 H NMR spectra (rt, 500 MHz, nitromethane - d 3 ) of 2 b (bottom, green), 2 b + excess DBU (middle, brown) and 2a (top, red).Comparison between the spectrum obtained after addition of NH3 to 2b and the spectrum of 2a in NM, shows that 2a is the main product of the reaction at room temperatures. The resonances at 1.65, 2.05, 2.97, 3.51 and 3.46 ppm are related to the added DBU. ................................ ................................ ................................ ................................ . 105 xii Figure 4.5.5 1 H NMR spectra (rt, 500 MHz, nitromethane - d 3 ) of 2 b + excess NH 3 (bottom, black), 2e (middle, blue) and 2a (top, red). By comparing the spectra, the major product of the reaction between 2b and NH 3 in NM is 2a . ................................ ................................ .............................. 106 Figure 4.5.6 Top: Changes in the electronic absorption spectra of a green starting solution of 7.65 10 - 5 M 2b absorbance at 490 nm due to generation of 2a happens simultaneously with the decrease in absorbance at 730 nm due to consumption of 2b . ................................ ................................ ...... 108 Figure 4.5.7 Comparing the electronic absorption spectra obtained at the end of the titration, i.e. 1.4 equiv. DBU added (orange) with the constructed spectrum for 0.5 e quiv. 2a in NM (red) and constructed spectrum for 0.5 equiv. 2a + DBU in NM (black). ................................ ................. 110 Figure 4.6.1 The schematic structure of an RR DE electrode is illustrated on the left. On the right, the redox events at the disk and ring are displayed for a redox active species with a generic CV response as shown on the top. ................................ ................................ ................................ ..... 112 Figure 4.6.2 Top: CVs obtained for a solution containing 5.57 10 - 4 M 2 a in THF collected at the ring and the disk show that the E 1/2 of the complex does not change with the material of the electrode. Bottom: The currents associated with the disk (i D ) and ring (i R ), when the potential of the disk is being scanned linearly and the ring is set at a constant reductive potential of 0.00 V vs. NHE. Three rotation speeds are shown: 200 rpm (red), 5 00 rpm (green) and 1000 rpm (blue). N = 24% (rotation speed independent), consistent with the value reported by the manufacturer. .... 115 Figure 4.6.3 Top: CVs obtained after addition of DBU to the solution containing 5.57 10 - 4 M 2 a in THF collected at the disk. Bottom: The currents associated with the disk and ring for the experiment with [DBU]= 1.10 10 - 5 M. Th ree rotation speeds are shown: 200 rpm (red), 500 rpm (green) and 1000 rpm (blue). N < 24% (rotation speed independent). ................................ ....... 117 Figure 4.6.4 RRDE LSVs for 5.57 10 - 4 M 2 a in THF with the disk being held at a potential more positive than the E 1/2 of 2a . Constant oxidative disk currents are labeled as i D . The product of the disk oxidation is being collected and reduced at the ring when E ring is lower than E 1/2 of 2a . Three rotation speeds are shown: 200 rpm (red), 500 rpm (green) and 1000 rpm (blue). N = 24% (rotation speed independent). ................................ ................................ ................................ ..................... 119 Figure 4.6.5 RRDE LSVs for the experiment with 5.57 10 - 4 M 2 a and [DBU]= 1.10 10 - 5 M in THF. Three rotation speeds are shown: 200 rpm (red), 500 rpm (green) and 1000 rpm (blue). The potential of the disk is held at the constant value of 1.0 V vs. NHE, while the potential of the ring was scanned linearly at a rate of 10 mVs - 1 . ................................ ................................ ................ 120 Figure 4.6.6 CVs taken at the disk without DBU (red) and after two additions of DBU. Top: raw data, Bottom: after the reversible processes (s econd peaks) were superimposed. ...................... 121 Figure 5.2.1 1 H NMR reference spectrum ( 600 MHz, THF - d 8 , 25 °C) of complex 2 e with (top) and without (bottom) excess NH 3 . The starred peak is solvent residual. The displacement of the bound hydrazine with free ammonia was not observed. ................................ ............................. 148 Figure 5.3.1 CVs of M 1.85 10 - 3 M 2 e in DCM. Top: the initial two cycles, Bottom: Three successive scans taken around 1 min after the CV in the top. Peak III appears after a couple scans xiii are taken in a freshly made solution, but it goes away w ith successive cycling of the potential. CVs taken after the one shown in the bottom, behave the same way. ................................ ................ 149 Figure 5.3.2 CVs of 1.8 5 10 - 3 M 2 e in DCM in the absent (black. Three successive cycles) and presence of ammonia (red). The red curve is very similar to the catalytic oxidation of ammonia in the presence of 2a . ................................ ................................ ................................ ...................... 150 Figure 5.3.3 CVs of 1.22 10 - 3 M 2 e in THF. Top: three successive cycles in a freshly made solution. Bottom: Changes in the CVs with scan rate (three cycles are shown for each scan rate). ................................ ................................ ................................ ................................ ..................... 151 Figure 5.3.4 CVs of 2.50 10 - 3 M 2 e in THF in the absence (black) and presence (red) of NH 3 . A shoulder that is evident at around - 0.35 is marked with an arrow. ................................ ............. 152 Figure 5.3.5 Cyclic voltammograms of a mixture solution of 1.05 10 - 3 M 2 a and 1.05 10 - 3 M 2 e in THF. Bottom: CVs of a mix ture solution of 2a and 1 equiv. NH 3 in THF. ........................... 153 Figure 5.3.6 Electro - oxidation of 4.0 10 - 2 M N 2 H 4 in THF at the glassy carbon electrode. .... 155 Figure 5.4.1 1 H NMR spectrum ( 500 MHz, THF - d 8 , 25 °C) of the reaction solution of 2 b and excess NH 3 (top). 1 H NMR spectra (500 MHz, THF - d 8 , 25 °C) of complex 2a (middle) and 2e obtained under same conditions, shown for comparison. ................................ ........................... 156 Figure 5.4.2 1 H NMR spectra ( 500 MHz, CD 2 Cl 2 , - 75 °C) of complex 2 a with ~ 4000 equiv. NH 3 (top), complex 2e with ~4000 equiv. NH 3 (middle), and the reaction mixture that results when ~ 4000 equiv. of NH 3 is added to 2b at - 85 °C. ................................ ................................ ............. 157 Figure 5.4.3 Change in the 1 H NMR ( 500 MHz, CD 2 Cl 2 ) spectral features as the temperature of the solution containing 2b and excess NH 3 is increasing. As the temperature is being increased, the resonances related to the hydrazine complex vanish and the NMR spectrum at room temperature o nly contains 2a . ................................ ................................ ................................ ..... 158 Figure 5.5.1 Changes in the absorption spectrum of a solution of 9.70 10 - 5 M 2 b with the addition of NH 3 . The green curve is the spectrum of the starting solution and the final orange spectrum is when the titration was ended (the point where the addition of NH 3 did not change the spectra). All absorption spectra are corrected for dil ution. ................................ ................................ ............. 160 Figure 5.5.2 Comparison between the absorption spectra recorded at the end of titration (orange), the constructed spectra f or 0.5 equiv. 2a (black) and 0.5 equiv. 2e (blue) and the arithmetic summation of the spectra of 0.5 equiv. 2a (black) and 0.5 equiv. 2e (blue) shown in dashed red. ................................ ................................ ................................ ................................ ..................... 161 Figure 6.1.1 Top: CVs of 3 a in the absence of NH 3 (sat d) in THF. Scan rates: 0.05 , 0.1 , 0.25 and 0.5 Vs - 1 . Bottom: The catalytic current (black) in a solution 2.7 mM 3a in THF. Dotted green line is the non - catalytic NH 3 oxidation. Scan rate is 0.1 Vs - 1 and the onset of the catalytic current is - 0.1 V versus Fc +/0 . ................................ ................................ ................................ ....................... 172 Figure 6.1.2 Top: Catalytic currents obtained in solutions of 2.7 mM 3 a and NH 3 (sat d) in THF and their dependence on the scan rate. Bottom: the catalytic currents normalized for scan rate. ................................ ................................ ................................ ................................ ..................... 173 xiv Figure 6.2.1 CVs of 4 a in the absence (top) and presence (bottom) of NH 3 (sat d) in THF. Scan rates for t he CVs on the top: 0.05, 0.1, 0.25, 0.5 and 0.8 Vs - 1 . On the bottom, the scan rate is 0.1 Vs - 1 and the onset of the catalytic current is - 0.15 V versus Fc +/0 . ................................ ............. 176 Figure 6.2.2 Top: Catalytic currents obtained in solutions of 2.51 mM 4 a and NH 3 (sat d) in THF and their dependence on the scan rate. Bottom: the catalytic curre nts normalized for scan rate. ................................ ................................ ................................ ................................ ..................... 177 Figure 6.3.1 The CVs obtained in solutions: 4.05 10 - 3 M 5 a in NM (red), 4.05 mM 5 a in the presence of NH 3 in NM (black) and a solution of NH 3 ................................ .......... 180 Figure 6.3.2 Top : The CVs of the ca talytic ammonia oxidation in NM in the presence of 4.05 10 - 3 M 5a , with different scan rates. Bottom: The normalized currents shown for the same set of data. The magnitudes of the normalized currents remain relatively constant at different scan rates. . 181 Figure 6.3.3 Top : CVs for 1.68 10 - 3 M 3 a in NM in the absence (red) and presence (black) of NH 3 . Scan rate 0.1 Vs - 1 . Bottom : CVs of the catalytic ammonia oxidation in NM in the presence of 1.68 10 - 3 M 3a , with different scan rates (left). The normalized currents shown for the same set of data (right). ................................ ................................ ................................ ............................. 182 xv LIST OF SCHEMES Scheme 2.1 Two possible mechanisms for coupling of terminal nitrides: (top) coupling of two nitridyl radicals, (bottom) nucleophilic nitride addition to an electrophilic terminal nitride. ...... 13 Scheme 2.2 Formation of high valence Ru(IV) and Ru(V)oxo species in catalytic water oxidation. Only the coordinating N centers in the polypyridyl ligands are shown for simplicity. ................ 14 Scheme 2.3 Catalytic water oxidation where O - O bond formation happens via nucleophilic attack of water to a Ru(V) oxo intermediate as proposed by Concepcion et a l. ref [53]. ....................... 15 Scheme 5.1 The hydrazine pathway to formation of the N - N bond. The nucleophilic attack of the free NH 3 to an imido nitrogen leads into formation of a hydrazine intermediate. The polypyridyl ligands are omitted. ................................ ................................ ................................ ..................... 145 Scheme 6.1 Synthesis of the dmaptrpy ligand. ................................ ................................ ........... 178 Scheme 7.1 The proposed catalytic cycle. The formation of the complexes highlighted in red has not yet been directly confirmed. ................................ ................................ ................................ . 195 xvi KET TO ABBREVIATIONS A Amperes, Surface Area (cm 2 ), Absorption ATR Attenuated Total Reflectance b Pathlength of light, 1 cm BE Bulk Electrolysis bpy 2,2´ - bipyridine BTU British Thermal Unit Bu Butyl C Coulomb, Concentration (M), Temperature (celsius) CE Counter Electrode CPE Controlled Potential Electrolysis CV Cyclic Voltammetry D Diffusion Coefficient (cm 2 s - 1 ) DBU 1,8 - diazabicyclo [5.4.0] undec - 7 - ene DCM Dichloromethane DI - MS Direct Infusion Mass Spectrometry DMSO Dimethyl sulfoxide dmabpy 4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridine dmaptrpy (4 - N,N - dimethylaminophenyl) - 2,2',6',2'' - terpyridine e - Electron o Formal Potential E 1/2 Half - Wave Potential xvii E p Peak Potential EDS Energy - dispersive X - ray Spectroscopy Molar Absorptivity (M - 1 cm - 1 ) ESI + Electrospray Ionization Positive Mode F Faraday Number Fc +/0 Ferrocenium/Ferrocene FE Faradaic Efficiencies FT Fourier Transform 1 H NMR Proton Magnetic Nuclear Resonance HRMS High Resolution Mass Spectrometry i p Peak Current K eq Equilibrium Constant LMCT Ligand - to - Metal Charge Transfer Me 3 trpy 4',4,4'' - trimethyl - 2,2':6',2'' - terpyridine MHz Mega Hertz MLCT Metal - to - Ligand Charge Transfer MS Mass Spectrometry NHE Normal Hydrogen Electrode nm Nanometer NM Nitromethane Ox Oxidized Form PCET Proton Coupled Electron Transfer ppm Part Per Million xviii QTOF Quadrupole Time - of - Flight rad Radian RE Reference Electrode Red Reduced Form RDE Rotating Disk Electrode RRDE Rotating Ring Disk Electrode rpm Revolution Per Minute Saturated SHE Standard Hydrogen Electrode SOFC Solid Oxide Fuel Cell t Bu Tert - Butyl t Bu 3 trpy 4',4,4'' - tri - tert - butyl - 2,2':6',2'' - terpyridine TCD Thermal Conductivity Detector THF Tetrahydrofuran TOF Turnover Frequency (s - 1 ) TON Turnover Number trpy 2,2':6',2'' - terpyridine UV Ultraviolet V Volts Vis Visible vs. Versus VT Variable Temperature WE Working Electrode xix XPS X - ray Photoelectron Spectroscopy 1 INTRODUCTION 2 1.1 . The Essence of Carbon Neutrality Greenhouse gases including carbon dioxide (CO 2 ), methane (CH 4 ), nitrous oxide (N 2 O) and temperature and consequently causes harmful impacts on the ecosystem. 1 The first three gases are mainly introduced to the atmosphere through the burning of fossil fuels (coal, natural gas, and oil) and agricultural activities. This is inevitable, as the burning of these fuels is the predominate way to fulfill that growing demand of our thriving population and industrial activities. The global emission of greenhouse gases has a fast - increasing trend, with just the fossil fuel - related CO 2 emissions, reached the al l - time high in 2018 and is projected to hit a record 37.1 billion metric tons by the end of this year. 2 Collectively, CO 2 , CH 4, and N 2 O compose the majority of the greenhouse emissions in the U.S. alone ( Figure 1.1 . 1 ). 3 This alarming release of greenhouse gasses has put the planet and its inhabitants in a critical situation and has urged us to look for alternative energy resources to replace the dependency on fossil fuel. Nevertheless, moving from n onrenewable to sustainable and clean energy sources is still a huge challenge. Figure 1.1 . 1 Relative amounts (in percent) of U.S. greenhouse gas emissions in 2016 , shows the large contribution of CO 2 . Figure reproduced from ref. [2]. British Thermal Unit Total = 97.7 quadrillion Btu (1) 3 The dependence of the industrial sector on efficient, cheap and easily accessible energy sources has restrained the demands in the renewable energy market. In 2017, around 80% of the U. S. energy sources were based on fossil fuels ( Figure 1.1 . 2 ), releasing 5.14 billion metric tons of CO 2 as a byproduct. 4 Figure 1.1 . 2 The U.S. energy consumption in 2017 from various sources shows the significance reliance on the fossil fuels . Figure reproduced from ref. [2]. The energy information agency projects that the global energy consumption will rise by 28% between 2015 and 2040, resulting in atmospheric energy - related CO 2 concentrati ons to increase by a rate of 0.6% per year. 5 However, global energy - related CO 2 emissions increased higher than expectations by 1.7% in 2018 to reach a historic high of 33.1 Gt CO 2 . It was the highest rate of growth since 2013, and 70% higher than the average increase since 2010. 6 To stop or slow down this growth, extensive efforts and inves tments has to be made to replace fossil fuels with renewable sources of energy. Strategies regarding integrating renewable energy sources in energy systems, including developing technologies to harvest and store renewable energy are already being pursued. However, issues regarding the cost and storage are not easy to be solved. With operating costs considered, onshore wind plants and large - scale photovoltaic plants are 4.6 and 14.1 times as expensive as gas plant, respectively. 7 Not only, the financial costs of building the 100% 4 renewable energy world are enorm ous, but the land area needed to accommodate the equipment required to harvest such diffuse sources of energy supply is just as daunting. Despite these concerns, the potential of wind and solar energy sources to expand their applicability is substantial, especially if efficient ways for their storage and transportation are existent. Nature stores sunlight in the chemical bonds of the products of photosynthesis: sugar, which is further synthesis to use photons of sunlight and atmospheric CO 2 amount of research. However, another not fully developed approach is to store sunlight in the chemical bonds of a carbon - free compound wit h a feasible storage and transportation technology. 1.2 . Carbon - Neutral Fuels Hydrogen (H 2 ) and ammonia (NH 3 ) are two, known carbon - free energy carriers that have been used and applied commercially. While hydrogen is not abundant in its molecular pure for m, it is mass. 8 H 2 th at can be sustainably produced from electrolysis of water. It is combusted with oxygen to reproduce water in zero - emission fuel cells with high efficiencies. 9 Due to its small molecular mass, H 2 has the highest gravimetric energy density (140.4 MJ kg - 1 ) of any fuel. This value is three times of the gravimetric energy density of gasoline (48.6 MJ kg - 1 ). 10 However, even when stored as compressed gas (700 bar, 15 C), its volumetric energy density is as low as 5.3 MJ L - 1 , significantly lower than the value for gasoline (31.1 MJ L - 1 ). 11 The safety issues regarding transportation of such pressurized H 2 tanks dema nd different safety considerations for onboard applications in hydrogen driven vehicles compared to electric or internal combustion cars. Such challenges are less prominent for NH 3 . Being a polar molecule with the capability of hydrogen 5 bonding, NH 3 liquif ication takes place at much lower pressures (10 bar at 20 C) compared to H 2 , resulting in relatively lower storage costs. 12 This also results in volumetric energy density of ammonia being almost three times higher th an H 2 , as well. In terms of driving range, a fuel tank containing 16 gal of ammonia provides a driving range of 470 miles, while the same volume of liquid hydrogen provides a driving range of 259 miles. 13 Table 1 summarizes the energy - related properties of H 2 and NH 3 , compared with gasoline as conventional fuel. Another attractive feature of NH 3 as a fuel and hydrogen carrier is its availability and accessibility. Ammonia is the second largest produced chemical in the world via the Haber - Bosch process, a revolutionary industrial method 14 developed in the first decade of the 20 th century that currently is responsible for the production of more than 200 million metric tons of ammonia per annum. 15 Additionally, the required infrastructure already exists for ammonia transportation and storage in large quantities. For example, a very large well - established transportation system based on pipelines, trains, trucks and barges exists in the Midwest and retail ammonia outlets exist in practically every state. 16 However, extension of the distribution network to more urban and industrial areas across the U.S. is still required. Table 1 . 1 Comparison of energy densities between hydrogen and ammonia as fuels . Gasoline is also shown as the reference. Energy Carrier H (%wt) Volumetric Energy Density (MJ l - 1 ) Gravimetric Energy Density (MJ kg - 1 ) H 2 ( l ), - 253 C, 1 bar 100 5.3 140.4 NH 3 ( l ), 25 C, 10 bar 17.6 13.6 17 22.5 6 Gasoline ( l ), 25 C, 1 bar 15.7 31.1 48.6 6 1.3 . Energy - Related Applications of Ammonia 1.3.1 . Using NH 3 in Internal Combustion Engines Ammonia can be directly burned in an internal combustion engine with minor modifications, emitting nitrogen and water vapor in the exhaust gas. Unburned ammonia and nitrogen oxides in i.e., a secondary fuel such as gasoline to overcome the low flammability of NH 3 a nd slow propagation of the flame. In 2007, an NH 3 Car departed Detroit and arrived in San Francisco a few days later, recording the longest trip made by an ammonia driven vehicle. In an NH 3 Car , a mixture of gasoline/ammonia was used to spark the ignition, but as the engine started the fuel was predominantly ammonia. 18 1.3.2 . Ammonia - Fueled Solid Oxide Fuel Cells (SOFC) Ammonia SOFC are characterized by their high operational temperatures (500 - 700 C) and are based on diffusion of gaseous NH 3 and air to an anode and cathode chamber, respectively. 19 However, in addition to the need of high operational temperatures, employment of SOFC for onboard applications is limited by the low stability of the catalysts, brittleness of the ceramic parts in the SOFC, cathode p oisoning arising from other components of the cathodic air feed and the possibility of formation of undesirable NO x byproducts. 20 22 1.3.3 . Ammonia as a Hydrogen Carrier A more interesting approach to a nitrogen fuel economy wo uld be to take advantage of the relatively practical storage and distribution of ammonia and to use it as a medium to store hydrogen which consequently allows storage of large amounts of energy. 23 Extraction of H 2 from NH 3 and 7 feeding the pure g enerated H 2 to an onboard hydrogen fuel cell is gaining attention. Recently Australian scientists at CSIR O ( ) were able to power vehicles with ammonia - derived hydrogen. of zero emission transportation. Although Haber - Bosch process is affording large - scale production of ammon ia, it i8s highly carbon - intensive. Not only the H 2 feed for this process is derived from natural gas (methane), also the energy production for the elevated temperatures and high pressures required for the chemical reaction is based on fossil fuels. Ammoni a manufacturing is not only a very energy consuming process (2% of global energy consumption), is it also responsible for %1 of the global CO 2 emissions. 24 When the challenges in decarbonization of NH 3 synthesis are overcome, the energy cycle provi ded by ammonia would be totally carbon neutral. In that regards, various strategies are currently being 25 through reduction of atmospheric N 2 coupled with splitting of H 2 O as the source of protons using renewable energy sources such as wind 26 , solar 27 29 and electrical energy 8, 30 . Once produced, ammonia is easily liquefied, stored and transported to the consumer, where the energy stored in N - H bonds has to be extracted. After splitting NH 3 i nto its elemental components, N 2 and H 2 , the latter can be further fed to a hydrogen fuel cell where it combusts with O 2 , leaving clean H 2 O vapor as the exhaust gas. Coupling photoelectrochemical water splitting with (photo)electrochemical N 2 reduction in the NH 3 synthesis plant would close the energy cycle as shown in Figure 1.3 . 1 . The Commonwealth Scientific and Industrial Research Organization, www.csiro.au 8 Figure 1.3 . 1 The ideal c arbon - neutral energy cycle of ammonia as a hydron carrier . 9 1.4 . Ammonia Splitting While a hypothetical carbon neutral cycle of ammonia as a fuel or hydrogen carrier is not far from reality, the challenge remains the r apid and efficient conversion of ammonia into nitrogen and hydrogen from ammonia: Thermal cracking, ammonolysis of metal hydrides and electrolysis of ammonia. 1.4.1 . Thermal Cracking of Gaseous NH 3 Decomposition of NH 3 ( g ) over a heterogeneous catalyst such as Ni at elevated temperatures 31 is the reverse reaction of the Haber - Bosch synthesis of ammonia: (1) This reaction is the primary choice for the production of H 2 that is used in alkaline hydrogen fuel cells, since the NH 3 remainings in the product gas mixture is problematic in the case of acidic prototypes, and the generated H 2 has to go through extra post - production purification processes. 32,33 Despite being mildly endothermic, catalytic decomposition of NH 3 requires temperatures as high as 600 C since the catalyst is more active at elevated temperatures. In addition to product impurity and high temperature operational co nditions, problems associated with the scale of equipment ( Figure 1.4 . 1 ) limit the application of thermal cracking to mainly industrial purposes rather than onboard hydrogen production. The challenge of sufficiently scaling down the reforming unit while yet maintai ning the capability to decompose ammonia at a rate in accordance with the consumption, has to be advised. 13 10 Figure 1.4 . 1 An ammonia cracker by Sam Gas Projects Ptv. Ltd. (India) which is maintained at 850 o C using an electric furnace . The p urity of the produced hydrogen is 99.5% (0.5% nitrogen contamination ) . Visit: https://www.psa - nitrogen.com 1.4.2 . Ammonolysis of Alkali Metal Hydrides The reaction between NH 3 and alkali metal hydrides, MH (M = Li, Na, and K), is an exothermic reaction that rel eases hydrogen at room temperature along with formation of alkali metal amides (MNH 2 ) as byproducts: 34 (2) The regeneration of the metal hydride is performed at high temperatures (300 C) and under a flow of H 2 over the metal amide product. During this process, decomposition of the MNH 2 to inactive products such as M 3 N and MNH via ammonia desorption deactivates the material. Also, the repetitive recycling of the material results in brittleness. 35 Thus, such systems fall short on fast, reversible generation of H 2 especially for onboard applications. 36 1.4.3 . Electrolysis of Ammonia Electrochemical splitting of ammonia in solution phase is a very attractive approach to extract high purity hydrogen for mobile applications. Most advances are related to electrolysis o f aqueous 11 alkaline solutions of ammonia, where at the anode ammonia undergoes oxidation to produce N 2 and protons, while H 2 is evolved at the cathode: (3) (4) (5) The thermodynamic potential for the oxidation reactions is - 0.77 V versus standard hydrogen electrode (SHE). The reduction of water in basic conditions requires an energy input of - 0.83 V versus SHE. Thus the driving potential for the overall reaction is +0.06 V, comparably less than the 1.23 V required for the analogous water splitting reaction. 37 The electrolysis is done at room temperature, and ambient pressure and the thermodynamic energy input is achievable by renewable sources such as electricity or solar. 37 Feasible coupling of ammonia electrolysis with hydrogen fuel cells is also another advantage with makes this method an excellent choice for on - demand hydrogen production. Though the electrolysis of ammonia appears to be very promising, there are some important issues associated with it. One is the highly corrosive nature of alkaline solutions of ammonia which abates the applications of this procedure. Also, while the thermodynamic potential of the ov erall reaction is small, large overpotentials are required to run the half reactions. This is more significant for the anodic oxidation of ammonia since the kinetics of the 6 electron/6 proton reaction of NH 3 oxidation are sluggish and large anodic overpot entials are required to initiate the oxidation. The selection of a suitable anode material is also a challenge since it has to simultaneously meet two essential criteria: providing a catalytic interface to reduce the kinetic overpotential and to be resisti ve to poisoning effects caused by strongly adhered NH x species. 38 Competitive formation of oxygen - containing products such as nitrate with N 2 is also an undesired process that lowers the faradaic 12 efficiencies. Electrolysis of pure NH 3 in its liquid state is an alternative approach taken to eliminate the interference of water in the electrolysis process. Several recent reports highlight the promising aspects of NH 3 ( l ) splitting. The thermodynamic potential of the overall splitting of NH 3 ( l ) is 0.1 V, only slightly higher than the value for aqueous electrolysis. 39 The problem with the high anodic overpotentials could be lessened by choosing a suitable anode material. For example, Dong et a l . were able to reduce the NH 3 ( l ) electrolysis voltage from 1.2 V to 0.47 V by switching from a Pt anode to a more active Rh - Pt - Ir alloy electrode. 40 However, the deactivation of the electrode at high current densities, which are mainly due to developments of a passivating layer of strongly adhered species on to the surface of the electrode, was still problematic. 41 Since the performance of most heterogeneous active surfaces seem to dimini sh due to poisoning, going towards homogenous catalysis has to be considered. A homogenous catalyst is a molecular substance added to the solution of ammonia to facilitate the kinetics of the reaction. In this scenario, the surface of the electrode is not actively involved in bond breaking and forming and only serves as a platform for delivery, or removal of electrons; thus, the poisoning effects have eliminated. Homogeneous systems also provide a molecular - level probe to understand the mechanism of the oxi dation of NH 3 to N 2 . The general methodology is based on activation of N - H bonds in an NH 3 ligand that is coordinated to a transition metal center. The energetics of the metal complex can be tuned by the right choice of spectator ligands, which consequentl y enables easier access to high oxidation states for the metal. As a result, lower onset potentials for oxidation of ammonia can be possibly achieved. However, a true molecular transition metal catalyst for the electro - oxidation of ammonia to N 2 has not be en reported. The first challenge is to achieve an actual catalytic turnover. A goal for which three hydrogens must be abstracted from the coordinated NH 3 and a triple bond has to form between two nitrogen molecules followed by the 13 evolution of the final N 2 atoms. Most previous reports of molecular transition metal complexes capable of ammonia activation, did not exhibit a full catalytic turnover. 42 44 More details and examples are discussed in the next section. 1.5 . Transition Metal Complexes as Catalysts for Ammonia Oxidation Chemical or electrochemical oxidization of NH 3 to terminal nitrides, bridged - N 2 or N 2 have been reported using metal complexes of Mn, 43 Mo, 45 Ir, 46 Ru, 47 and Os 48 . The formation of the terminally bound or bridged N 2 in the final stage is usually the consequence of coupling of two terminal nitride nitrogens via one of two coupling mechanisms: coup ling of the terminal nitridyl radicals, and nucleophilic - electrophilic coupling of the terminal nitrides ( Scheme 1 . 1 ). 43 Scheme 1 . 1 Two possible mechanisms for coupling of terminal nitrides: (top) coupling of two nitridyl radicals, (bottom) nucleophilic nitride addition to an electrophilic terminal nitride. While significant progress has been made towar ds N - H bond activation and NH 3 oxidation to various products, a fully closed catalytic cycle was not reported. However, mono - or binuclear transition metal complexes based on metals such as Ru and Os are very well studied for their catalytic role in chemic al or electrochemical oxidation of water to O 2 . 49 52 Most molecular 14 ruthenium water oxidation catalysts use polypyridyl spectato r ligands that were selected with careful attention given to their rigidity as well as the positioning of substituent groups having different steric and electronic effects. Polypyridines chelate with Ru through multidentate sites preventing ligand displace ment by water under acidic or alkaline conditions. Additionally, the polypyridine ligands are generally believed to be redox inactive when the Ru(II) metal centers are exposed to oxidizing conditions, allowing for the assumption that all electron transfer processes of the complexes are metal - based. Starting Ru(II) aqua catalysts, when treated with an chemical oxidant, undergo a series of proton and electron losses to achieve high valance Ru(IV or V)oxo species ( Scheme 1 . 2 ) that are considered to be the key intermediates in the catalytic cycle. Scheme 1 . 2 Formation of high valence Ru(IV) and Ru(V)oxo species in catalytic water oxidation. Only the coordinating N centers in the polypyridyl ligands are shown for simplicity. For example, Concepcion et al. suggested that changes in the UV - Vis absorption spectrum of an acidic solution of [Ru(trpy)(bpm)(OH 2 )] 2+ (See Appendix A1.2 for terminology) after addition of 3 equiv. of Ce(IV) as the chemical oxidizing agent, was due to the formation of a transient Ru(V) oxo species, which fur ther reacts with water to give a putative, not isolated [Ru III (trpy)(bpm)OOH] 2+ intermediate. Under catalytic conditions ( i.e., excess oxidant added) a 15 tentative [Ru IV (trpy)(bpm)O 2 ] 2+ species is formed from further oxidation of [Ru III (trpy)(bpm)OOH] 2+ . O 2 evolution is envisioned upon replacement of the coordinated dioxygen with water. The regeneration of the initial [Ru II (trpy)(bpm)OH 2 ] 2+ catalyst enables a closed catalytic cycle with 7.5 turnovers ( Scheme 1 . 3 ). 53 Scheme 1 . 3 Catalytic water oxidation where O - O bond formation happens via nucleophilic attack of water to a Ru(V) oxo intermediate as proposed by Concepcion et al. ref [53]. High valent ruthenium intermediates are hard to be isolated due to being very unstable and short - lived under catalytic conditions. Distinguishing between the two pathways of O - O formation (nucleophilic attack of water to Ru=O or the coupling between two Ru=O radicals which is more observed for bi - or multi - nuclear catalysts) is usually accomplished by isotopic labeling experiments using either the labeled catalyst, Ru= 18 O or labeled substrate 18 OH 2 . 54,55 Also, the radical coupling mechanism would present a second order kinetics in concentration of catalyst compared to the substrate attacking mechanism. 56 We focused our attention to the [Ru II (trpy)(bpy)OH 2 ] 2+ (See Appendix A1.2 for terminology) catalyst th at were intensively studied mostly by Meyer and his coworkers. 51 Berlinguette et al. 16 - donating methoxy grou ps enhanced the catalytic efficiency of the catalyst compared to the parent [Ru II (trpy)(bpy)OH 2 ] 2+ reflected by increased turnover frequencies (TOF). The rich catalytic chemistry of [Ru II (trpy)(bpy)X] 2+ complexes (X is the reactive neutral ligand) towards activation of small molecules was not only limited to ruthenium and/or O - O formation. Meyer and coworkers were simultaneously studying equivalent [Os II (trpy)(bpy)OH 2 ] 2+ complexes in the 1980s and expanded their work to analogous amine complexes a decade after. Interestingly, they observed that when electrooxidation of [Os II (trpy)(bpy)NH 3 ] 2+ at pH 7 had proceeded in the presence of excess concentrations of a secondary amine, an osmium hydrazido product was formed that was further isolated from the e lectrolysis solution and characterized: 57 , (8) They proposed the nucleophilic attack of the secondary amine on a transient Os(IV) imido species to be the N N bond - forming step. When the electro - oxidation was performed in the presence of primary amines or ammonia, an N 2 complex [Os II (trpy)(bpy)N 2 ] 2+ was yielded. 58 Meyer and Thompson in 1981 reported that exh austive oxidation of [Ru II (trpy)(bpy)NH 3 ] 2+ in aqueous solutions (pH 6.8) yielded the formation of a N - O bond in a [Ru II (trpy)(bpy)NO 2 ] + product, identified based on spectrophotometric and cyclic voltammetry experiments. 59 Inspired by the chemistry observed for N - N bond formation in osmium polypyridyl amine complexes, we hypothesized that similar molecular species would be able to act as catalysts for electro - oxidati on of NH 3 in dry organic solvents where the interference of H 2 O as an attacking nucleophile is prevented. 17 1.6 . This work The objective of this study is to evaluate ruthenium - based polypyridyl amine complexes as molecular electrocatalysts for oxidation of N H 3 . In that regard, several Ru amine complexes ( Figure 1.6 . 1 , 1 a to 5 a ) were synthesized and characterized, and their catalytic activity was examined using electrochemical methods. Amongst the prepared catalysts, compound [Ru II (trpy)(bpy)NH 3 ] 2+ ( 2a ) was selected for further mechanistic studies for this project. Gaseous products of the headspace were quantified by gas chromatography and a variety of spectroscopic techniques were employed to provide insights on the mechanistic details. The synthesis and exp erimental setups are fully explained in Chapter 2. In Chapter 3, the catalytic activity of [Ru II (trpy)(bpy)NH 3 ] 2+ ( 2a ) and some insights into the chemistry of catalytic oxidation of ammonia are discussed. Chapter 4 is devoted to one of the isolated interme diates and its reactivity towards additional bases. Chapter 5 and 6 include the primary results on other prepared catalysts and their catalytic activity for NH 3 oxidation. 18 Assignment Chemical Structure Chemical formula (1a) [Ru(trpy)(bpy)NH 3 ] 2+ (2a) [Ru(trpy)(dmabpy)NH 3 ] 2+ (3a) [Ru(Me 3 trpy)(dmabpy)NH 3 ] 2+ (4a) [Ru( t Bu 3 trpy)(dmabpy)NH 3 ] 2+ (5a) [Ru(dmaptrpy)(dmabpy)NH 3 ] 2+ Figure 1.6 . 1 Ruthenium catalysts 1 a - 5 a used in this study. 19 APPENDIX 20 A 1.1 . Terminology TON (Turnover Number) The number of moles of substrate that a mole of catalyst can convert before becoming inactivated and is the amount of substrate converted per the amount of catalyst used. TOF (Turnover Frequency), (time) - 1 The total number of moles transformed into the product by one mole of the active site per unit of time. 21 A 1.2 . Ligands Table A 1 . 1 . 2 Ligands mentioned or used in this study . Abbreviation Full name Structure trpy - terpyridine bpy - bipyridine bmp - bipyrimidine dmabpy - dimethylamoni - - dipyridine Me 3 trpy 4',4,4'' - trimethyl - 2,2':6',2'' - terpyridine t Bu 3 trpy 4',4,4'' - tri - tert - butyl - 2,2':6',2'' - terpyridine dmaptrpy (4 - N,N - dimethylaminophenyl) - 2,2',6',2'' - terpyridine 22 REFEREN CE S 23 REFERENCES (1) Greenhouse Gases Effect on Global Warming . https://svs.gsfc.nasa.gov/20114. (2) Jackson, R. B.; Le Quéré, C.; Andrew, R. M.; Canadell, J. G.; Korsbakken, J. I.; Liu, Z.; Peters, G. P.; Zheng, B. Environ. Res. Lett. 2018 , 13 (12), 120401. (3) U.S. Environmental Protection Agency (2018). Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 - 2016 https://www.epa.gov/sites/production/files/styles/large/public/2018 - 04/total_type.png. (4) U.S. Energy Information Administration. U.S. Energy Facts Explained https://www.eia.gov/energyexplained/?page=us_energy_home. (5) U.S. Energy In formation Administration. International Energy Outlook 2017 ; 2017. (6) International Energy Agency. Global Energy & CO 2 Status Report ; 2018. (7) Lyman, R. Friends of Science . 2016 . https://friendsofscience.org/assets/documents/Renewable - energy - cannot - replace - FF_Lyman.pdf (8) Los Alamos National Lab. PERIODIC TABLE OF ELEMENTS: LANL. https://periodic.lanl.gov/index.shtml (9) St affell, I.; Scamman, D.; Abad, V.; Balcombe, P.; Dodds, P. E.; Ekins, P.; Ward, K. R. Energy Environ. Sci. 2019 , 12 , 463 491. (10) Siddharth, K.; Chan, Y.; Wang, L.; Shao, M. Curr. Opin. Electrochem. 2018 , No. 5. (11) Gupta, R. B. Hydrogen fuel: production, transport and storage , 1st ed.; CRC Press, 2008. (12) Brown, T. The new generation of fuel cells: fast, furious, and flexible. https://www.ammoniaenergy.org/the - new - generation - of - fuel - cells - fast - furious - and - flexible/. (13) Zamfirescu, C.; Dincer, I. Fuel Process. Technol. 2009 , 90 (5), 729 737. (14) Erisman, J. W.; Sutton, M. A.; Galloway, J.; Klimont, Z.; Winiwarter, W. Nat. Geosci. 2008 , 1 (10), 636 639. (15) Giddey, S.; Badwal, S. P. S.; Kulkarni, A. Int. J. Hydrogen Energy 2013 , 38 (34), 14576 14594. 24 (16) Bartels, J . R . , "A feasibility study of implementing an Ammonia Economy" (2008). Graduate Theses and Dissertations. 11132. https://lib.dr.iastate.edu/etd/11132 . (17) Cheddie, D. In Hydrogen Energy - Challenges and Perspectives ; Minic, D., Ed.; IntechOpen, 2012; pp 333 362. (18) The NH 3 Car Project , www.NH3CAR.com. (19) Okanishi, T.; Okura, K.; Srifa, A.; Muroyama, H.; Matsui, T.; Kishimoto, M.; Saito, M.; Iwai, H.; Yoshida, H.; Saito, M.; Koide, T.; Iwai, H.; Suzuki, S.; Takahashi, Y.; Horiuchi, T.; Yamasaki, H.; Matsumoto, S.; Yumoto, S.; Kubo, H.; Kawahara, J.; Okabe, A.; Kikkawa, Y.; Isomura, T.; Eguchi, K. Fuel Cells 2017 , 1 8. (20) Simons, E. L.; Cairns, E. J.; Surd, D. J. J. Electrochem. Soc. 1969 , 116 (5), 556. (21) Cinti, G.; Discepoli, G.; Sisani, E.; Desideri, U. Int. J. Hydrogen Energy 2016 , 41 (31), 13583 13590. (22) Lan, R.; Tao, S. Electrochem. Solid - State Lett. 2010 , 13 (8), B83. (23) Schüth, F. Chemie Ing. Tech. 2011 , 83 (11), 1984 1993. (24) Service, R. F. Science (80). 2018 . DOI 10.1126/science.aau7489 . (25) Cherkasov, N.; Ibhadon, A. O.; Fitzpatrick, P. Chem. Eng. Process. Process Intensif. 2015 , 90 , 24 33. (26) Malmali, M.; Reese, M.; McCormick, A. V.; Cussler, E. L. ACS Sustain. Chem. Eng. 2018 , 6 (1), 827 834. (27) Shiraishi, Y.; Shiota, S.; Kofuji, Y. ; Hashimoto, M.; Chishiro, K.; Hirakawa, H.; Tanaka, S.; Ichikawa, S.; Hirai, T. ACS Appl. Energy Mater. 2018 , 1 , 4169 4177. (28) Wang, L.; Xia, M.; Wang, H.; Huang, K.; Qian, C.; Maravelias, C. T.; Ozin, G. A. Joule 2018 , 1 20. (29) Bi, Y.; Wang, Y.; Dong, X.; Zheng, N.; Ma, H.; Zhang, X. RSC Adv. 2018 , 8 (39), 21871 21878. (30) Shipman, M. A.; Symes, M. D. Catal. Today 2016 , 1 12. (31) Bell, T. E.; Torrente - Murciano, L. Top. Catal. 2016 , 1 20. (32) Chellappa, A. S.; Fischer, C. M.; Thomson, W. J. Appl. Catal. A Gen. 2002 , 227 (1 2), 231 240. 25 (33) Choudhary, T. V.; Sivadinarayana, C.; Goodman, D. W. Catal. Letters 2001 , 72 (3), 197 201. (34) Yamamoto, H.; Miyaoka, H.; Hino, S.; Nakanishi, H.; Ichikawa, T.; Kojima, Y. Int. J. Hydrogen Energy 2009 , 34 (24), 9760 9764. (35) Yamaguchi, S.; Miyaoka, H.; Ichikawa, T.; Kojima, Y. Int. J. Hydrogen Energy 2017 , 42 (8), 5213 5219. (36) Vegge, T.; Klerke, A.; Christensen, H.; Nørskov, K.; Vegge, T. J. Mater. Chem. 2008 , 18 (20), 2304 2310. (37) Vitse, F.; Cooper, M.; Botte, G. G. J. Power Sources 2005 , 142 (1 2), 18 26. (38) Bunce, N. J.; Bejan, D. Electrochim. Acta 2011 , 56 (24), 8085 8093. (39) Little, D. J.; Smith, III, M. R.; Hamann, T. W. Energy Environ. Sci. 2015 , 8 , 2775 2781. (40) Dong, B. X.; Tian, H.; Wu, Y. C.; Bu, F. Y.; Liu, W. L.; Teng, Y. L.; Diao, G. W. Int. J. Hydrogen Energy 2016 , 41 (33), 14507 14518. (41) Ye, J.; Lin, J.; Zhou, Z.; Hong, Y.; Sheng, T.; Rauf, M. J. Electroanal. Chem. 2018 , 819 (November 2017), 495 501. (42) Margulieux, G. W.; Bezdek, J.; Turner, Z. R.; Chirik, P. J. J. Am. Chem. Soc. 2017 , 139 , 6110 6113. (43) Keener, M.; Peterson, M.; Hernández Sánchez, R.; Oswald, V. F.; Wu, G.; Ménard, G. Chem. Eur. J. 2017 , 23 , 11479 11484. (44) Chang, Y.; Nakajima, Y.; Tanaka, H.; Yoshizawa, K.; Ozawa, F. J. Am. Chem. Soc. 2013 , 135 , 11791 11794. (45) Bhattacharya, P.; Heiden, Z. M.; Wiedner, E. S.; Raugei, S.; Piro, N. A.; Kassel, W. S.; Bullock, R. M.; Mock, M. T. J. Am. Chem. Soc. 2017 , 139 (8), 2916 2919. (46) Scheibel, M. G.; Abbenseth, J.; Kinauer, M.; Heinemann, F. W.; Würtele, C.; De Bruin, B. ; Schneider, S. Inorg. Chem. 2015 , 54 (19), 9290 9302. (47) Ishitani, O.; Ando, E.; Meyer, T. J. Inorg. Chem. 2003 , 42 (5), 1707 1710. (48) Demadis, K. D.; Meyer, T. J.; White, P. S. Inorg. Chem. 1998 , 37 , 838 83 9 . (49) Tseng, H. - W.; Zong, R.; Muckerman, J. T.; Thummel, R. Chem. Sci. 2016 , 47 (24), 11763 11773. 26 (50) Wada, T.; Nishimura, S.; Mochizuki, T.; Ando, T.; Miyazato, Y. Catalysts 2017 , 7 (2), 56. (51) Takeuchi, K.; Thompson, M.; Pipes, D. W.; Meyer, T. J. Inor g. Chem. 1984 , 23 (13), 1845 1851. (52) Gilbert, J. A.; Eggleston, D. S.; Murphy, W. R.; Geselowitz, D. A.; Gersten, S. W.; Hodgson, D. J.; Meyer, T. J. J. Am. Chem. Soc. 1985 , 107 , 3855 3864. (53) Concepcion, J. J.; Jurss, J. W.; Templeton, J. L.; Meyer, T. J. J. Am. Chem. Soc. 2008 , 130 (49) , 16462 16463. (54) Bozoglian, F.; Romain, S.; Ertem, M. Z.; Todorova, T. K.; Sens, C.; Mola, J.; Benet - buchholz, J.; Fontrodona, X.; Cramer, C. J.; Gagliardi, L.; Llobet, A. J. Am. Chem. Soc. 2009 , 131 , 15176 15187. (55) Romain, S.; Bozoglian, F.; Sal a, X.; Llobet, A. J. Am. Chem. Soc. 2009 , 131 , 2768 2769. (56) Shaffer, D. W.; Xie, Y.; Concepcion, J. J. Chem. Soc. Rev. 2017 , 46 , 6170 6193. (57) Coia, G. M.; White, P. S.; Meyer, T. J.; Wink, D. A.; Keefer, L. K.; Davis, W. M. J. Am. Chem. Soc. 1994 , 116 (4), 3649 3650. (58) Coia, G. M.; Devenney, M.; White, P. S.; Meyer, T. J.; Wink, D. a. Inorg. Chem. 1997 , 36 (11), 2341 2351. (59) Meyer, T. J.; Thompson, M. S. J. Am. Chem. Soc. 1981 , 103 , 5579 5581. 27 EXPERIMENTAL DETAILS 28 In this work, a variety of electrochemical and spectrochemical techniques were employed. This chapter focuses on the experimental procedures followed for those experiments. Preparation of some chemicals are also explained here. 2.1 . General Materials and Methods Concentrated ammonium hydroxide, hydrazine monohydrate, and solvents used in the synthesis of ruthenium complexes were reage nt grade and used as received without further purification. Deuterated solvents and solvents used in electrochemistry were rigorously dried according to standard procedures prior to use. 1 Ruthenium trichloride hydrate, RuCl 3 ·xH 2 O, was purchased from Sigma - - Terpyridine (trpy), 97% was purchased from Alfa Aesar (USA) and recrystallized from hexane prior to use as follows: - terpyridine (%97) was dissolved in 150 mL boiling hexane. The solution was fi ltered hot through a filter paper. Afterward, the volume was reduced to one - third by rotovaping. The flask content was heated to 80 C to dissolve the solids that were formed during the solvent evaporation step. After all the solids were dissolved, the fla sk was left under a slow flow of nitrogen at room temperature to cool down overnight. Pale yellow crystals formed. The flask was attached to the vacuum line of a Schlenk line to evaporate the residual hexane and then was vacuum - dried for 48 h to complete d ryness. Tetrabutylammonium hexafluorophosphate [Bu 4 N](PF 6 ) 97% was purchased from Alfa Aesar (USA) and was recrystallized from ethanol following a similar procedure used for trpy. Ammonium hexafluorophosphate 99.5% [NH 4 ](PF 6 ) was obtained from Alfa Aesar and was used without further recrystallization but was dried in a vacuum oven set at 50 C for 48h and then stored in the glovebox. 15 NH 4 PF 6 was made by exchanging the chloride ion in 15 NH 4 Cl (99% atom 15 N, Cambridge Isotope Laboratories, Inc., USA) with hexafluorophosphate by mixing 15 NH 4 Cl 29 and AgPF 6 (Sigma Aldrich, USA) 1:1 in deionized water. The AgCl precipitate was filtered twice using a Celite column. The product was collected as a solid after water was evaporated and then dried under vacuum overnight. - bis(dimethylamino) - - bipyridine (dm - bipyridine (bpy) were purchased from Hetcat (Switzerland) and Sigma - Aldrich (USA), respectively and used without further purification. Nuclear magnetic resonance (NMR) spectra were recorded on a Varian Innova 600 MHz spectrometer equipped wi th a triple resonance indirect probe or a Varian 500 MHz DD2 spectrometer equipped with a Pulsed Field Gradient (PFG) Probe. Variable temperature NMR experiments were performed on a Varian Unity Plus 500 MHz spectrometer equipped with a NALORAC 5 mm PFG pr obe. CHN analyses were performed by Midwest Micro Lab (IN, USA). High - resolution mass spectra (HRMS) were obtained at the Michigan State University Mass Spectrometry Core using quadrupole time - of - flight instruments (QTOF). IR spectra was collected using a Jasco FT/IR 6600 spectrophotometer equipped with ATR PRO ONE Single - reflection ATR Accessory. 2.2 . Preparation of Dry Liquid Ammonia Ammonia gas (Airgas, USA) was condensed over metallic sodium in a previously dried one - neck round bottom flask equipped w ith two gas inlet arms, immersed in a methanol bath under an inert atmosphere of N 2 or Ar at ambient pressure ( Error! Reference source not found. ). The tem perature of the methanol bath was set t o - 55 ° C using a temperature control chiller equipped with a metal thermometer probe. The dry ammonia gas was then cannula transferred into another receptacle which could be either the electrochemistry cell or a secondary storage flask, depending on the experiment purpose. The quenching of the nonreacted sodium residuals in the leftover NH 3 30 was pursued by dropwise addition of isopropyl alcohol, followed by acetone and finally water until the final solution was clo udy white. This solution was then diluted with 1 L of water and was disposed in to an aqueous waste container. Figure 2.2 . 1 The apparatus used for the storage of dry liquid ammonia under ambient pressure. 2.3 . Preparation of Saturated 15 NH 3 Solutions in Tetrahydrofuran (THF) 15 N - labeled ammonia gas (98% atom 15 N) was purchased from Sigma Aldrich in 1 L quantities contained in a lecture - sized cylinder and used as received. A high - vacuum Schlenk line was used to transfer measured volumes of ammonia gas into heavy walled reaction vessels and gas - tight, Teflon - valve d, medium pressure, high vacuum NMR tubes. Labeled ammonia in THF was prepared as a supersaturated solution by vacuum transfer of excess 15 NH 3 (9 mmol, 250 mL at 1 atm and 25 ° C) into 5.5 mL dry THF in a 20 mL heavy walled, high - pressure vessel. Before the transfer of 15 NH 3 , THF was subjected to 3 cycles of freeze/thaw degassing. The 15 NH 3 /THF solution was thawed in a dry ice/ethanol bath. Before transfer to the electrochemical cell, the cold bath was removed, and the vessel was filled with argon. The Teflo n plug was replaced with a Suba - Seal® septum under an argon counterflow and the solution was stirred for approximately 1 min under a static argon atmosphere after closing off the counterflow. Then, the required volume of the solution 31 was withdrawn by syrin ge and injected into the electrochemical cell while the solution was still cold. 2.4 . Electrochemistry All electrochemical experiments were conducted under inert atmosphere (N 2 and Ar for voltammetry and chronoamperometry experiments, respectively). Note : Due to significant interference of residual water, it is crucial to rigorously dry all solvents and materials prior to electrolysis. 2.4.1 . Cyclic Voltammetry (CV) either 0.1 M [Bu 4 N](PF 6 ) or 0.1 M [NH 4 ](PF 6 ) as the supporting electrolyte. For electrochemical experiments in nitromethane (NM) 0.2 M [NH 4 ](PF 6 ) was used as the supporting electrolyte. Glassy carbon (disk: geometric surface area = 0.07 cm 2 , plate: geometric surface area=2.0 cm 2 ), a custom silver/ saturated silver nitrate/methanol and a platinum mesh were used as the working, reference and counter electrodes, respectively. The diagram of the electrochemistry cell is shown in Figure 2.4 . 1 . Figure 2.4 . 1 Diagram of the three - electrode electrochemical cell used for CV experiments. The side arms are inert gas inlet and outlets. CE: Counter Electrode, WE: Working Electrode and RE: Reference Electrode. 32 The potential of the reference electrode was measured ver sus ferrocene/ferrocenium (Fc +/0 ) in each solvent. Figure 2.4 . 2 shows how E 1/2 of ferrocene was measured versus Ag/AgNO 3 / methanol reference electrode in THF and NM. Figure 2.4 . 2 CVs (three cycles) of ferrocene in THF and NM using the Ag/AgNO 3 reference electrode . Top : 3.0 10 - 3 M ferrocene in THF , and Bottom : 1.6 10 - 3 M ferrocene in NM. CVs were obtained at 0.1 V s - 1 scan rate. E p,a and E p,c stand for ano dic and cathodic peak potentials, respectively. The E 1/2 is defined as the midpoint between E p,a and E p,c and is calculated as their arithmetic average. 33 2.4.2 . Determin ation of the Onset Potential The onset potential of the current is defined as the intersection of the baseline current with the linear portion of the oxidation current. Figure 2.4 . 3 represents the onset potential determination for non - catalytic oxidation of ammonia at the surface of the electrode in THF. Figure 2.4 . 3 Determination of the onset potential for oxidation of ammonia at the surface of the glassy carbon electrode in TH F. The potential of the cross - point between the baseline and the oxidation current is considered as the onset potential. 2.4.3 . Hydrodynamic Voltammetry Two types of hydrodynamic voltammetry experiments were used in this study, one with a rotating disk electrode (RDE) and the other with a rotating ring - disk electrode (RRDE). A flat bottom glass cell was used fo r those experiments ( Figure 2.4 . 4 ). The speed of rotation was controlled remotely using an Autolab RDE consisted of the rotating mechanical unit (RDE - 2) and the motor control unit (MCU). 34 RRDE electrodes (Autolab) were used as WE which the tip was fitted with a 5 mm glassy carbon collection efficiency of the RRDE was 24.9%. The rotation speeds ( ) were applied in revolution per minute (rpm) and were converted to rad s - 1 using the following calculation: (1) Figure 2.4 . 4 Diagram of the three - electrode electrochemical cell used in RDE and RRDE experiments. The side arms are inert gas inlet and outlets. CE: Counter Electrode, WE: Working Electrode and RE: Reference Electrode. 2.4.4 . Controlled Po tential Electrolysis (CPE) CPE or Bulk electrolysis (BE) experiments were performed with a constant applied potential to the working electrode, while the solution was vigorously stirring. For the electrochemical synthesis of [Ru(trpy)(dmabpy)NH 3 ] 3+ , 2b , the counter electrode was consist ed of a silver coil immersed in a solution of 0.1 M AgNO 3 and 0.1 M [Bu 4 N](PF 6 ) in acetonitrile. This solution was separated from the electrolysis solution with a fritted glass bridge. For the b ulk electrolysis of ammonia in THF in the presence of the catalyst, the counter electrode was platinum mesh and was directly inserted into the electrolysis solution in order to allow the reduction of anodically produced protons. Desired masses of the catal yst and [NH 4 ](PF 6 ) were weighed out and placed in the cell 35 ( Figure 2.4 . 5 ). The cell was sealed and kept under a slow flow of Ar overnight. Prior to solvent transfer, the cell was disconnected from the Ar line and was kept closed for 3 hours, during which les of the cell headspace were injected to the gas chromatograph instrument (See Section 2. 5 ) every hour to check for leaks. Then 2.5 mL of saturated NH 3 /THF was transferred to the cell by injecting through the injection/sampling port ( Figure 2.4 . 5 , inset). Three steps of BE were performed by applying a constant potential of 0.15 V versus Fc +/0 to the working electrode taken out with a glass syringe for U V - Vis measurements. Bulk electrolysis experiments using isotopic labeled reagents were performed in the same fashion. Some other BE experiments were conducted in a cell set up similar to what is represented in Figure 2.4 . 1 . Figure 2.4 . 5 The cell used in some controlled potential electrolysis experiments. An empty cell is presented here for better visibility. The counter electrode is Pt mesh, and working is glassy carbon plate. The inset shows the position of the sampling port. 36 2.5 . Gas C hromatography (GC) Gas quantifications were conducted with a gas chromatograph (Agilent 7820A) equipped with an Agilent - PLOT/U capillary column (Agilent 19095P - UO4PT 30 m - molsieve 5Å capillary column (Agilent CP7539 50 m conductivity detector (TCD). Argon was used as the carrier gas. The PLOT/U capillary column is connected to the injector and exits to a pneumatic valve that switches between the CP - molsieve 5Å capillary column and the detector. Ten mi nutes after injection, the pneumatic valve switches to isolate the faster eluting, volatile diatomic gases in the molsieve column, allowing the slower eluting volatiles THF and NH 3 to pass from the PLOT/U column directly to the detector. At ca. 17 min, the valve resets, and the diatomics in the molsieve column then elute to the detector. The full chromatogram is shown in Figure 2.5.1. Figure 2.5 . 1 Full TCD Chromatograms obtained for two headspace injections. Injection A (black): Headspace of a cell containing THF and the catalyst. Injection B (red): Headspace of the cell after 2 h of electrolysis in the presence of NH 3 and the catalyst. At point 1, a GC column valve switches to isolate N 2 , H 2 , and residual O 2 in the molsieve column, while heavier volatiles, THF, and NH 3 , elute through the PLOT/U column to the detector during time window 2. At time point 3, the column isolation valve resets and diato mic gases elute through the molsieve column to the thermal conductivity detector as seen in time window 4. 37 2.5.1 . GC Calibration Evolution of H 2 and N 2 was quantified based on an independent calibration obtained separately for each gas by direct injection of kn own volumes of a gaseous mixture of 5% V hydrogen and 95% V nitrogen (Airgas, USA) to the GC. The moles of N 2 and H 2 are calculated using the ideal gas law with T = 293 K and P = 1 atm ( Table 2 . 1 ). The data points for 0 mol N 2 and H 2 in Figure 2.5 . 2 gon gas (Blank), corrected for residual N 2 in the argon gas used in during electrolysis giving a peak with an area of 28.84. Table 2 . 1 Experimental data used for GC signal calibration for moles of injected N 2 and H 2 . Red and blue colored numbers are used to construct the corresponding calibration lines in Fig. 2.5.2. Volume of gas injected Volume of H 2 in the injection Volume o f N 2 in the injection Moles of H 2 injected a Moles of N 2 injected a H 2 peak area N 2 peak area Corrected N 2 peak area b 50 2.5 47.5 1.04 10 - 7 1.97 10 - 6 98.62 295.32 266.48 80 4.5 76.0 1.66 10 - 7 3.16 10 - 6 157.67 466.79 437.95 100 5.0 95.0 2.08 10 - 7 3.95 10 - 6 207.20 589.95 561.11 a Calculated as n = (PV)/(RT). b Corrected N 2 peak area = N 2 peak area 28.84. 38 Figure 2.5 . 2 Gas Chromatography calibration lines obtained for N 2 ( top , blue ) and H 2 ( bottom , red ) based on data in Table 2. 1 . 2.6 . GC - MS Experiment Isotopic N 2 production was monitored by mass spectrometry with a Shimadzu QP2010SE GC/MS spectrometer modified with an Agilent J & W CP - molecular sieve column installed (5Å, 50 m y = 1.41 10 8 x 4.80 y = 9.81 10 8 x 1.81 39 530 µm 50 µm) and helium as the carrier gas. The N 2 was chromatographically separate d from other gases in the mixture, and then mass - to - charge values of 28 ( 14 N 14 N), 29 ( 15 N 14 N) and 30 ( 15 N 15 N) were determined by quantitative integration of parent peak intensities. The amounts of gases were not quantified in GC - MS experiments. The instru ment was modified, and the analysis was carried out according to the following description. An Agilent J & W CP - molecular sieve column was installed (CP7539 50 m 0.25 mm ID into the detector. The g uard column was cut from a Varian FactorFour WCOT fused silica 0.25 mm ID column and connected with a Silitek treated MXT union connector. The injection volume of the cell headspace was 100 µL with split ratio 50:1 and a flow rate of 3 mL min - 1 , with an in jector temp = 100 ° C. 2.7 . X - Ray Spectroscopy Scanning Electron Microscope (SEM) images were obtained using a Carl Zeiss EVO LS25 microscope. Energy - dispersive X - ray spectroscopy (EDS) measurements of selected areas were performed using an EDAX instrument equipped with a Pegasus camera. 40 REFERENCES 41 REFERENCES ( 1 ) Armarego, W. L. T. Purification of Laboratory Chemicals , 6th ed.; Butterworth - Heinemann, 2009. 42 CATALYTIC ACTIVITY O F [RU(TRPY)(DMABPY)NH 3 ] 2+ IN THF 43 Inspired by Meyer and Thompson results using Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 , 1 we started this study based off of the similar ruthenium (II) ammine com plex and investigated the chemistry when it was exhaustively oxidized in the presence of ammonia in organic solvents. T his chapter focuses on initial electrochemical studies of complexes [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1a ) and [Ru(trpy)( dma bpy)NH 3 ](PF 6 ) 2 ( 2a ) with and without ammonia in THF. 3.1 . Synthesis of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1a) 1a was synthesized in three steps as described below . 3.1.1 . Synthesis of (2,2´:6´,2´´ - terpyridyl) trichloro ruthenium (III), [Ru(trpy)Cl 3 ]. [Ru(trpy)Cl 3 ] was prepared by a literature procedure and characterized by ESI + - MS: m/z: 404.9 [M - Cl + ]. 2 Trichloro complexes of ruthenium (III) with o ther derivatives of trpy were synthesized by the same procedure. 3.1.2 . (2,2´:6´,2´´ - terpyridyl) (2,2´ - bipyridyl) chloro ruthenium (II) chloride, [Ru(trpy)(bpy)Cl]Cl. 44 [Ru(trpy)(bpy)Cl]Cl was synthesized from [Ru(trpy)Cl 3 ] using a previously reported procedure. 3 1 H NMR (500 MHz, DMSO - d 6 ) (d, J = 8.1 Hz, 2H), 8.70 (d, J = 8.0 Hz, 2H), 8.64 (d, J = 8.1 Hz, 1H), 8.36 (td, J = 7.9, 1.6 Hz, 1H), 8.22 (t, J = 8.1 Hz, 1H), 8.07 (ddd, J = 7.3, 5.6, 1.3 Hz, 1H), 7.99 (td, J = 7.8, 1.5 Hz, 2H), 7.78 (td, J = 7.8, 1.4 Hz, 1H), 7.62 (dd, J = 5.6, 1.5 Hz, 2H), 7.38 (ddd, J = 7.2, 5.5, 1.3 Hz, 2H), 7.32 (dd, J = 5.8, 1.4 Hz, 1H), 7.08 (ddd, J = 7.3, 5.7, 1.3 Hz, 1H). 3.1.3 . (2,2´:6´,2´´ - Terpyridyl)(2,2´ - bipyridyl)ruthenium(II) ammine dihexafluoro phosphate, [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 , (1a). [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 was synthesized from [Ru(trpy)(bpy)Cl]Cl and characterized by 1 H NMR spectroscopy using a previously reported procedure. 4 3.2 . Catalytic Activity of 1a in THF Cyclic voltammetry of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 in THF shows a reversible one - electron redox process at E 1/2 = 0.40 V versus Fc +/0 corresponding to oxidation of Ru(II) center to Ru(III) ( Figure 3.2 . 1 ) . In this CV experiments, a s the potential is scanned in the positive direction, oxidation of the analyte results in a n anodic current . In the reverse scan, reduction gives rise to a catho dic c urrent . The half - potential of the redox process ( E 1/2 ) is determined as the midpoint between the potentials corresponding to the maximum anodic and cathodic currents (i p,a and i p,c , respectively) . 45 Scans to more positive potentials were constrained by the oxidation potential window of the THF solvent. versus square root of the scan rate are linear with a ratio of 1.03. From the slope of the anodic line , a diffusion coefficient, D o x , of 4.06 10 - 6 cm 2 s - 1 was calculated according to equation 1. 5 (1) Where n is number of transferred electrons (1 e - ), A is the surface area of the working electrode ( 0.07 cm 2 ), is the bulk concentration of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 2.13 10 - 6 mol cm - 3 ), is the scanning rate (V s - 1 ), T = 23 C, R = 8.314 J mol - 1 K - 1 and F = 96485 C. The onset potential of the current arising from direct oxidation of NH 3 at the glassy carbon working electrode was measured vs. Fc +/ 0 as shown in Figure 3.2 . 2 . Electro - oxidation of saturated solutions of NH 3 in THF ([NH 3 ]=0.34 M) in the presence of 2.13 10 - 3 M [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 results in an enhanced current (red solid line curve in Figure 3.2 . 2 ) with an onset potential lower than the direct NH 3 oxidation in the absence of the complex . 46 Figure 3.2 . 1 Top : Scan rate dependence of the current in 2.13 × 10 - 3 M [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1 a ) in THF shown for sev en scan rates. Bottom : Plots of anodic and cathodic peak currents obtained from CVs on the top versus square root of the scan rate. From the slope of the anodic branch a diffusion coefficient of D ox = 4.06 10 - 6 cm 2 s - 1 is calculated for 1a . 47 Figure 3.2 . 2 Cyclic voltammograms for a solution containing ferrocene in THF with (red) and without (blue) NH 3 (0.34M). The onset of the direct ammonia oxidation was measured as +0.25 V vs . ferrocene. In solutions that do not contain fer rocene, the oxidation of ammonia appears at the same potential using a n Ag/ A gNO 3 reference electrode that is separately calibrated with Fc +/0 in THF. The disappearance of the return current in the CV of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1a ) and the lower onset potential both are consistent with the enhanced current being catalytic and not merely being the mathematical summation of currents for [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1a ) and direct NH 3 oxidation. Additionally, when the currents are normaliz ed for the scan rate, the catalytic activity appears to be improved for longer time scales, i.e. slower scan rates, suggesting that the current is controlled by a slow reaction which is improved when the time scale of the reaction is increased ( Figure 3.2 . 3 , bottom). 48 Figure 3.2 . 3 Top : Cyclic voltammograms for THF solutions of 2.13 × 10 - 3 M catalyst 1 a (blue) , 0.34 M NH 3 added to the solution of 2.13 ×10 - 3 M catalyst 1a (red) and direct NH 3 oxidation in THF in the absence of 1a (dotted green). Scan rate 0.1 V s - 1 . Bottom : Normalized catalytic (i / ) currents when NH 3 (0.34 M) is added to a solution of 2.13×10 - 3 M 1a obta ined for mentioned scan rates. The magnitude and the onset of the normalized current s are being improved as the scan rate is decreasing. 49 S ince the onset of the catalytic current is only 80 mV more negative than noncatalytic ammonia oxidation, it is hard to assess the catalytic current without the interference of background NH 3 oxidation. All these factors are related to the relatively positi ve E 1/2 for [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1a ) . It is expected that catalysts with redox potentials more negative than 0.25 V versus Fc +/0 (NH 3 oxidation onset) c ould enable a clear observation of a background - free catalytic current and hence more precise insights to the catalysis mechanism. Controlling the electron density at the ruthenium metal center via structural modifications of the polypyridyl ligands is a well - stablished approach to tune th e catalytic activity of ruthenium polypyridyl catalysts. 6 , 7 Installing electron - donating substituent groups on the trpy and bpy moieties, directly affects the redox potential of the complex and therefore the catalytic activity via making the higher oxidation states of the metal more accessible. 2 Aiming to lowe r the redox potential, [Ru(trpy)( dma bpy)NH 3 ](PF 6 ) 2 ( 2a ) was employ ed by utilizing a more electron rich 4,4´ - bis(dimethylamino) - - bipyridine (dmabpy) ligand. 3.3 . Synthesis of [Ru(trpy)( dma bpy)NH 3 ](PF 6 ) 2 ( 2a ) 2a was synthesized from [Ru(trpy)Cl 3 ] in two steps as described below . 3.3.1 . (2,2´:6´,2´´ - Terpyridyl)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl)chloro ruthenium(II) chloride, [Ru(trpy)(bdmabpy)Cl]Cl. 50 (2,2´:6´,2´´ - Terpyridyl)trichloro ruthenium (200 mg, 0.45 mmol, 1 equiv.), 4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridine (110 mg, 0.45 mmol, 1 equiv.), and LiCl (95 mg, 2.25 mmol, 5 equiv.) were weighed into a 150 mL round bottom flask equipped with a stir bar. A 1:1 mixture of methanol/water (50 mL) was added as a solvent, and the resulting dark slurry was stirred under a positive pressure of N 2 . Triethylamine (0.5 mL) was added by syringe. The reaction was heated to gentle reflux under nitrogen with stirring for 4 hours. The volume of the solution was reduced by half on a rotary e vaporator, resulting in the precipitation of dark, red - brown solids. The flask was cooled in an ice bath, and the dark solid was collected by suction filtration. The solid was washed with 1 mL of ice - cold 3 M HCl, then 3 1 mL portions of cold acetone fol lowed by 3 1 mL diethyl ether. A black - purple crystalline solid was collected with mass of 230 mg (0.36 mmol) after drying under high vacuum at 50 ° C for 12 hours. (79% yield). 1 H NMR (500 MHz, DMSO - d 6 ) Hz, 1H), 7.13 7.00 (m, 5H), 6.74 (d, J = 2.9 Hz, 1H), 6.55 (t, J = 6.6 Hz, 2H), 6.42 (dd, J = 6.8, 2.8 Hz, 1H), 5.69 (d, J = 6.9 Hz, 1H), 5.39 (dd, J = 6.9, 2.8 Hz, 1H), 2.56 (s, 6H), 2.19 (s, 6H). 3.3.2 . ( 2,2´:6´,2´´ - Terpyridyl)( 4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl)ruthenium(II) ammine dihexafluoro phosphate, [Ru(trpy)(dmabpy)NH 3 ](PF 6 ) 2 , (2a). [Ru(tpy)(dmabpy)Cl]Cl was we ighed into a 25 mL heavy walled reaction flask equipped with a stir bar (200 mg, 0.31 mmol, 1 equiv.). Concentrated aqueous ammonium hydroxide was added as 51 solvent (10 mL, excess). The reaction was closed with a Teflon screw plug and heated as a closed sys tem for 2 h at 90 ° C. After heating was complete, the reaction and stir bar were poured into a 125 mL beaker open to air in a fume hood. Excess ammonium hexafluorophosphate (100 mg, 0.6 mmol, 2 equiv.) was added as a solid into the beaker with stirring. The mixture was stirr ed for an hour until a thick slurry of precipitated brown solids resulted. The solid was collected by filtration and was washed with small portions of deionized water until the initial brown filtrate ran a lighter tan color. The mass of the resultant brown solid was 239 mg (0.27 mmol) after vacuum drying, 88 % isolated yield. 1 H - NMR (500 MHz; CD 3 CN): 2 (d, J = 6.7 Hz, 1H), 8.49 (d, J = 8.1 Hz, 2H), 8.38 (d, J = 8.0 Hz, 2H), 8.09 (t, J = 8.1 Hz, 1H), 7.94 (td, J = 7.8, 1.3 Hz, 2H), 7.88 (d, J = 5.5 Hz, 2H), 7.69 (d, J = 2.8 Hz, 1H), 7.41 - 7.38 (m, 3H), 7.17 (dd, J = 6.7, 2.8 Hz, 1H), 6.48 (d, J = 6.8 Hz, 1H), 6.16 (dd, J = 6.9, 2.8 Hz, 1H), 3.35 (s, 6H), 2.99 (s, 6H), 1.6 5 (s, 3H). 13 C NMR (126 MHz; CD 3 CN): .7, 149.7, 148.6, 137.0, 133.2, 127.4, 123.4, 122.6, 109.1, 108.6, 105.8, 105.2, 39.2, 38.9. 19 F NMR (470 MHz; CD 3 CN): - 72.8 (d, JF - P = 706 Hz). 31 P NMR (202 MHz; CD 3 CN): - 144.6 (septet, JP - F = 706 Hz). Anal. Calcd for C 29 H 32 F 12 N 8 P 2 Ru: C, 39.42 H, 3.6 5; N, 12.68. Found: C, 39.37; H, 3.66; N, 12.52. 8 3.4 . Catalytic Activity of 2a in THF 3.4.1 . Assessment of the Catalytic Current by Cyclic Voltammetry In dry THF, the E 1/2 of the Ru(III)/Ru(II) couple for [Ru(trpy)( dma bpy)NH 3 ](PF 6 ) 2 ( 2a ) appears at 0.095 V vs. Fc +/0 (See Appendix, Figure A3. 2 .1 ), 305 mV more negative than the E 1/2 of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 ( 1 a ) and 155 mV more negative than the onset of direct NH 3 oxidation at the working electrode. An enhanced current is observed for solutions containing 2a and NH 3 (sat d) in THF, with an onset potential of approximately - 0.025 V vs. Fc +/ 0 ( Figure 3.4 . 1 ). 52 Similar to what was observed for 1a , a negative shift in the onset potential as well as an increase in the normalized currents for slow scan rates, agree with a complicate catalytic mechanism which enhances with longer experiment time scales at low scan rates (See Appendix, Figure A3.2. 2 ). Figure 3.4 . 1 Cyclic voltammograms for THF solutions of 2.5 × 10 - 3 M 2 a (black), 2.5 × 10 - 3 M 2 a and 0.34 M NH 3 (red), uncatalyzed NH 3 oxidation in THF, 0.34 M (green) and the electrolyte background (gray). Scan rate 0.1 Vs - 1 . This figure shows that upon the presence of the catalyst 2a , a catalytic current is appeared at lower overpotentials relative to direct oxidation of ammonia in THF. The catalytic current has a linear dependence on the concentration of [Ru(trpy)( dma bpy)NH 3 ](PF 6 ) 2 catalyst (See Appendix , Fig. A3. 2 . 3 ). To assess the dependence of the catalytic current on the concentration of ammonia, THF solutions of 2.5 10 - 3 M 2a in NH 3 were prepared while maintaining the [NH 3 ] larger than 3 - fold excess to assure that pure catalytic conditions are held. A plot of the catalytic currents at 0.2 V versus concentration of ammonia ( Figure 3.4 . 2 ) shows a nonlinear trend. 53 Figure 3.4 . 2 Dependence of catalytic current on the concentration of ammonia when [ 2 a ] = 2.5 × 10 - 3 M, [NH 3 ] = (a) 0 M, (b) 0.008 M, (c) 0.04 M, (d) 0.07 M, (e) 0.17 M and (f) 0.34 M. Inset : Peak current s (at 0.90 V for catalytic currents) versus concentration of ammonia. The first data point (gray) is the anodic peak current for the catalyst in the absence of ammo nia and is shown for comparison. Scan rate 100 mV s - 1 . 3.4.2 . Quantification of the Products of the Electrolysis Using Gas Chromatography Controlled potential electrolysis of NH 3 saturated THF solutions were performed under argon in the presence of [Ru(trpy)( dma bpy)NH 3 ](PF 6 ) 2 ( 2a ) (2.7 × 10 - 3 M, 6.8 × 10 - 6 mol). A potential of 0.15 V vs. Fc +/0 was applied to the working electrode that was immersed in the solution for a total time of 180 min , during which 9.1 C of charge (9.4 × 10 - 5 moles of electrons) was passed ( Figure 3.4 . 3 ). The 1 H NMR of the solution, after the electrolysis only contains [Ru(trpy)( dma bpy)NH 3 ](PF 6 ) 2 ( 2a ) (See Appendix, A3. 5.1 ) and when the concentration of the catalyst 2a was measured before, and after the electrolysis by UV - Vis spectrophotometry, a modest decrease of 6.7% in [ 2a ] was calculated (See Appendix , A3. 5 .2). So, the gradual decrease in the current with time could be attributed to the consumption of the substrate , NH 3 , as oppose d to the catalyst degradation. 54 Figure 3.4 . 3 Current pass ing during three steps of controlled potential electrolysis (each 3600 s) while the solution was being stirred under argon. [ 2a ] = 2.7 × 10 - 3 M, [NH 3 ] = 0.34 M, 0.1 M [NH 4 ](PF 6 ) in dry THF. Numbers show the amount of charge (q, in coulombs) passed in each ste p , calculated based on the integration of the area under each curve (q = i × t). Comparison of the cyclic voltammograms of the solution obtained before the electrolysis and after each 30 min interval also shows a decrease in current , as well as slight reappearance of a cathodic return current at higher scan rates (See Appendix , A3. 5 .3 ), suggesting that upon consumption of the NH 3 substrate the CVs do not further depict the characteristics of a pure kinetic regime. Consistently, when N H 3 is bubbled in the solution after 180 min of electrolysis, the catalytic current is restored (See Appendix, Figure A3. 5 .3.2). For the electrolysis experiment shown in Figure 3.4 . 4 , the headspace gas was analyzed at 60 - minute intervals by gas chromatography. 55 Figure 3.4 . 4 Gas chromatograms obtained after injection of 100 L of the electrolysis headspace before applying potential and after 60, 120 and 180 min of electrolysis. The peaks associated with N 2 and H 2 are growing in after each step. The residual O 2 in the headspace sample is attributed to the leaks. For full chrom atogram See Appendix, Figure A3.4.4.1. Quantification of the products in the gaseous phase show s that after 180 min of electrolysis, 1.4 × 10 - 5 moles of N 2 and 3.9 × 10 - 5 moles of H 2 were evolved at 86% and 77.6% faradaic efficiencies for the respective anodic and cathodic reactions with the H 2 :N 2 ratio equal to 2.74:1. Similar faradaic efficiencies and H 2 :N 2 ratio results were obtained for a few bulk electrolysis replicate experiment s that were carried out with different [ 2a ] and duration of electrolysis ( See Appendix , A3. 5 .4 ). Control electrolysis experiment in which electro - oxidation of NH 3 in THF was conducted at 0.15 V versus Fc +/0 in the absence of the catalyst did not generate N 2 and H 2 in the gaseous headspace. 56 3.5 . Homogeneous vs. Heterogeneous Catalysis T o verify the homogenous nature of the catalysis, the surface of the glassy carbon electrode was examined with XPS. The Ru3d peak appears at 280 eV, overlapping with the C1s peak at 284.8 eV. Because the electrode is glassy carbon, the very large C1s peak obscures unambiguous identification of the Ru3d peak. However, when the region between 270 to 300 eV is scanned separately, the characteristic peak for Ru 3 d is not ev ident ( Figure 3.5 . 1 , bottom ). 57 Figure 3.5 . 1 XPS spectra of the glassy carbon electrode after the BE. Top : full spectrum . The source of silicon and oxygen is elect rode contamination with silicon grease during the removal of the electrode from the electrolysis cell . The grease was heavily applied around the joints to prevent leaks. Bottom : the scan around 285 eV where the characteristic peak for ruthenium is expected to appear if depositions have had taken place . 58 Additionally, a rinse test was performed by removing the glassy carbon electrode from the electrolysis solution, rinsing with dry THF and then inserting it into a cell containing a fresh NH 3 /THF/ 0.1 M [NH 4 ](PF 6 ) solution. No enhanced catalytic current was passed at an applied potential of 0.15 V versus Fc +/0 , and the headspace sampling of the electrolysis cell at the end of 3 hours of electrolysis showed no formation of N 2 and H 2 ( Figure 3.5 . 2 ). These results confirm that the catalytic activity is not due to electrodeposition of ruthenium species on to the surface of the glassy carbon electrode. Figure 3.5 . 2 Rinse test results. Left: The CV recorded for a solution of NH 3 (sat d)/THF in which the rinsed glassy electrode was immersed. Scan rate 0.1 Vs - 1 . Right: Current vs. times for three electrolysis steps of electrolysis. The inset is the gas chromatogram of the cell headspace injected to GC before the electrolysis and after the third hour of electrolysis. No H 2 is generated and the N 2 :O 2 ratio matches with leaks from the air. Examination of the glassy carbon electrode by SEM after the rinse test reveals a ruthenium - deposition free surface and the EDS measurements of the selected areas do not show any ruthenium deposition s on the surface ( See Appendix, A3. 5 .5). Such depositions, when existent, are very easily noticeable. For instance, in an electrolysis experiment which failed due to the presence of water residuals in the NH 3 /THF solution, microscopic examination of the surface showed that large 59 grains of a ruthenium - containing material have developed on to the surface. This layer is visible with naked eye in the form of a thin orange layer on the electrode that does not get rinsed with THF ( Figure 3.5 . 3 ). To prevent this issue, extensive care should be taken regarding the dryness of the reagents and equipment. Figure 3.5 . 3 EDS analysis of the selected area (red circle in the inset) on the glassy carbon electrode at the end of a failed electrolysis experiment. The insets are the SEM images of the electrode with two different magnifications . These results show that a rutheniu m containing material has deposited on the surface, possibly acting as an active catalytic surface. 3.6 . Isotopic Labeling Experiments In order to understand the origin of the evoleved N 2 in bulk electrolysis experiments, nitrogen - labeling experiments we re performed. In th e se experiments, electrolysis of 15 NH 3 was conducted in the presence of 15 N - labeled catalyst under the same conditions applied in section 3.4 .1 . 60 3.6.1 . Synthesis of [Ru(trpy)( dma bpy) 15 NH 3 ](PF 6 ) 2 ( 15 N - 2a ) (2,2´:6´,2 - t erpyridyl)(4,4´ - bis( N , N - dimethylamino) - 2,2´ - bipyridyl) chloro ruthenium(II) chloride was weighed into a 25 mL heavy walled reaction flask (160 mg, 0.25 mmol, 1 equiv . ) containing 10 mL deionized water. 15 NH 4 Cl (670 mg, 12.5 mmol, 50 equiv . ) was weighed and added to the solution as a solid. Pellets of KOH (690 mg, 12.5 mmol, 50 equiv . ) were transferred into the solution and the flask was capped and sealed immediately. The reaction was heated for 2 h at 100 ° C while stirring. After 2 h, the flask was cooled down to room temperature and the solution was transferred to a beaker. Ammonium hexafluorophosphate (80 mg, 0.5 mmol, 2 equiv . ) was added as a solid and the precipitate was collected on a glass filter. The solid was washed with portions of deionized water and dried overnight in a vacuum oven at 60 ° C. The mass of product was 182 mg (0.20 mmol), 83 % isolated yield. 1 H NMR (500 MHz, CD 3 CN ) : 8.69 (d, J = 6.7 Hz, 1H), 8.45 (d, J = 8.1 Hz, 2H), 8.35 (d, J = 8.1 Hz, 2H), 8.06 (t, J = 8.1 Hz, 1H), 7.91 ( td, J =7.8, 1.4 Hz, 2H), 7.85 (d, J = 5.4 Hz, 2H), 7.65 (d, J = 2.8 Hz, 1H), 7.37 - 7.35 (m, 3H), 7.14 (dd, J = 6.7, 2.8 Hz, 1H), 6.45 (d, J = 6.8 Hz, 1H), 6.12 (dd, J = 6.9, 2.8 Hz, 1H), 3.31 (s, 6H), 2.96 (s, 6H), 1.61 (d, J = 68.3 Hz, 3H). 13 C NMR (126 MHz ; CD 3 CN): 154.9, 153.8, 152.7, 149.7, 148.6, 137.0, 133.2, 127.4, 123.4, 122.6, 109.1, 108.6, 105.8, 105.2, 39.2, 38.9. 3.6.2 . Electrolysis of 15 NH 3 using 2a and 15 N - 2a Controlled potential electrolysis of a 15 NH 3 saturated THF solution containing 15 N - 2a generated 15 N 15 N (m/z = 30) judged by GC - MS analysis ( See Appendix, Figure A3. 5 .6. 2 ). As described by reaction 2 below (polypyridyl ligands are not shown) , 15 N 15 N (m/z = 30) is expected to be the result of the N - N bond formation between the coordinated ammonia nitrogen and the free ammonia in the solution: 61 (2) Bulk electrolysis of a 15 NH 3 saturated THF solution with non - labeled 2a generated N 2 that was initially 15 N 14 N (m/z = 29) and was eventually dominated by 15 N 1 5 N (m/z = 30) as the 14 NH 3 ligand in the complex was displaced by 15 NH 3 in the solution over the course of the electrolysis ( See Appendix, Fig. A3. 5 . 6 .3 ). Reactions 3 - 5 describe these pathways: This finding is of a great importance because it precludes the possibility of coupling pathways and provides evidence for a single ruthenium site catalytic mechanism. Inc orporation of the NH 3 substrate in the N - N bond formation step also suggests that this step is governed by a nucleophilic - electrophilic mechanism, guiding us towards the essentiality of an intermediate bearing a very electrophilic nitrogen. Insights to the possible i ntermediates are discussed in the following chapter s . 3.7 . Conclusions The catalytic activity of two complexes 1a and 2a were confirmed by cyclic voltammetry. More detailed studies on the electro - oxidation of NH 3 in THF solutions in the presence of 2a shows that N 2 and H 2 are generated while the substrate NH 3 is being consumed throughout the electrolysis. Isotope labeling e xperiments provide evidence that the generated nitrogen is the product of the anodic reaction and the possibility of a heterogeneous catalytic layer was ruled out via control (3) (4) (5) 62 experiments. In the next two chapters, our findings towards the mechanistic detai ls of the catalysis are discussed. 63 APPENDIX 64 A 3.1 . 1 H NMR Spectra Figure A 0 . 1 .1 . 1 H NMR spectrum of [Ru(trpy)(bpy)Cl]Cl in DMSO - d 6 . Top: full spectrum, Bottom: magnified aromatic region. (ppm) (ppm) 65 Figure A 3 .1 .2. 1 H NMR spectrum of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 in NM - d 3 . Top: full spectrum, Bottom: blow - up of the aromatic region. Starred peaks are solvent related. (ppm) (ppm) 66 Figure A 3 .1. 3 1 H NMR spectrum of [Ru(trpy)(dmabpy)Cl]Cl in DMSO - d 6 . Top: full spectrum, Bottom: blow - up of the aromatic region. (ppm) (ppm) 67 Figure A 3 .1. 4 1 H NMR spectrum of [Ru(trpy)( dma bpy)NH 3 ](PF 6 ) 2 in acetonitrile - d 3 . Top: full spectrum, Bottom: blow - up of the aromatic region. Starred peaks are solvent related. (ppm) 68 A 3.2 . Cyclic Voltammetry Studies of 2a in THF. Figure A3. 2 .1 Top : Scan rate dependence of the current in 5.2 × 10 - 3 M 2a in THF for (a) 50, (b) 100, (c) 250, (d) 500 and (e) 1000 mV s - 1 . Bottom : Plots of anodic and cathodic peak currents obtained from CVs on the top versus square root of the scan rate. From the slope of the anodic branch a diffusion coefficient of D ox = 4.26 10 - 6 cm 2 s - 1 was calculated for 2a . 69 A 3.3 . Scan Rate Dependence of the Catalytic Current Figure A3. 3 . 1 Top : CVs collected for 2.5 × 10 - 3 M 2a in saturated ammonia solutions in THF, with different scan rates, Bottom : Normalized catalytic currents in the solution of 2.5 × 10 - 3 M 2a and NH 3 (0.34 M) . Both magnitude and the onset of the normalized currents are improved at lower scan rates. 70 A 3.4 . Concentration Dependence of the Catalytic Current Figure A3. 4.1 Top : Catalytic current versus concentration of 2a . In all cases [NH 3 ] = 0.34 M. Bottom : Catalytic peak currents versus concentration of the catalyst 2a . (a) 7.8 × 10 - 5 M catalyst, (b) 3.1 × 10 - 4 M catalyst, (c) 1.2 × 10 - 3 M catalyst and (d) 2.5 × 10 - 3 M catalyst. Scan rate 0.1 V s - 1 . The correlation between the catalytic current and the con centration of the catalyst is almost linear. 71 A 3.5 . Catalytic Electrolysis of NH 3 in THF 1 H NMR Spectrum of the Solution After the Electrolysis After the electrolysis, the solvent was pumped out for 3 hours, and the residual solids were dissolved in acetonitrile - d 3 . Figure A3.5.1.1 Room temperature 1 H NMR spectrum of the electrolysis sell contents in acetonitrile - d 3 . The resonance for the coordinated ammine is labeled with a black circle. The singlet at 2.46 ppm (blue circle) is unidentified. Figure A3.5.1.2 1 HNMR spectra of (A) [NH 4 ](PF 6 ) and (B) an authentic sample of 2a in acetonitrile - d 3 . This figure could be used to identify the species observed in the spectrum in Figure A3.5.1.1. The starred peak is due to solvent residuals. (ppm) (ppm) 72 Spectrophotochemical Determination of [2a] Before and A fter Electrolysis n was drawn before and after the using a glass syringe and was transferred to a 5 mL volumetric flask . The volume was made up to the volume by THF . The electronic spectra of these solutions were collected against a THF reference, and the absorbance values at 490 nm were converted to concentration using the absorption coefficient of 2a in THF ( = 9945 M - 1 cm - 1 ). The calculated concentration in the solution after BE is with the consideration of the change in the solution volum e during the electrolysis. Calculations are summarized in Table A3.5 .2. 1. Table A3. 5 .2. 1 The spectrophotochemical measurement of 2a concentration before and after BE. sample Abs. @496 nm [ 2a ] a in soln.2 mol 2a in soln. 2 mol 2a in BE solution % loss b (t=0) 0.149 1.49 10 5 M 7.49×10 8 6.24 10 - 6 b =6.7% a (t=180 min) 0.151 1.51 10 - 5 M 7.59×10 8 5.82 10 - 6 c a using A= bC, where b=1 cm and = 9945 M - 1 cm - 1 at 496 nm. b total volume of the solution at t=0 is 2.5 mL. c total volume of the solution at t=180 min is 2.3 mL. 73 Figure A3. 5.2 .1 Top : Electronic absorption spectrum of 2a in THF. Bottom : Absorption spectra of the solution used in the electrolysis of NH 3 (a) after the electrolysis (t=180 min) and (b) before the electrolysis (t=0). 74 CVs of the Solution Over the Course of Electrolysis Fig ur e A3.5.3.1 Cyclic voltammograms of the electrolysis solution obtained under static conditions (without stirring). (a) before electrolysis, (b) after 60 min, (c) after 120 min, (d) after 180 min of electrolysis. [ 2a ] = 2.7 × 10 - 3 M, Scan rate 0.1 V s - 1 . A= 2.0 cm 2 . The change in the CVs are related to the consumption of ammonia with time. 75 Figure A3. 5 . 3 . 2 Top : CVs of the solution at the end of a bulk electrolysis experiment (6.1 C charge passed, 1 × 10 - 6 mol 2a ) with different scan rates. Solid black: 1 00 mV s - 1 , dotted blue: 500 mV s - 1 and dashed red: 1000 mV s - 1 . Bottom : Effect of NH 3 addition on the anodic current in the solution after the electrolysis (scan rate 100 mV s - 1 ). (a) CV of the solution after BE is complete, (b) current recovery after NH 3 is bubbled through the solution in (a), and (c) CV of the solution before electrolysis. A= 0.07 cm 2 . 76 Quantification of the Gaseous Products of the Electrolysis in the Headspace Figure A3. 5 .4.1 Full gas chromatograms from 100 microliter injections of cell headspace gases during electrolysis. (i) Before electrolysis (t=0), and after (blue) 60 min, (green) 120 min, and (orange) 180 min of electrolysis. The two peaks between 10 and 15 min are THF and NH 3 . The baselines are superimposed on to p of each other for the time 17 min < t < 30 min window. Prior to the experiment, the ratio of N 2 :O 2 of air to the instrument and finding the peak integral ratios. For the BE experiments discussed in 3.4.2 , the N 2 :O 2 ratio was found to be 2.98. The peak integrals for N 2 in the cell headspace samples were corrected for leaks and N 2 residuals in argon gas according to equation A 1 , ( A 1) Where the N 2 (sample) is the peak area of N 2 measured in the injection, N 2 (blank) is the nitrogen gas present in the headspace before the electrolysis begins (=0.88), O 2 (sample) is the oxygen observed in the sample injection which is coming in to the cell through leakage over time, and the 77 is the 2.98 that was obtained experimentally. The t erm in the bracket in equation A1 is to eliminate the amount s of N 2 that is brought in to the cell via air leaks. No correction was applied to H 2 signals since no H 2 was present in the air and blank samples. Table A3. 5 .4.1. Actual signals obtained for e ach gas and the corrected signals for N 2 (in the gray - shaded column). Duration of Electrolysis (min) Peak Integrals H 2 O 2 N 2 Corrected N 2 t=0 0 0 0.88 0 t=60 135.38 1.06 12.05 8.01 t=120 286.51 1.18 19.52 15.12 t=180 550.28 1.72 34.95 28.95 The H 2 and corrected N 2 using the calibration lines that were obtained for each gas (Chapter 2). The number of moles of each gas was then back - calculated for the 6.5 mL headspace of the BE ce ll that was employed for the experiment. Faradaic efficiencies (FE) were calculated by dividing the theoretical number of moles of each gas (based on the transferred charge) by their corresponding experimental values. Results are shown in Table A3. 5 .4.2. Table A3 .5 .4.2. Results of the Electrolysis Experiment. a Duration of electrolysis Charge passed (C) Measured in Headspace H 2 : N 2 % FE H 2 N 2 H 2 N 2 60 min 3.7 8.9 3.8 2.34 46.8 59.3 120 min 6.8 19.0 7.0 2.71 54.2 59.8 180 min 9.1 36.5 13.4 2.72 77.6 86.0 a 6.8 10 - 6 mol 2a 78 Two other BE experiments were also performed, and the results were analyzed in the same way: Table A3. 5 .4.3. Results of the Electrolysis Experiments #2 and #3. Experiment Duration of electrolysis Charge passed (C) Measured in Headspace H 2 : N 2 % FE H 2 N 2 H 2 N 2 #2 a 120 min 7.2 29.5 11.2 2.63 79.0 90.3 #3 b 240 min 8.0 32.4 13.8 2.67 78.2 87.6 a 5.9 10 - 6 mol 2a , b 5.6 10 - 6 mol 2a Microscopic Examination of the Surface of the Glassy Carbon Electrode The surface of the working electrode was examined with XPS and EDS for possible precipitation of catalytic layers. Figure A3. 5 .5.1 SEM images of the glassy carbon electrode before and after being used in the electrolysis. 79 Figure A3.5.5.2 A more magnified SEM image of the working electrode after the electrolysis. The pores in the electrode were present before the electrolysis and are remains of the unknown history of various uses. The green arrows show that the pores are not only on the sur face but are visible underneath the surface. These buried pores may be exposed in the future upon polishing. The areas 1 and 2 are selected for EDS measurements. 80 Figure A3.5.5.3 EDS analysis of selected areas 1 (top) and 2 (bottom). Al, P, and F ar e probably from polishing paste and PF 6 salts. The source of Si and oxygen is probably the silicon grease. 81 Data of the Double and Single Labeling Experiments Figure A3.5.6.1 1 H NMR spectra for unlabeled 2a (top , blue ) and 15 N - 2a (bottom , black ) in acetonitrile - d 3 . The starred resonances are associated with the coordinated NH 3 ligand. Figure A3. 5 .6.2 Mass spectrometry data for two injections of headspace mixture of the bulk electrolysis experiments of 15 NH 3 solutions before and after passing 3.58 C. 82 Figure A3. 5 .6.3 MS spectra for the injections of the headspace mixture of an electrolysis experiment w hen 15 NH 3 was electrolyzed in the presence of 4.5 10 - 5 M 2a . Top: The headspace of the blank cell. The peak at m/z=29 (green) shows an abundance of 0.8% relative to N 2 (m/z=28), the natural abundance of m/z=29 in the air. Middle: The percent of m/z=29 has increased to 1.2% after passing 1.34 C. m/z=30 (red) is also growing. Bottom: After 6.47 C charge is passed the generated N 2 is dominantly m/z=30 83 REFERENCE S 84 REFERENCES (1) Murphy, W. R.; Takeuchi, K.; Barley, M. H.; Meyer, T. J. Inorg. Chem. 1986 , 25 (9), 1041 1053. (2) Wasylenko, D. J.; Ganesamoorthy, C.; Henderson, M. A.; Koivisto, B. D.; Osthoff, H. D.; Berlinguette, C. P. J. Am. Chem. Soc. 2010 , 132 (45), 16094 16106. (3) Takeuchi, K.; Thompson, M.; Pipes, D. W.; Meyer, T. J. Inorg. Chem. 1984 , 23 (13), 1845 1851. (4) Ridd, M. J.; Keene, F. R. J. Am. Chem. Soc. 1981 , 103 (19), 5733 5740. (5) Bard , A . J .; Faulkner , L . R . Electrochemical Methods: Fundamentals and Applications, 2nd Edition , Wiley Global Education , 2000 . (6) Yagi, M.; Tajima, S.; Komi, M.; Yamazaki, H. Dalt. Trans. 2011 , 40 , 3802 3804. (7) Hetterscheid, D. G. H.; Reek, J. N. H. Angew. Chem. Int. Ed 2012 , 51 , 9740 9747. (8) Habibzadeh, F.; Miller, S. L.; Hamann, T. W.; Smith, M. R. Proc. Natl. Acad. Sci. 2019 , 3 (l), 2 6. 85 THE RU(III) INTERMED IATE 86 With the redox potential of [Ru(trpy)( dma bpy)NH 3 ] +2 ( 2a ) being more negative than the direct oxidation of ammonia at the glassy carbon electrode in THF, we can c onc lude that the oxidation of Ru(II) to Ru(III) occurs under catalytic conditions to give the first intermediate of the catalytic cycle, [Ru(trpy)( dma bpy)NH 3 ] 3+ ( 2b ) . In the presence of a base, such as NH 3 , [Ru(trpy)( dma bpy)NH 3 ] 3+ is unstable and reacts further to give other intermediates, but it eventually goes back to [Ru(trpy)( dma bpy)NH 3 ] +2 at the end of the cycle. However, when the electrolysis was performed in the absence of a base, [Ru(trpy)( dma bpy)NH 3 ] 3+ was produced upon one - electron oxidation of [Ru(trpy)( dma bpy)NH 3 ] +2 and , was isolated as a green solid in dichloromethane (DCM), with Ag + ion serving as the sacrificial oxidant in a separated compartment. Alternatively, [Ru(trpy)( dma bpy)NH 3 ] + 3 was synthesized from [Ru(trpy)( dma bpy)NH 3 ] +2 using a chemical oxidant in DCM and was thoroughly characterized using a variety of techniques. 4.1 . Synthesis of [Ru(trpy)( dma bpy)NH 3 ] (PF 6 ) 3 , (2b) 4.1.1 . Electrochemical Synthesis Exhaustive oxidation of solutions containing [Ru(trpy)( dma bpy)NH 3 ] +2 in DCM at the applied constant potential of 0.35 V versus Fc +/0 afforded a green precipitate adhered to the surface of the glassy carbon electrode. To perform this experiment, a 1.2 10 - 3 M solution of [Ru(trpy)( dma bpy)NH 3 ] +2 in DCM was prepared with 0.1 M [Bu 4 N](PF 6 ) serving as the supporting electrolyte. The counter electrode was prepared by inserting a freshly polished silver coil in a solution of 0.5 M AgPF 6 in acetonitrile , containing 0.1 M [Bu 4 N](PF 6 ) . The CE sliver coil and solution were separated from the electrolysis sol ution using a fritted glass membrane ( Figure 87 4. 1 . 1 ). A glassy carbon plate with a large surface area was used as the working electrode, and the solution w as stirred during the electrolysis. Figure 4. 1 . 1 The diagram of the three - electrode electrolysis cell used for chemical oxidation of 2 a to 2 b . E very 10 min the glassy carbon electrode was removed from the cell, rinsed with fresh DCM and the solid precipitate was scraped off the electrode into a centrifuge tube. At the end of the experiment, 5 mL fresh DCM was added to the collected solid in the centrifuge tub e, the tube was placed in a centrifuge machine and was spun at a speed of 60 rpm for 20 min. Then the solvent was slowly decanted, and the rinsing step was repeated three times. The washed solid was then dried under vacuum overnight. Since these experiment s were never performed to completion, i.e., to the point where all [Ru(trpy)( dma bpy)NH 3 ] +2 were oxidized, a yield was not calculated. 4.1.2 . Chemical Synthesis Addition of 1 equiv. of the chemical oxidant tris( para - bromophenyl) ammonium cation radical, [N( p Br - ph) 3 ]PF 6 , to solutions of [Ru(trpy)( dma bpy)NH 3 ] +2 in DCM yielded a green precipitate after 30 min of stirring under nitrogen . The solid was collected on a Fine glass filter and was rinsed several times with fresh DCM. For a starting solutio n of 1.13 10 - 2 M 88 [Ru(trpy)( dma bpy)NH 3 ] (PF 6 ) 2 (100 mg, 1.13 10 - 4 mol 2a in 10 mL DCM) a yield of 94% (110 mg, 1.06 10 - 4 mol 2b ) was obtained. 4.2 . Characterization of 2b 4.2.1 . Proton NMR Complex [Ru(trpy)( dma bpy)NH 3 ] + 3 as a hexafluorophosphate salt is not soluble in dry THF or DCM and reacts (judged by the color change from green to orange upon mixing ) with electron donating solvents such as methanol or dimethyl sulfoxide (DMSO). On the other hand, it fully dissolves in solvents with slight acidic propertie s such as dry acetonitrile and nitromethane (NM). However, 1 H NMR experiments show that while the solutions of [Ru(trpy)( dma bpy)NH 3 ] + 3 in dry acetonitrile remain green, the coordinated NH 3 is replaced by the solvent to some extent (See Appendix , Figure A4. 1.1). Addition of a small concentration of nitric acid to the acetonitrile prevents ligand displacement and affords clear 1 H NMR spectra of [Ru(trpy)( dma bpy)NH 3 ] + 3 that reflect the paramagnetic nature of the complex (See Appendix, Figure A4.1.2 ). 1 H NMR (500 MHz, 0.1% HNO 3 / CD 3 CN): (s, 2 H), 12.99 (s, 1 H), 9.34 (s, 2 H), 7.99 (s, 2 H), 4.67 (s, 2 H), - 18.20 (s, 1 H), - 20.53 (s, 1 H), - 23.14 (s, 1 H), - 23.31 (s, 1 H), - 35.59 (s, 1 H) , - 43.30 (s, 1 H). 4.2.2 . Elemental Analysis The results of CHN analysis for both electrochemically and chemically made [Ru(trpy)( dma bpy)NH 3 ] + 3 samples are identical a nd are as follows: Anal. Calcd. for C 29 H 32 N 8 P 3 F 18 Ru: C, 33.81; H, 3.13; N, 10.88. Found: C, 33.65; H, 3.10; N, 10.98. 89 4.2.3 . Magnetic Susceptibility The Evans method was used to calculate the effective magnetic moment, eff , for [Ru(trpy)( dma bpy)NH 3 ] + 3 using ferrocene as the standard. 1 A standard solution of saturated ferrocene in nitromethane - d 3 was prepared and transferred to a glass capillary. The capillary tube was sealed and inserted into an NMR tube containing 0.40 ml nitromethane - d 3 . A sample solution was prepared by dissolving 3.0 mg of [Ru(trpy)( dma bpy)NH 3 ] (PF 6 ) 3 2b ) in 0.25 m L nitromethane - d 3 gas - tight glass syringe and the NMR spectrum was acquired after the addition. The shift in the ferrocene 1 H resonance upon addition of 2b was measured (Hz) and used to calculate the molar magnetic susceptibility ( M ) using equation 1. (1) 0 is the spectrometer frequency (500 MHz), and C is the concentration of [Ru(trpy)( dma bpy)NH 3 ] + 3 (mol L - 1 ). The experimental value for 10 - 6 M solution of the sample in the NMR tube ( Figure 4.2 . 1 ). 90 Figure 4.2 . 1 The 1 H NMR resonance shift observed for the standard ferrocene after addition of paramagnetic 2 b . The diamagnetic contribution to the magnetic susceptibility, D , was calculated using tabulated below. 2 The appropriate substitutions in eq. 1 yield (2) (3) (4) (5) (6) (7) ( 8 ) 91 The effective magnetic momentum, eff , was therefore determined directly from M according to equation 9 with T = 21 ° C for the sample temperature. (9) The experimental value for eff is slightly higher than the spin - B for one unpaired electron and is typical for lo w spin d 5 complexes. 3 4.2.4 . Electrochemistry CVs of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM show a reversible redox process at E 1/2 = 0.085 V versus F c +/ 0 , identical to the redox potential of [Ru(trpy)( dma bpy)NH 3 ] + 2 in NM ( Figure 4.2 . 2 , top ). NM acts as an excellent solvent for electrochemical studies of [Ru(trpy)( dma bpy)NH 3 ] + 3 since it provides a very wide working oxidation potential window. W hen the scanning potential in th e cyclic voltammetry experiments of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM are extended to almost 1.5 V versus Fc +/ 0 , another redox wave is evident ( Figure 4.2 . 2 , bottom , wave II). Hydrodynamic linear scan voltammograms (See Appendix , Figure A4.3.1 ) of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM reveal that the first redox process corresponds to the reversible reduction of [Ru(trpy)( dma bpy)NH 3 ] + 3 to [Ru(trpy)( dma bpy)NH 3 ] +2 , while the second process is an oxidative process that was assigned to quasi - reversible one - electron oxidation of Ru(III) to Ru(IV) which was not observable in solvents with a smaller potential window such as THF or dichloromethane. 92 Figure 4.2 . 2 Top : CV of 4.5 10 - 3 M 2 b (green) compared to the CV of M 4.12 10 - 3 M 2 a (orange) in NM . Both complexes appear at the same E 1/2 , Bottom : Extended potential window CVs of 2.10 10 - 3 2b in NM shown for three scan rates. Here a second oxidation process is seen at 0.95 V vs. Fc +/0 (labeled as II). Bulk electro - reduction of [Ru(trpy)(dmabpy)NH 3 ] +3 back to [Ru(trpy)(dmabpy)NH 3 ] +2 was also performed in THF, using a similar cell diagram as shown in Figure 4. 1 . 1 , to further investigate the 93 relation between these two complexes. In this case, the counter electrode was modified to a platinum wire immersed in a solution of decamethyl ferrocene ( Me 10 Fc ) in acetonitrile ( E 1/2 = - 0.51 V versus Fc +/0 , See Appendix, Figure A4.4.1), so that the oxidation of decamethyl ferrocene occur during the reduction of [Ru(trpy)(dmabpy)NH 3 ] +3 . The glassy carbon plate working electrode was replaced with a rotating disk electrode after 1000, 2000 and 5000 s of electrolysis while a slow flow of Ar was maintaine d to keep the solution from exposure to air, and hydrodynamic CVs were taken to track the changes in each complex after electrolysis . However, d ue to the low solubility of [Ru(trpy)(dmabpy)NH 3 ] +3 in THF, the hydrodynamic currents associated with [Ru(trpy)( dmabpy)NH 3 ] +3 do not reflect the actual concentration. 1.03 equivalent of electrons were transferred for the full conversion of [Ru(trpy)(dmabpy)NH 3 ] +3 to [Ru(trpy)(dmabpy)NH 3 ] +2 along with the color of the solution changing to orange. The E 1/2 of the fina l product is also the same as what is measured for [Ru(trpy)(dmabpy)NH 3 ] +2 in the THF solvent (See Appendix, Figures A4.4.2 and A4.4.3). 4.3 . Deprotonation of [Ru(trpy)( dma bpy)NH 3 ] + 3 Addition of NH 3 to solutions of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM ( or suspensions in case of THF) leads t o a rapid reaction which changes the color of the solution (or the suspension in case of THF) from green to orange. Therefore, [Ru(trpy)( dma bpy)NH 3 ] + 3 could not be detected from the solution under catalytic conditions where the concentration of NH 3 is very large because the presence of the base promotes further reaction s . As discussed in chapter 1, the acidity of the complexes increases with the oxidation state of the ruthenium metal center. The acidic [Ru(tr py)( dma bpy)NH 3 ] + 3 (compared to [Ru(trpy)( dma bpy)NH 3 ] +2 ) is expected to undergo deprotonation in the presence of a base of suitable strength. With [Ru(trpy)( dma bpy)NH 3 ] +2 having 94 a oxidation potential within the applied potential for the catalytic electrolys is experiments, it is speculated that the first step towards N 2 formation, is the one electron oxidation of [Ru(trpy)( dma bpy)NH 3 ] +2 to [Ru(trpy)( dma bpy)NH 3 ] + 3 . The the latter, 2b , is stable in the absence of a base and was isolated as discussed earlier. The reversibility of the [Ru(II/III)NH 3 ] 2+/3+ couple in THF suggests that the p K a of [Ru(trpy)( dma bpy)NH 3 ] + 3 is not low enough to be deprotonated by THF and changing the electrolyte from [Bu 4 N](PF 6 ) to more acidic [NH 4 ](PF 6 ) does not affect the E 1/2 . However, when a base such as NH 3 is added, [Ru(trpy)( dma bpy)NH 3 ] + 3 undergoes de protonation to give an amido complex, [Ru(trpy)( dma bpy)NH 2 ] +2 , 2c . (10) The acid - base equilibrium constant for reaction (10) depends on the strength and concentration of the proton acceptor and the nature of the solvent due to its effects on the K b of the base. Relatively a cidic [Ru(trpy)( dma bpy)NH 2 ] + 3 is unstable when NH 3 is present in the solution and reacts further to give other intermediates that were formed via nucleophilic attack of the free NH 3 ( vide infra ). However, with a suitable adjust ment of the type and concentration of the base, as well as the nature of the solvent it might be possible to observe [Ru(trpy)( dma bpy)NH 2 ] +2 . To do so, the sterically hindered 1,8 - diazabicyclo [5.4.0] undec - 7 - ene (DBU) was selected as the noncoordinating proton acceptor to prevent possible nucleophilic/electrophilic reactions since the bulky structure of DBU limits the approach of the base to the electron deficient intermediates which leads to the formation of other products . Nitromethane was chosen as the solvent due to complete solubility of [Ru(trpy)( dma bpy)NH 3 ] + 3 . Being an acidic solvent, NM may compet e with [Ru(trpy)( dma bpy)NH 3 ] + 3 in losing protons to DBU. I n solutions containing DBU and NM, the ( 2b ) ( 2c ) 95 amounts of deprotonated DBU is defined by the equilibrium constant of the reaction represented below: Scheme 4.1 The acid - base equilibrium between DBU (pK a = 11.9 (DMSO)) and NM (pK a = 17.2 (DMSO)). pK a values from Refs. 14 and 15. Based on the pK a values, this equilibrium is more towards the reactants. However, a slight yellow color is observed when DBU is added to NM , which might indicate the some reaction occurs between the two . Direct infusion of a diluted solution consist ing of [Ru(trpy)( dma bpy)NH 3 ] + 3 in anhydrous nitromethane mixed with sub - stoichiometric concentrations of DBU to an ESI + QTOF mass spectrometer ( ) shows the presence of four ruthenium - based species ( Figure 4.3 . 1 ). For this experiment stock solutions of 1.16 10 - 3 M [Ru(trpy)( dma bpy)NH 3 ] + 3 (6 mg in 5 mL NM) and 1.34 10 - 3 solutio n was diluted to a final volume of 5 mL to give a secondary stock solution of 1.07 10 - 3 M. NM and then the volume was made up by adding fresh NM. The final concentration s for [Ru(trpy)( dma bpy)NH 3 ] + 3 and DBU were 1.16 × 10 6 M and 1.072 × 10 6 M, respectively (DBU: 2b = 0.92:1). 100 of this solution was then diluted to 1 mL in a glass vial equipped with a septum cap, the vial was wrapped with Teflon tape and was taken to the mass spectrometer, and the content was directly injected to the instrument. Positive mode electrospray ionization (ESI + ) quadrupole time - of - flight (QTOF) mass spectrometer 96 Figure 4.3 . 1 Total ion spectrum obtained for a solution containing 1.16 × 10 7 M 2 b and 1.072 × 10 7 M DBU in NM. Peaks 1 and II, and VI and VII are related to doubly charges species, while peaks III and IV are triply charges species. Peak V is matching with the dmabpy ligand, which comes off in the mass analyzer. The dmabpy - less complexes are seen as in signals I and II. The mass to charge (m/z) values observed for ruthenium species that are expected to be existent in the solution and their assignments to observed signals III, IV, VI and VII are listed in Figure 4.3 . 2 . For other assignments see Appendix, A4.5. 97 III IV VI VII Figure 4.3 . 2 Assignments to four selected m/z values III, V, VI, and VII. The spectra in red are the experimentally obtained spectra and the inset shows the simulated spectra. 98 Direct infusion techniques, even when integrated with high - resolution mass analyzers such as QTOF, suffer from inevitable matrix effects since the samples are infused together without separation. This may result in reduced sensitivity and vague accuracy for analyte identification. 4 Nevertheless, direct infusion of control samples containing [Ru( trpy)( dma bpy)NH 3 ] +2 and [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM result in the formation of ion fragments IV and VI , respectively (See Appendix , A4.5). Thus, III and VII are species generated under experimental conditions, i.e., presence of DBU . These peaks are close to fragmentations of [Ru(trpy)( dma bpy)NH 2 ] +2 , the deprotonated form of [Ru(trpy)( dma bpy)NH 3 ] + 3 in the presence of a base , so they were cautiously assigned to [Ru(trpy)( dma bpy)NH 2 ] +2 . 4.4 . Regeneration of 2a from 2b One of the products de tected by mass spectrometry was [Ru(trpy)( dma bpy)NH 3 ] +2 . Regeneration of [Ru(trpy)( dma bpy)NH 3 ] +2 upon treatment of [Ru(trpy)( dma bpy)NH 3 ] + 3 with a base has a significant role in elucidating the reaction mechanism. Similar results have already been observed and reported for ruthenium catalysts used for water oxidation . According to Shaffer et al . , for two ruthenium complexes [Ru II (trpy)(bpm)(OH 2 )] 2+ and [Ru II (trpy)(bpz)(OH 2 )] 2+ - bipyrimidine, bpz= - bipyrazine), addition of one equivalent of an oxidizing agent yielded [Ru(III)OH] 2+ (polypyridyl ligands are omitted) as the dominant species even at low pH s . 16 They further discussed that while the Ru(II)OH 2 /Ru(III)OH 2 redox potential was relatively high due to the electron withdrawing effects of the bpm and bpz ligands, the oxidation of [Ru(III)OH] 2+ /[Ru(II)O] 2+ occurred at lower potentials, a phenomenon that thermodynamically enables a rapid disproportionation reaction to take place following the fo rmation of [Ru(III)OH] 2+ : (11) 99 Coia et al . have also reported similar chemistry for [Os(trpy)(bpy)NH 3 ] 2+ in aqueous conditions with pH 10. 5 A redox disproportionation reaction is a reasonable rationale for the regeneration of [Ru(trpy)( dma bpy)NH 3 ] +2 from [Ru(trpy)( dma bpy)NH 3 ] + 3 in the presence of a base (NH 3 or DBU) when a chemical oxidant is absent (polypyridyl ligands are omitted in eq. 12) : (12) This reaction should show second - order kinetics in [Ru(trpy)( dma bpy)NH 3 ] + 3 and is easy to monitor by spectrophotometry since the electronic spectra of the reactant ( [Ru(trpy)( dma bpy)NH 3 ] + 3 ) and one of the products ( [Ru(trpy)( dma bpy)NH 3 ] + 2 ) is fully known. The Ru(IV)imido complex , [Ru(trpy)( dma bpy)NH ] + 2 , 3d , is isoelectronic to analogous Ru(IV) oxo intermediates that are reported for catalytic water oxidatio n reactions . Huynh and Meyer, summarize the Ru(IV/III) redox behavior with pH in aqueous systems for [Ru(bpy) 2 (py)H 2 O] 2+ (py = pyridine) catalysts using Pourbaix diagrams as described briefly below: At pH = 0.8, the redox couple is cis - [Ru IV (bpy) 2 (py)(O)] 2+ /cis - [Ru III (bpy) 2 (py)(H 2 O)] 3+ . As the pH is increased above 0.85, the couple becomes cis - [Ru IV (bpy) 2 (py)(O)] 2+ /cis - [Ru III (bpy) 2 (py)(OH)] 2+ . At pH =12.8, the pH - dependent Ru(IV/III) and pH - independent Ru(III/II) couples intersect. As the pH is increased further, E 1/2 of Ru(III/II) becomes more positive that the E 1/2 of Ru(IV/III) and cis - [Ru II I (bpy) 2 (py)(OH)] 2+ is unstable due to disproportionation. In this pH region, cis - [Ru III (bpy) 2 (py)(OH)] 2+ is a stronger oxidant than cis - [Ru IV (bpy) 2 (py)(O)] 2+ because of the pH dependence of the Ru(IV/III) couple. 6 Stabilization of a higher oxidation state by metal - ligand ( 2 b ) ( 2c ) ( 3d ) ( 2a ) 100 multiple bonding or change in pH can become sufficient , so E (VI/III) < E (III/II) for sequential couples, which le = E (VI/III) - E (III/II) < 0 (E is the formal redox potential). If thermodynamically because it is unstable with respect to disproportionation, e.g. , 2Ru (III) Ru(II) + Ru(IV). 6 Examples of isolated Ru(IV) oxo species are existent in the literature. For instance, addition of excess concentrations of (NH 4 ) 2 [Ce IV (NO 3 ) 6 ], CAN, oxidant to solutions of [Ru(TPA)(bpy)] 2+ (TPA= tris ( 2 - pyridyl - methyl)amine, Figure 4.4 . 1 ) in H 2 O (pH= 2 ) at room temperature afforded light green paramagnetic [Ru IV (O)(H + TPA)(bpy)] 3+ through partial ligand dissociation. 7 The product was isolated and the structure was determined by X - ray diffraction. The electronic absorption spectra of [Ru IV (O)(H + TPA)(bpy)] 3+ was featureless within the wavelength window of 400 700 nm and its ESI + - MS spectrum in CH 3 CN contained a peak cluster at m/z=709.1 assigned to {[Ru(O)(H + TPA)(bpy)] 3+ - H + - 2(PF 6 - ))} + . 7 In another case, UV - Vis titration o f [Ru II (bpy) 2 (py)(OH 2 )] 2 + ( Figure 4.4 . 1 ) with CAN in 1 M HClO 4 (aq) exhibited a decrease of the MLCT (metal to ligand charge transfer) band at 470 nm with occurrence of two isosbestic points at 322 and 585 nm. 8 Upon addition of 1 equiv. of CAN, a weak band was observed at 587 nm, which was assigned to the LMCT (ligand to metal charge transfer) band of the Ru(III) complex, [Ru III (bpy) 2 (py)(OH 2 )] 3+ . Further addition of CAN caused a decrease in th e LMCT band at 587 nm and the reaction ended on the addition of 2 equiv. of CAN. The obtained Ru(IV)O complex, [Ru IV (O)(bpy) 2 (py)] 2+ , did not show any specific absorption bands in the visible region. 9 Diamagnetic Ru(IV) = O species have also been reported. In seven - coordin ated [Ru IV (O)(N4Py)(OH 2 )] 2+ (N4Py = (1 , 1 di (pyridin - 2 - yl) N , N - bis (pyridin - 2 - ylmethyl) 101 methanamine), Figure 4.4 . 1 ) the splitting of degenerate d orbitals into a stable pair of d yz and d zx and unstable d xy orbitals causes a singlet diamagnetic ground state. 10 Figure 4.4 . 1 The structure of the mentioned Ru(IV)oxo species . Compared to ruthenium (IV) oxo, the chemistry and characterization of terminal Ru(IV) imido complexes are less developed, which makes it difficult to investigate their potential role in the catalytic NH 3 oxidation cycle. While we were not successful in obtaining direct evidence for their generation in the catalytic cycle, the clues that direct us to the consideration of such interm ediates would be discussed later in this chapter. 4.5 . Reaction of [Ru(trpy)( dma bpy)NH 3 ] + 3 With Bases 4.5.1 . Electrochemistry Addition of DBU to solutions of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM results in a rapid color change from green to orange, and the voltammograms exhibit new redox features as shown in Figure 4.5 . 1 . 102 Figure 4.5 . 1 CVs of a solution of 5.36 10 - 3 M 2 b in NM before (green) and after (orange) addition of 1.34 10 - 2 M DBU. Scan rate 0.1 Vs - 1 . After the reaction with the base, the solution contains 2a and three new red ox active species labeled as I, II and III. According to the starting currents being anodic, the solution after addition of DBU does not contain [Ru(trpy)( dma bpy)NH 3 ] + 3 but instead has two new species that are easier to be oxidized (irreversibly) than [Ru(trpy)( dma bpy)NH 3 ] + 2 (peaks I and II in Figure 4.5 . 1 ). These feature s are absent when excess concentrations of NH 3 is added as the base to solutions of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM, the case which the voltammograms resemble of catalytic oxidation of NH 3 by [Ru(trpy)( dma bpy)NH 3 ] + 2 ( Figure 4.5 . 2 ). While due to the possibility of N - N bond formation, the chemistry is more complicated when NH 3 is used, the very negative onset potential of the catalytic current respect to t he E 1/2 of [Ru(trpy)( dma bpy)NH 3 ] + 3 can be related to more feasible oxidative processes. In the following sections the reaction of [Ru(trpy)( dma bpy)NH 3 ] + 3 with the two bases, NH 3 and DBU are discussed in more details. 103 Figure 4.5 . 2 CVs of a solution containing 4.30 10 - 3 M [Ru(trpy)( dma bpy)NH 3 ] + 3 , 2 b , in NM before (green) and after (red) addition of NH 3 (saturated, 0.93 M, determined by NMR measurements). The black curve shows the non - catalytic oxidation of NH 3 in NM. By adding stoichiometric concentrations of NH 3 to solutions of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM, another redox peak appears at 0.26 V versus Fc +/ 0 ( Figure 4.5 . 3 ), corresponding to a new species in a lower concentration than the starting [Ru(trpy)( dma bpy)NH 3 ] + 3 , judged by the relative anodic peak currents of the newly appeared peak and the redox peak fo r [Ru(trpy)( dma bpy)NH 3 ] + 3 . Due to the possibility of N - N formation in the presence of ammonia which would lead to a hydrazine intermediate, CVs of an authentic ruthenium hydrazine complex were taken in NM. Two redox processes with matching E 1/2 potentials are observed for the Ru(II) - N 2 H 4 , suggesting that its formation under reaction conditions has to be considered. The hydrazine intermediates are discussed in chapter 5 in more details. 104 Figure 4.5 . 3 Vs of 4.30 10 - 3 M [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM without NH 3 (green), and after addition of NH 3 (4.60 10 - 3 M, 1.06 equiv.) (brown).CV obtained for 2.80 10 - 3 M [Ru(trpy)(dmabpy)N 2 H 4 ](PF 6 ) 2 in NM in the absence of NH 3 . 4.5.2 . 1 H NMR Studies Addition of DBU to solutions of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM - d 3 , as mentioned before, is followed by a fast color change and the NMR spectrum shows [Ru(trpy)( dma bpy)NH 3 ] + 2 as the only diamagnetic ruthenium - based species present ( Figure 4.5 . 4 ). When NH 3 is bubbled into an NMR tube containing [Ru(trpy)( dma bpy)NH 3 ] + 3 , the final red solution has two ruthenium species appeared in the room temperature 1 H NMR spectrum, with the major one having resonances matching with [Ru(trpy)( dma bpy)NH 3 ] + 2 except for the singlet at 1.65 ppm being moved to 2.20 ppm in the presence o f NH 3 . The resonances related to the minor species are modestly close to the hydrazine complex, [Ru(trpy)( dma bpy)N 2 H 4 ] + 2 , 2 e ( Figure 4.5 . 5 ). 105 Figure 4.5 . 4 1 H NMR spectra (rt, 500 MHz, nitromethane - d 3 ) of 2 b (bottom, green), 2 b + excess DBU (middle, brown) and 2a (top, red). Comparison between the spectrum obtained after addition of NH3 to 2b and the spectrum of 2a in NM, shows that 2a is the main product of the reaction at room temperatures. The resonances at 1.65, 2.05, 2.97, 3.51 and 3.46 ppm are related to the added DBU. ( 2a ) ( 2 b ) ( 2b ) 106 Figure 4.5 . 5 1 H NMR spectra (rt, 500 MHz, nitromethane - d 3 ) of 2 b + excess NH 3 (bottom, black), 2 e (middle, blue) and 2a (top, red). By comparing the spectra, the major product of the reaction between 2b and NH 3 in NM is 2a . 4.5.3 . UV - Vis Spectrophotometric Titrations Monitoring the regeneration of [Ru(trpy)( dma bpy)NH 3 ] + 2 , 2a , from [Ru(trpy)( dma bpy)NH 3 ] + 3 , 2b , via treatment with bases, provides us with a great tool to investigate the mechanistic details . In this section we focus on titration of [Ru(trpy)( dma bpy)NH 3 ] + 3 with DBU in NM and in chapter 5, similar titrations are discussed when NH 3 is u sed as the base. Absorption spectra of [Ru(trpy)( dma bpy)NH 3 ] + 3 in dry NM has two LMCT bands 11 at 730 nm ( = 5600 M - 1 cm - 1 ) and 840 nm ( = 5800 M - 1 cm - 1 ). Electronic absorption spectrum of [Ru(trpy)( dma bpy)NH 3 ] + 2 in NM has a characteristic Ru(d ) dmabpy( *) MLCT band at 490 ( 2a ) ( 2e ) ( 2b ) 107 nm ( = 15660 M - 1 cm - 1 ) ( See Appendix, Figure A4.6.1). 12 The decrease and increase in the absorption values at 730 nm and 490 nm were used to track the fates of species [Ru(trpy)( dma bpy)NH 3 ] + 3 and [Ru(trpy)( dma bpy)NH 3 ] + 2 , respectively, i n spectrophotochemical titrations of [Ru(trpy)( dma bpy)NH 3 ] + 3 with DBU in NM. For this experiment, a stock solution of 1.11 10 - 3 M [Ru(trpy)( dma bpy)NH 3 ] + 3 (5.7 mg 2b in 5 mL 2.0 mL NM to give 2.15 mL of a final solution of 7.73 10 - 5 M [Ru(trpy)( dma bpy)NH 3 ] + 3 . Based on the absorption reading at 730 nm and using the molar a bsorptivity of [Ru(trpy)( dma bpy)NH 3 ] + 3 in NM, the actual concertation was determined as 7.65 10 - 5 M [Ru(trpy)( dma bpy)NH 3 ] + 3 . The cuvette was capped and sealed in the glove box. A solution of DBU in NM was also prepared by on1) which was further diluted by a factor of 50 to a give a final solution of 2.68 10 - 5 M DBU (solution2). Aliquots of these solutions (solution2 for the first 10 additions, then solution1 until the end of the titration) were added step by step to the cuv ette containing [Ru(trpy)( dma bpy)NH 3 ] + 3 during the experiment and the spectr a were recorded after each addition. The changes in the spectral features are shown in Figure 4.5 . 6 . 108 Figure 4.5 . 6 Top : Changes in the electronic absorption spectra of a green starting solution of 7.65 10 - 5 M 2 b with the addition of DBU t Bottom : Increase in the absorbance at 490 nm due to generation of 2a happens simultaneously with the decrease in absorbance at 730 nm due to consumption of 2b . 109 As data in Fig ure 4. 5.6 ( top ) represents, the LMCT bands of [Ru(trpy)( dma bpy)NH 3 ] + 3 gradually disappears with addition of base, and the reappearance of a new band at 490 nm occurs simultaneously. The negligible initial change in the absorbances in the plot of absorbance (at 730 nm an d 490 nm) versus the equivalent of added DBU (=moles of DBU present in the solution divided by the starting moles of [Ru(trpy)( dma bpy)NH 3 ] + 3 ) ( Figure 4.5 . 6 , bottom ) is attributed to the acid - base equilibrium that is established between the solvent and the base, DBU. The reaction is complete when approximately 1.4 equiv. of DBU is present in the so lution, judging by the spectra not changing with further additions of DBU. Based on room temperature NMR results which showed [Ru(trpy)( dma bpy)NH 3 ] + 2 as the product of the reaction between [Ru(trpy)( dma bpy)NH 3 ] + 3 and excess DBU, the absorption spectrum obt ained at 1.4 equiv. of DBU ( Figure 4.5 . 7 , orange) was compared to the spectrum of a solution of [Ru(trpy)( dma bpy)NH 3 ] + 2 in NM. If it is considered that regeneration of [Ru(trpy)( dma bpy)NH 3 ] + 2 is via the deprotonation of [Ru(trpy)( dma bpy)NH 3 ] + 3 , followed by redox disproportionation of [Ru(trpy)( dma bpy)NH 2 ] + 2 , we can estimate the concentration of the produced [Ru(trpy)( dma bpy)NH 3 ] + 2 at the end of the titration to be 3.46 10 - 5 M in the final solution with the volume of 2.402 mL , based on the stoichiometry of equation 12. The comparison between the constructed a bsorption spectrum for 3.46 10 - 5 M [Ru(trpy)( dma bpy)NH 3 ] + 2 ( Figure 4.5 . 7 , red) and the spectrum collected at the end point of the titration ( 1.4 equiv. DBU) is shown in Figure 4.5 . 7 (red and orange spectra) . To account for the effect of DBU, a solution of [Ru(trpy)( dma bpy)NH 3 ] + 2 and DBU in NM was made and its absorption spectrum was recorded. To prepare this solution, a stock solution of 2.68 10 - 3 M DBU was diluted to a final volume of 5 mL to give a solution of 1.3 4 10 - 4 M. 50 L of a solution containing 4.2 10 - 4 M [Ru(trpy)( dma bpy)NH 3 ] + 2 (3.7 mg 2a in 10 mL NM) was diluted to 0.5 mL by the 110 1.31 10 - 4 M DBU in NM as the solvent. The final concentrations of [Ru(trpy)( dma bpy)NH 3 ] + 2 and DBU in this solution were 4.20 10 - 5 M and 1. 21 10 - 4 M, respectively ( 2a : DBU = 1 : 2.80 or 0.5 : 1.40 ). The absorption spectrum for this solution was obtained by dividing the absorbance by [ 2a ] (=4.20 10 - 5 M) (See Appendix , Figure A4.6.2 ) and was used to construct the spectrum that mimics the final solution of the titration by multiplying molar absorptivity by 3.46 10 - 5 M ( Figure 4.5 . 7 , black). Figure 4.5 . 7 Comparing the electronic absorption spectra obtained at the end of the titration, i.e. 1.4 equiv. DBU added (orange) with the constructed spectrum for 0.5 equiv. 2a in NM (red) and constructed spectrum for 0.5 equiv. 2a + DBU in NM (black). As seen in Figure 4.5 . 7 , neither of the simulated spectra perfectly match with that of the final solution, implying that some other absorbing species are present in th e solution . Mathematical subtraction of the constructed spectra (red and black curves in Figure 4.5 . 7 ) from the spectrum of the endpoint solution results in residual absorption spectra ( See Appendix, Figure A4.6.3). While the identit ies of such co - products are not known at this point, the obser vations in the steady - state titrations of [Ru(trpy)( dma bpy)NH 3 ] + 3 with DBU in NM are consistent with a half equivalence of 111 [Ru(trpy)( dma bpy)NH 3 ] + 2 being generated by this reaction, supporting the concept of the redox disproportionation. 4.6 . Rotating Ring Disk Electrode (RRDE) Experiments Electrochemical experiments using two working electrodes in the solution provide invaluable information about the intermediates of a reaction. Those experiments are performed under hydrodynamic conditions achieved by enforced convection to overcome migration - related mass transport limitations, thus leaving diffusion to be the only factor affecting the mass transport. Under such circumstances, when an intermediate is generated at one working electrode it can be detected at the other closely - positioned working electrode. The potential of the second working electrode is often set to a constant value at which the intermediate could undergo another redox reaction, generating a current that carries information about that desi red species. Commercially available RRDEs are as excellent tool s for such experiments since the two working electrodes are closely positioned respect to each other and the rotation of the electrodes is precisely controlled . The RRDE used in this study cons isted of a glassy carbon disk electrode (WE1) embedded into a Teflon shaft. Concentric to the disk, a platinum ring electrode was placed (WE2). The micrometer scale gap between the two electrodes allows the flow of the disk products to the ring ( Figure 4.6 . 1 ). The shaft was connected to an electric motor that enabled a very fine control of the electrode's angular speed. A bipotentiostat control led the pote ntial of both WE2 and WE1 separately and the current of the of each electrode was measured against a platinum mesh counter electrode. 112 Figure 4.6 . 1 The schematic structure of an RRDE electrode is illustrat ed on the left. On the right, the redox events at the disk and ring are displayed for a redox active species with a generic CV response as shown on the top. As shown in Figure 4.6 . 1 , for a solution containing a reducing species, Red , the rotation of the electrode provides a steady state flow of Red to the diffusion layer of the disk electrode through convection. When the p otential of the disk (WE1) is positive respect to the E 1/2 , e.g., E(WE2) in Figure 4.6 . 1 , Red undergoes oxidation to produce Ox at the surface of the elec trode. The redox events at the disk electrode are normally unaffected by the ring; while the current at the ring is totally determined by the products of the disk reaction. If the lifetime of the species Ox is ~ (d = the gap between the disk and the ring in cm, D= diffusion constant of Ox in cm 2 s - 1 ), it would reach to the ring electrode. In cases where the potential of the ring, WE2 is set to a value more negative than the E 1/2 , e.g. , E(WE 1 ) in Figure 4.6 . 1 , Ox can be reduced back to Red , causing a reductive current to be observed for the ring. However, n ot all of the generated Ox could reach to WE2 (ring) and onl y a fraction of the disk products is collected by the ring. The term correction efficiency, N, is used to describ e the fraction of the collected products at the ring and is defined as 113 the ratio of the ring limiting current to the disk limiting currents (where the current is plateaued). N is governed by the geometry of the RRDE or more specifically the gap between the two electrodes and should be independent from the rotation . If in an experiment, the observed N is rotation speed d ependent, one can conclude that the lifetime of the intermediates are short and major losses occur within the gap between the disk and the ring. 13 In this work, two types of experiments were performed in RRDE measurements. In one set of experiments, t he RRDE was immersed in a solution containing [Ru(trpy)( dma bpy)NH 3 ] + 2 in THF and the potential of the disk electrode was scanned linearly. The ring was set a constant potential, and the redox events at the rings were studied. The other set of experiments w ere carried out by experiments were running both in the absence and presence of DBU base. For these experiments, the potentials are reported vs. NH E (conv ersion was done based on Fc +/0 in THF) . 4.6.1 . Titration with DBU: Scanning E D , Constant E R Before the addition of the DBU, CVs were taken in a solution of 5.57 10 - 4 M [Ru(trpy)( dma bpy)NH 3 ] + 2 in THF at both disk and ring electrodes ( Figure 4.6 . 2 , top ). The s ame redox potentials were observed for Ru(II/III)NH 3 couple at each electrode, with the ring showing less currents due to its smaller surface area. Then , the electrodes were rotated with three rotation speeds, and the potential of the disk (E D ) was scanned at a rate of 10 mVs - 1 , exhibiting a typical hydrodynamic s teady - state i - E curve ( Figure 4.6 . 2 , bottom ) for the disk . The potential of the ring (E R ) was set at a constant value of 0 .00 V vs. NHE, negative to the E 1/2 of [Ru(trpy)( dma bpy)NH 3 ] + 2 to collect the products of the disk reaction. Under these conditions, for E D < E 1/2 no oxidation occurred at the disk (i D =0), consequently no currents were observed for the ring because there was 114 no products to be collected . When E D > E 1/2 , oxidation of [Ru(trpy)( dma bpy)NH 3 ] + 2 to [Ru(trpy)( dma bpy)NH 3 ] + 3 took place at the disk and produced [Ru(trpy)( dma bpy)NH 3 ] + 3 was reached to the ring and reduced back to [Ru(trpy)( dma bpy)NH 3 ] + 2 , exhibiting a reductive current (i R ). 115 Figure 4.6 . 2 Top : CVs obtained for a solution containing 5.57 10 - 4 M 2 a in THF collected at the ring and the disk show that the E 1/2 of the complex does not change with the material of the electrode . Bottom : The currents associated with the disk (i D ) and ring (i R ) , when the potential of the disk is being scanned linearly and the ring is set at a constant reductive potential of 0.00 V vs. NHE . Three rotation speeds are shown: 200 rpm (red), 5 00 rpm (green) and 1000 rpm (blue). N = 24% (rotation speed independent), consistent with the value reported by the manufacturer. 116 Step by Step ad dition of aliquots of a DBU solution in THF to the electrochemistry cell resulted in the appearance of a pre - peak in the CV s of the disk ( Figure 4.6 . 3 , top ). The magnitude of the pre - peak current increas ed linearly with the concentration of DBU ( See Appendix , Figure A4.7.1) while its potential demonstrat ed a positive shift as [DBU] was increa sing . Linear scan voltammograms (LSVs) were then obtained after each addition. For instance, for the first addition of DBU where [DBU] = 1.10 10 - 5 M, the potential of the disk was scanned and the reduction of the products of the disk reaction(s) were colle cted reductively at the ring ( Figure 4.6 . 3 , bottom ). 117 Figure 4.6 . 3 Top : CVs obtained after addition of DBU to the solution containing 5.57 10 - 4 M 2 a in THF collected at the disk. Bottom : The currents associated with the disk and ring for the experiment with [DBU]= 1.10 10 - 5 M. Three rotation speeds are shown: 200 rpm (red), 500 rpm (green) and 1 000 rpm (blue). N < 24% (rotation speed independent). As seen in Figure 4.6 . 3 , bottom the LSVs of the disk show two diffusion limited currents, similar to the CVs in Figure 4.6.3, top 118 when compared to Figure 4.6 . 2 , bottom . The product(s) of the process responsible for the pre - peak are not reduced at the ring. This could be either due to their instability and very short lifetimes (being consumed in a rapid coupled chemic al reaction) or because their reduction takes place at more negative potentials and is not observable within the selected potential window. The ring, however, shows small reductive currents at potentials more positive than 0.7 V versus NHE. Since the Ru(II /III)NH 3 peak still has a return peak in the corresponding CV, this reductive ring current might be due to the reduction of Ru(III)NH 3 species. However, the very low collection efficiencies ed to what was seen in the control experiment ( Figure 4.6 . 2 , bottom ) may be indications of the reduction of a new species. 4.6.2 . Titration with DBU: Scanning E R , Constant E D In another experiment, the potential of the disk electrode was held at a contact oxidative potential of 1 .00 V vs NHE (E D > E 1/2 ( 2a )) in a solution containing [Ru(trpy)( dma bpy)NH 3 ] + 2 in THF . ( ) At this potential, oxidation of Ru(II ) to Ru(III) occur ed , thus a constant, mass transport - limited oxidative current was seen for the disk electrode. Meanwhile, the potential of the ring was scanned linearly to phish for any redox - active products of the d isk reaction. As seen in Figure 4.6 . 4 , when E R < E 1/2 /( 2a ), reduction of [Ru(trpy)( dma bpy)NH 3 ] + 3 product ( 2b ) was taking place at the ring. At E R > E 1/2 /( 2a ), oxidation of the [Ru(trpy)( dma bpy)NH 3 ] + 2 , 2a , to [Ru(trpy)( dma bpy)NH 3 ] + 3 , 2b , happen ed at the ring. See Appendix A4 for experiments with different E D values. 119 Figure 4.6 . 4 RRDE LSVs for 5.57 10 - 4 M 2 a in THF with the disk being held at a potential more positive than the E 1/2 of 2a . Constant oxidative disk currents are labeled as i D . The product of the disk oxidation is being collected and reduced at the ring when E ring is lower than E 1/2 of 2a . Three rotation speeds are shown: 200 rpm (red), 500 rpm (green) and 1000 rpm (blue). N = 24% (rotation speed independent). Applying the same electrochemical conditions to a solution containing 5.57 10 - 4 M [Ru(trpy)( dma bpy)NH 3 ] + 2 and 1.10 10 - 5 M DBU results in LSV responses that are shown in Fig ure 4.6 . 5 . First , we noticed that while the currents at the disk remain almost constant, they do not change with rotation speed and their magnitude is smaller than what was observed for the control experiment in Figure 4.6 . 4 . One possible scenario is that the disk current is controlled by a sluggish kinetically controlled process. The currents at the ring also do appear to be independent from the events at the disk. 120 Fig ure 4.6 . 5 RRDE LSVs for the experiment with 5.57 10 - 4 M 2 a and [DBU]= 1.10 10 - 5 M in THF . Three rotation speeds are shown: 200 rpm (red), 500 rpm (green) and 1000 rpm (blue). The potential of the disk is held at the constant value of 1.0 V vs. NHE, while the potential of the ring was scanned linearly at a rate of 10 mVs - 1 . It has to be noted that only in RRDE experiments, a positive shift was observed after each addition of DBU . Figure 4.6 . 6 , top shows that with increasing [DBU], the redox potentials are shifted positive ly. This, reduces the accuracy of the applied E D . When the peaks are moved so that the second ( Figure 4.6 . 6 , bottom ), reversible peaks lay on top of each othe r , the CVs match with other experiments that were conducted using a stationary (different) glassy carbon electrode (no ring) (See Appendix , Fig ure A4.7.5). It is still unclear that whether these issues are related to the instability of the reference electrode or the conditions of the RRDE experiments, i.e., presence of a platinum ring (since it might catalyze the produ ction of interfering spieces ) or convection which might affect the electrochemistry at the glassy carbon disk. More careful RRDE experiments need to be conducted, specifically polishing of the electrodes in between DBU addition s and p er forming EDX examinations of the surface of the d isk at the end of the experiment are strongly 121 recommended. Nevertheless, multiple LSV measurements were performed between every two DBU additions and the CVs taken before and after each RRDE experiments do not agree with depositions on to the surface. Stil l, more investigations regarding these experiments have to be done. Figure 4.6 . 6 CVs taken at the disk without DBU (red) and after two additions of DBU. Top : raw data, Bottom : after the reversible processes (second peaks) were superimposed. 122 4.7 . Conclusions Chemical synthesis of the compound [Ru(trpy)(bdmabpy)NH 3 ](PF 6 ) 3 , 2b , as the first intermediate in the catalytic cycle provided a convenient path to investigate the role of this complex i n the catalytic cycle. Treatment of 2b with DBU as a non - coordinating base resulted in regeneration of [Ru(trpy)(bdmabpy)NH 3 ](PF 6 ) 2 , 2a , which was relate d to a redox disproportionation following the deprotonation of 2b . Reactions of 2b with NH 3 , yielded an other intermediated with characteristics similar to a hydrazine complex as observed in CV experiments. These findings would confirm the one electron oxidation of 2a to 2b to be the initial step in the catalytic cycle. Under catalytic conditions this interm ediate would rapidly react with ammonia to form a hydrazine complex. Since the selected solvent, NM, was not innocent towards the added base, overcoming the solubility of 2b in solvents such as THF or DCM is recommended. A dditionally , reactions o f 2a with DBU were conducted using RRDE experiments in THF, which initially suggested that very short - living intermediates were generated. RRDE experiments with NH 3 as the base (not presented in this dissertation ) were also performed, however the results w ere complicated, probably due to the material of the ring (Pt) which seems to have a catalytic effect on hydrazine decomposition. Thus, for those conditions a RRDE with both the disk and the ring being made of glassy carbon are required. 123 APPENDIX 124 A 4.1 . 1 H NMR Spectra Figure A4.1.1 Full 1 H NMR spectrum (500 MHz, acetonitrile - d 3 , 25 ° C) of 2b (top) and the magnified region between 5 and 10 ppm (bottom). The red square highlights the resonances related to ammonium that is produced after the coordinated NH 3 was replaced with acetonitrile. (ppm) (ppm) 125 Figure A4.1. 2 1 H NMR spectrum (500 MHz, acetonitrile - d 3 with 0.1% HNO 3 , 25 ° C) of the paramagnetic complex 2b added to prevent redox disproportionation. 17 (ppm) 126 Figure A4.1. 3 1 H NMR spectrum (500 MHz, nitromethane - d 3 ) of 2b . The Inset shows the magnified upfield region between 0 to 15 ppm. (ppm) 127 Figure A4.1. 4 The upfield region of the 1 H NMR spectrum (500 MHz, nitromethane - d 3 ) of 2b before (green) and after (red) addition of NH 3 which shows the reappearance of methyl and amine proton resonances, related to regeneration of 2a . A 4.2 . Infrared Spectroscopy Figure A4.2.1 ATIR Spectrum obtained for a solid sample of 2b . The stretches assoc iated to the coordinated ammine are shown. (ppm) 128 A 4.3 . Hydrodynamic Linear Scan Voltammetry of 2b in NM Using an RDE Figure A4. 3 .1 RDE linear scan voltammograms of a solution of 2.10 10 - 3 2b in NM, scan rate 0.01 V s - 1 . The first process ( E 1/2 =0.085 V vs. Fc +/0 ) is the reduction of Ru(III) to Ru(II) and the second process at 0.96 V vs. Fc +/0 is oxidation of Ru(III) to Ru(IV). 129 A 4.4 . Bulk Electro - Reduction of 2b to 2a Figure A4.4.1 Half potential of Me 10 Fc +/0 measured by CV (two successive cycles) versus Fc +/0 in acetonitrile. Scan rate 0.1 Vs - 1 . 130 Figure A4. 4 .2 Currents passing for five steps of electrolysis of 8.0 mg 2b in 10 mL THF (7.77 10 - 4 M) at a constant potential of - 0.25 V versus E 1/2 . A total of 0.78 C charge (1.0 3 equiv. of electrons) was passed at the end of the electrolysis. Figure A4.4.3 Hydrodynamic linear scans obtained using a rotating disk electrode after three steps of reduction of 2b (7.7 × 10 - 6 mol) in THF. (a) 0, (b) 0.28, (c) 0.45 and (d) 0.78 C charge passed. Scan rate 10 mV s - 1 . 131 A 4.5 . DI - MS Data Figure A4. 5 . 1 Fragmentation pattern for protonated DBU. red: experimental and black: simulated spectra. This data shows that for DI - MS results, there is an almost significant difference between observed and simulated m/z values. Figure A4. 5 .2 Ions assigned to charge - to - mass values observed for peaks I and II in Fi gure 4 . 3.1. 132 Figure A4. 5 .3 Mass spectra obtained for direct infusion of a sample containing 2a in NM. The inset is mass - to - charge 280 - 310 the region magnified. This is similar to peak cluster VI in Figure 4.3.1. Figure A4.5.4 Mass spectra obtained for direct infusion of a sample containing 2b in NM. The inset is mass - to - charge 170 - 220 the region magnified. This is similar to peak clusters III and IV in Figure 4.3.1 (triply charges fragmentations). 133 A 4.6 . Vis - Spectrophotometric Da ta Figure A4. 6 .1 Electronic abruption spectra of 7.10 10 - 6 M 2a (red) and 7.65 10 - 5 M 2b (green) in NM . Figure A4. 6 .2 Electronic abruption spectrum of a solution containing 4.20 10 - 5 M 2a and 1.18 10 - 4 M DBU in NM . 2a 2b 134 Figure A4. 6 .3 Subtraction spectra (blue) for two cases: Top : The simulated spectrum for 0.5 equiv. 2a plus 2.8 (=2 1.4) equiv. DBU (black) is subtracted from the final absorption spectrum (orange), Bottom : The simulated spectrum for 0.5 equiv. 2a (red) is subtracted from the final absorption spectrum (orange). 135 A 4.7 . RRDE Experiments Figure A4. 7 .1 Dependence of the pre - peak current (i p ) on the concentration of DBU. Data extracted from Fig. 4.14. 136 Figure A4. 7 .2 Top : CVs dependence on the scan rate on solution contain ing 4.58 10 - 4 M 2a and 1.55 10 - 4 M DBU. Scan rates are 0.05, 0.1, 02, 0.3, 0.4, 0.6, 0.8 and 1.0 Vs - 1 . Bottom: dependence of the pre - peak potential in V vs. NHE (left) and current ( bottom ) on the scan rate. 137 Figure A4. 7 .3 CVs obtained for a solution containing 5.0 10 - 4 M 2a in THF coll ected at the ring and the disk. Scan rate 0.1 Vs - 1 . This CVs was taken prior to the RRDE control experiments shown in Figure A4.7.4. 138 Figure A4 .7 .4 RRDE LSVs obtained for a solution containing 5.0 10 - 4 M 2a in THF. The disk was set at a constant potential of 0 ( top ), 0.77 ( bottom ) and 0.9 V ( next page ) vs. NHE while the ring was scanned. Rotation speeds: 200 (red), 500 (green) and 1000 (blue) rpm. 139 Figure A4.7.4. It appears that when the disk potential is almost equal to E 1/2 , the disk potential is affected by the ring events! 140 Figure A4. 7 .5 Raw CVs obtained after additions of DBU to a solution of 7.5 mM 2a in THF, scan rate 0.1 Vs - 1 (top). No shifting in the potentials were observed (compared to the CVs taken using a RRDE) . The linear dependence of the pre - peak potential (bottom) and current (next page) to [DBU]. 141 Figure A4.7.5 142 REFERENCES 143 REFERENCES (1) Schubert, E. M. J. Chem. Educ. Educ. 1959 , 69 (1), 62. (2) Bain, G. A.; Berry, J. F. J. Chem. Educ. 2008 , 85 (4), 532. (3) Kettle, S. F. A. In Physical Inorganic Chemistry ; Springer Berlin Heidelberg, 1996. (4) Lin, L.; Yu, Q.; Yan, X.; Hang, W.; Zheng, J.; Huang, B. Analyst 2010 , 135 , 2970 2978. (5) Coia, G . M.; Demadis, K. D.; Meyer, T. J.; Carolina, N.; Hill, C.; Carolina, N.; January, R. V. 2000 , 2212 2223. (6) Huynh, M. H. V.; Meyer, T. J. Chem. Rev. 2007 , 107 , 5004 5064. (7) Kojima, T.; Nakayama, K.; Ikemura, K.; Ogura, T.; Fukuzumi, S. J. Am. Chem. Soc. 2011 , 133 , 11692 11700. (8) Meyer, T. J.; Moyer, B. a. J. Am. Chem. Soc. 1979 , No. 101, 1326 1328. (9) Ishizuka, T.; Kotani, H.; Kojima, T. Dalt. Trans. 2016 , 45, 16727 16750. (10) Ohzu, S.; Ishizuka, T.; Hirai, Y.; Jiang, H.; Miyuki, S.; Takashi, O.; Shunichi, F.; Kojima, T. Chem. Sci. 2012 , 3 , 3421 3431. (11) Concepcion, J. J.; Tsai, M. - K.; Muckerman, J. T.; Meyer, T. J. J. Am. Chem. Soc. 2010 , 132 (14), 1545 1557. (12) Meyer, T. J.; Thompson, M. S. J. Am. Chem. Soc. 1981 , 103 , 5579 5581. (13) Bard, a J.; Itaya, K.; Malpas, R. E.; Teherani, T. J. Phys. Chem. 1980 , 84 (1975), 1262 1266. (14) M atthews, W. S.; Bares, J. E.; Bartmess, J. E.; Bordwell, F. G.; Cornforth, F. J.; Drucker, G. E.; Margolin, Z.; Mccallum, R. J.; Mccollum, G. J.; Vanier, N. R. J. Am. Chem. Soc. 1975 , 97 (24), 7006 7014. (15) Bordwell pKa Table (Acidity in DMSO). https://www.chem.wisc.edu/areas/reich/pkatable/ (16) Shaffer, D. W.; Xie, Y.; Concepcion, J. J. Chem. Soc. Rev. 2017 , 46 , 6170 6193. (1 7 ) Habibzadeh, F.; Miller, S. L.; Hamann, T. W.; Smith, M. R. Proc. Natl. Acad. Sci. 2019 , 3 (l), 2 6. 144 THE [RU(TRPY)(DMABPY)N 2 H 4 ] 2+ INTERMEDIATE 145 To elucidate the catalysis mechanism, identification of the possible intermediates in the catalytic cycle is very important. Such intermediates were not detected under catalytic conditions, i.e., directly from the BE solution, since the spectrophotometric and NMR experiments after the electrolysis only showed the starting [Ru(trpy)(dmabpy)NH 3 ] 2 + , 2a . Thus, independent synthesis of some intermediates was pursued based on what was understood from the previous experiments. In the previous chapter [Ru(trpy)(dmabpy)NH 3 ] 3+ , 2b , as the first intermediate was discussed. Chapter 5 focuses on the chemistry of [Ru(trpy)(dmabpy)N 2 H 4 ] 2+ , 2e . The labeling experiments suggest that N - N bond formation is resulted from the nucleophilic attack of the ammonia to highly electrophilic nitrogen in of the intermediates. This provides evidence for a hydrazine pathway as illus trated in Scheme 5 . 1 . Scheme 5 . 1 The hydrazine pathway to formation of the N - N bond . The nucleophilic attack of the free NH 3 to an imido nitrogen leads into formation of a hydrazine intermediate. The polypyridyl ligands are omitted. While existence of 2d is not yet supported by direct evidence s , it is speculated that an intermediate with a high - valent ruthenium should form to provide a highly electrophilic nitrogen. Attack of the the N 2 H 4 ligand remains coordinated or leaves the metal center under catalytic conditions is still 146 question able. However, as we would discuss the formation of such hydrazine intermediates are very likely to be a part of the cycle. E lectrophilic - nucleophilic reactions between the catalyst and the substrate are reported and studied in several cases. The nucleoph ilic attack of the substrate water to Ru(V) or Ru(IV)oxo species is considered as the main path to O - O bond formation in homogeneous catalytic oxidation of water when single site ruthenium catalysts are employed . 1 3 Meyer and Thompson observed that when the oxidation of [Ru(trpy)(bpy)NH 3 ] 2+ in aqueous medium at pH > 6.8 is monitored by UV - Vis spectrophotometry, full conversion of [Ru(trpy)(bpy)NH 3 ] 2+ to [Ru(trpy)(bpy)NH 2 OH] 2+ is occurring. 4 Additionally, Os(V) hydrazido complexes were isolated from electrolysis solution in which [Os(trpy)(bpy)NH 3 ] 2+ was exhaustively oxidized in the presence of primary amines . 5 These findings are shedding light on the mechanistic details of reactions that fit into the hydrazine pathway. In this c hapter, the experimental results suggesting similar hydrazine pathways in N 2 formation in the introduced catalytic system are discussed. 5.1 . Synthesis of an Authentic Sample of [Ru(trpy)(bpy)N 2 H 4 ] 2+ Complex (2e) 2e was synthesized from the chloro complex following the procedure below: [Ru(trpy)(dmabpy)Cl]Cl weighed into a heavy walled reaction flask equipped with a stir bar (100 mg, 0.15 mmol, 1 equiv.). Hydrazine monohydrate was added as solvent (10 mL, excess). Th e reaction flask was closed with a Teflon screw plug and heated as a closed system for 2 h at 90 ° C. After heating was complete, the reaction mixture was poured into a 125 mL beaker open to the air in a fume hood. The residue in the reaction flask was ri nsed into the beaker with ~1 mL deionized water. Excess ammonium hexafluorophosphate (100 mg, 0.6 mmol, 4 equiv.) was added as a solid into the stirred mixture. After 1 h, brown solids precipitated. The solids were collected 147 by filtration with a disposable fritted Hirsch funnel of 10 - micron pore size. The filter cake was washed with small, dropwise portions of deionized water until the initial brown filtrate ran a lighter pink - tan color. The brown filter cake was dried in the Hirsch funnel for two hours unt il a free - flowing powder resulted when broken up. The mass of the brown solid was 124 mg (0.138 mmol) after vacuum drying, 92 % isolated yield. 1 H NMR (500 MHz; CD 3 CN): 1H), 8.50 (d, J = 8.1 Hz, 2H), 8.38 (d, J = 8.0 Hz, 2H), 8.12 (t , J = 8.1 Hz, 1H), 7.96 (td, J = 7.9, 1.5 Hz, 2H), 7.87 (dd, J = 5.5, 0.6 Hz, 2H), 7.68 (d, J = 2.8 Hz, 1H), 7.42 - 7.38 (m, 3H), 7.21 (t, J = 3.4 Hz, 1H), 6.42 (d, J = 6.8 Hz, 1H), 6.15 (t, J = 3.4 Hz, 1H), 4.41 (s, 2H), 3.35 (s, 6H), 2.99 (s, 6H), 2.62 (s , 2H). 13 C NMR (126 MHz; CD 3 CN): 153.7, 151.8, 149.3, 138.2, 134.5, 128.7, 124.5, 123.8, 110.1, 109.5, 106.7, 106.1, 40.1, 39.8 . 19 F NMR (470 MHz; CD 3 CN): - P = 706.5 Hz). 31 P NMR (202 MHz; CD 3 CN): - 144.6 (septet, JP - F = 706.4 Hz). HRMS (ESI + ) ( ) m /z [M - H] + Calcd for C 29 H 32 N 9 Ru 608.1824; Found 608.1834. Anal. Calcd for C 29 H 33 F 12 N 9 P 2 Ru: C, 38.76; H, 3.70; N, 14.03. Found: C, 39.00; H, 3.63; N, 13.92. IR (KBr) 3378, 3371 cm - 1 . 5 5.2 . Ligand Displacement in 2e / THF / NH 3 1 H NMR spectra of [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ in THF - d 8 corresponding to the bound hydrazine ( F igure 5.2 . 1 ) 6 . Addition of excess concentrations of ammonia to [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ in THF - d 8 does not change the chemical shift of those resonances which suggests that the bound hydrazine is not likely to be replaced by ammonia. The possibility of replacement of the coordinated hydrazine under catalytic conditions would be further discussed. High Resolution Mass Spectrometry (Electro Spray Ionization positive mode) 148 F igure 5.2 . 1 1 H NMR reference spectrum ( 600 MHz, THF - d 8 , 25 °C) of complex 2 e with (top) and without (bottom) excess NH 3 . The starred peak is solvent residual. The displacement of the bound hydrazine with free ammonia was not observed. 5.3 . Cyclic Voltammetry Studies [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ , when completely dry is not soluble in dry solvents without the help of sonication. The cyclic voltammograms of [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ in dry THF and DCM exhibit variable feature over time. The first CVs taken in a freshly made solution of [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ in DCM has two redox features, a quasi - reversible process at E 1/2 (I) = 0.30 V vs Fc +/0 (i p,a /i p,c = 1.35) and an irreversible anodic process at E 1/2 (II)=0.70 V vs Fc +/0 ( Figure 5.3 . 1 , top ). The origin of peak II is unknown, but peak I was assigned to Ru III/II (N 2 H 4 ) couple. When another CV is taken following the first CV, a new redox process appears (Peak III in Figure 5.3 . 1 , 149 bottom ) which gradually vanishes with th e number of scans. The position of this new peak matches with the E 1/2 of [Ru(trpy)(dmabpy) NH 3 ] 2+ but is not behaving reversibly. Peak III remains in the solutions and does not change with scan rate or potential window. It is only absent in the very first CV taken in the freshly made solution. Figure 5.3 . 1 CVs of M 1.85 10 - 3 M 2 e in DCM. Top : the initial two cycles, Bottom : Three successive scans taken around 1 min after the CV in the top . Peak III appears after a couple scans are taken in a freshly made solution, but it goes away with successive cycl ing of the potential. CVs taken after the one shown in the bottom, behave the same way. 150 Upon bubbling NH 3 into the solution, an enhanced current is observed with an onset potential close to peak III ( Figure 5.3 . 2 ). These results may suggest that the coordinated N 2 H 4 ligand does not remain intact when oxidation of the metal center is enforced, and it is probable that [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ is converting to [Ru(trpy)(dmabpy) NH 3 ] 2+ under oxidative conditions. Figure 5.3 . 2 CVs of 1.85 10 - 3 M 2 e in DCM in the absent (black. Three successive cycles) and presence of ammonia (red). The red curve is very similar to the catalytic oxidation of ammonia in the presence of 2a . CV experiments were then performed in THF since it was the solvent used in the electrolysis. Three peaks were present from the beginning of the experiments as shown in Figure 5.3 . 3 , top , with peak I being initially irreversible. The reversibility of peak I was restored at higher scan rates ( Figure 5.3 . 3 , bottom ). 151 Figure 5.3 . 3 CVs of 1.22 10 - 3 M 2 e in THF. Top : three successive cycles in a freshly made solution. Bottom : Changes in the CVs with scan rate (three cycles are shown for each scan rate). The differences in redox behavior of [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ in DCM and THF could be related to the basicity and/or coordinating ability of each solvent. When the potential window is changed 152 to - 0.6 to 0.5 V vs. Fc +/ 0 , a very light shoulder was visible between - 0.3 to - 0.4 V ( Figure 5.3 . 4 , black). Similar to the experiments in DCM, the addition of NH 3 to the solution of [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ in THF result ed in an enhanced current with an onset potential close to peak III ( Figure 5.3 . 4 , red), while the shoulder at ~ - 0.35 V disappe red . Figure 5.3 . 4 CVs of 2.50 10 - 3 M 2 e in THF in the absence (black) and presence (red) of NH 3 . A shoulder that is evident at around - 0.35 is marked with an arrow. By adding [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ to a solution containing [Ru(trpy)(dmabpy) NH 3 ] 2+ in THF, peaks I and II appear ed in the volta mmogram and the magnitude of the current associated with [Ru(trpy)(dmabpy) NH 3 ] 2 + increase d ( Figure 5.3 . 5 , top ), leading to the conclusions that first, peak II is related to [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ and second, peak III is very likely the amine complex, being generated from [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ under experimental conditions. Interestingly, very similar cyclic voltammograms were obtained when stoichiometric concentrations of NH 3 were added to a solution containing [Ru(trpy)(dmabpy) NH 3 ] 2 + ( Figure 5.3 . 5 , bottom ). 153 Figure 5.3 . 5 Cyclic voltammograms of a mixture solution of 1.05 10 - 3 M 2 a and 1.05 10 - 3 M 2 e in THF. Bottom : CVs of a mixture solution of 2a and 1 equiv. NH 3 in THF. Electro - oxidation of free hydrazine at the surface of the glassy carbon electrode in THF takes place at relatively low potentials ( Figure 5.3 . 6 , peak I). According to several studies on electrochemical III I II III I II 154 oxidation of N 2 H 4 in non - aqueous media, hydrazine undergoes 2 - electron oxidation in non - aqueous solutions to the final product diimide, N 2 H 2 . 7 9 Cao et al. argue d that the reduction of the diimide product back to N 2 H 4 is observed as a reductive peak. However, based on the number of electrons passed for the anodic process (n= 0.67), they inferred that only one - third of the hydrazine was converted to diimide in acetonitrile and the rest of the hydrazine molecules act ed a s the proton acceptor, according to equation 1: 7 (1) Upon addition of a strong base, triethylamine, to the hydrazine solutions in acetonitrile, the conversion of N 2 H 4 to N 2 H 2 was quantitative (n=2). While the oxidati on of hydrazine in THF exhibits similar redox features ( Figure 5.3 . 6 ), the anodic to cathodic current ratio is 1 : 1.5 (as opposed to 1: 3, reported in acetonitrile 7 ), which could be related to the better proton accepting properties of THF compared to acetonitri le. However, what is worth noticing is that peak I in Figure 5.3 . 6 is at the same potential where the small shoulder appears in Figure 5.3 . 4 , leaving us with the question of whether free hydrazine is generated under oxidat ive conditions or not. 155 Figure 5.3 . 6 Electro - oxidation of 4.0 10 - 2 M N 2 H 4 in THF at the glassy carbon electrode. 5.4 . Variable Temperature (VT) 1 H NMR Experiments 6 As mentioned in chapter 3, the 1 H NMR of the electrolysis solutions at room temperature only contains [Ru(trpy)(dmabpy) NH 3 ] 2+ as the ruthenium - based component. Treatment of [Ru(trpy)(dmabpy) NH 3 ] 3 + with NH 3 in THF - d 8 at room temperature g ave the same results as well. Figure 5.4 . 1 compares the spectrum of the reaction mixture ([Ru(trpy)(dmabpy) NH 3 ] 3 + + NH 3 ) with 1 H NMR spectra of authentic samples of the amine ([Ru(trpy)(dmabpy) NH 3 ] 2+ ) and hydrazine complexes ([Ru(trpy)(dmabpy) N 2 H 4 ] 2+ ) to show that under experimental conditions no traces of [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ were observed in the product solution. However, when [Ru(trpy)(dmabpy) NH 3 ] 3 + is reacted with excess concentrations of ammonia in DCM other intermedi ates were captured at low temperatures ( Figure 5.4 . 2 ). 156 Figure 5.4 . 1 1 H NMR spectrum ( 500 MHz, THF - d 8 , 25 ° C) of the reaction solution of 2 b and excess NH 3 (top). 1 H NMR spectra (500 MHz, THF - d 8 , 25 ° C) of complex 2a (middle) and 2e obtained under same conditions, shown for comparison. To do this experiment, [Ru(trpy)(dmabpy) NH 3 ] 3 + gas - tight, medium pressure, high vacuum Teflon - valved NMR tube followed by addition of CD 2 Cl 2 (0.5 mL) and the NMR tube was attached to a high vacuum manifold. Two cycles of freeze/pump/thaw were conducted in liquid nitrogen to de - gas the solvent. The NMR tube was maintained under static vacuum until a fixed volume of anhydrous, de - gassed ammonia was opened to the line and condensed into the NMR tube while in the D ewar of liquid nitrogen. The NMR tube was closed and removed from the line while behind a blast shield. The NMR tube was kept behind a blast shield in a D ewar of liquid nitrogen until ready to be inserted into the NMR (ppm) 157 spectrometer. The resonances of one of the trapped intermediates at - 75 C match es with those obtained for [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ under the same experimental conditions. Integration of the low - field resonances at 9.27 and 8.62 ppm in Fig ure 5. 4.2 attributed to [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ and [Ru(trpy)(dmabpy) NH 3 ] 2+ , respectively, g ave a [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ : [Ru(trpy)(dmabpy) NH 3 ] 2+ ratio of 1:5. However, the presence of other resonances in the spectrum in Figure 5.4 . 2 suggest that there are yet unidentified species that account for the remaining mass balance. Figure 5.4 . 2 1 H NMR spectra ( 500 MHz, CD 2 Cl 2 , - 75 °C) of complex 2 a with ~ 4000 equiv. NH 3 (top), complex 2 e with ~4000 equiv. NH 3 (middle), and the reaction mixture that results when ~ 4000 equiv. of NH 3 is added to 2b at - 85 ° C. (ppm) 158 It is still not fully understood why [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ does not appear in the NMR spectrum at room temperature, even though the solubility is expected to be enhanced compared to - 75 C. Heating up the reaction mixture ( Figure 5.4 . 2 , bottom) gradually to room temperature shows that the resonances related to the hydrazine intermediate disappear and the final solution only shows [Ru(trpy)(dmabpy) NH 3 ] 2 + in the 1 H NMR spectrum ( Figure 5.4. 3 ). Figure 5.4 . 3 Change in the 1 H NMR ( 500 MHz, CD 2 Cl 2 ) spectral features as the temperature of the solution containing 2b and excess NH 3 is increasing. As the temperature is being increased, the resonances related to the hydrazine complex vanish and the NMR spectrum at room temperature only contains 2a . (ppm) 159 The mechanism of these conversions is unclear. Nevertheless, this phenomenon has to be considered whe n the treatment of [Ru(trpy)(dmabpy) NH 3 ] 3 + with NH 3 is being studied by other techniques specially spectrophotochemical titrations. 5.5 . Spectrophotochemical Studies The reaction between [Ru(trpy)(dmabpy) NH 3 ] 2+ and NH 3 was also studied by UV V is - spectroscopy in THF containing 0.8% of NM to help the solubility of [Ru(trpy)(dmabpy) NH 3 ] 3 + . Separate solutions of complexes [Ru(trpy)(dmabpy) NH 3 ] 2+ , [Ru(trpy)(dmabpy) NH 3 ] 3 + , and [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ in these solvent solution of each complex in NM with THF to a final volume of 5 mL. Electronic absorption spectrum of [Ru(trpy)(dmabpy) NH 3 ] 2+ in NM(0.8%V)/THF has a characteristic Ru(d ) dmabpy( *) MLCT band at 496 nm wit h a molar absorption coefficient ( ) of 10490 M - 1 cm - 1 (See Appendix, Figure A5.1.1). This band is slightly shifted to higher energies for [Ru(trpy)(dmabpy)N 2 H 4 ] 2+ and is almost non - existent for solutions of [Ru(trpy)(dmabpy)NH 3 ] 3+ in the solvent mixture (See Appendix, Figures A5.1.2, and A5.1.3). Since the UV cut - off of NM is 380 nm, the spectra collection was limited to wavelengths between 400 - 900 nm. Stock solutions of NH 3 h 5 mL of the solvent mixture for 5 min and the concentration was assumed to be 0.34 M. Aliquots of the NH 3 stock solutions were transferred to a cuvette containing 9.70 10 - 5 M [Ru(trpy)(dmabpy) NH 3 ] 3 + in the mixture solvent using a glass syringe. All absorbance spectra were collected against an NM (0.8%V)/THF reference solution. Right after the addition of ammonia to [Ru(trpy)(dmabpy) NH 3 ] 3 + , the spectral features change and consumption of [Ru(trpy)(dmabpy) NH 3 ] 3 + is evident via gradual disappearing of t he bands at 730 nm and 840 nm. 160 Simultaneously, a band is reappearing between 480 - 500 nm that suggests Ru(II) species are being generated as the products of the reaction ( Figure 5.5 . 1 ). Figure 5.5 . 1 Changes in the absorption spectrum of a solution of 9.70 10 - 5 M 2 b with the addition of NH 3 . The green curve is the spectrum of the starting solution and the final orange spectrum is when the titration was ended (the point where the addition of NH 3 did not change the spectra). All absorption spectra are corrected for dilution. Analysis of the results of the titration experiment, were performed based on the assumption that t he product of the reaction between [Ru(trpy)(dmabpy) NH 3 ] 3 + and NH 3 is a mixture containing equal concentrations of [Ru(trpy)(dmabpy) NH 3 ] 2+ and [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ . Here we assume that the presence of NH 3 in the solution triggers the redox disproportionation and 1 equiv. of [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ is formed along with each equiv. of [Ru(trpy)(dmabpy) NH 3 ] 2+ and it survives long enough to be seen by UV V is - spectrophotometry. At the end of the titration 8.05 10 - 7 mol NH 3 is present in the solution, 4 - fold excess with respect to the initial moles of [Ru(trpy)(dmabpy) NH 3 ] 3 + , so the conversion of [Ru(trpy)(dmabpy) NH 3 ] 3 + to half equivalent of 161 each product [Ru(trpy)(dmabpy) NH 3 ] 2+ and [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ is expected to be quantitative. Based on the calculation, at the end of the titration the concentrations of [Ru(trpy)(dmabpy) NH 3 ] 2+ and [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ should be equal to 4.70 10 - 5 M (=1.21 10 - 7 mol / 2.583 10 - 3 L). The constructed absorption spectra for 4.70 10 - 5 M of [Ru(trpy)(dmabpy) NH 3 ] 2 + and 4.70 10 - 5 M [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ is compared to the final spectra in Figure 5.5 . 2 . Figure 5.5 . 2 Comparison between the absorption spectra recorded at the end of titration (orange), the constructed spectra for 0.5 equiv. 2a (black) and 0.5 equiv. 2e (blue) and the arithmetic summation of the spectra of 0.5 equiv. 2a (black) and 0.5 equiv. 2e (blue) shown in dashed red. When the constructed spectra for 0.5 equiv. of [Ru(trpy)(dmabpy) NH 3 ] 2+ and [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ were added together, the final spectr um of the solution after the titration was regenerated in terms of the position of bands. While this data is consistent with the formation of the hydrazine complex from [Ru(trpy)(dmabpy) NH 3 ] 3 + after treatment with NH 3 , it is still not 162 clear why [Ru(trpy)(dmabpy) N 2 H 4 ] 2+ is not observed in quantitative concentrations by 1 H NMR measurements. 5.6 . Conclusions The formation of a hydrazine intermediate is experimentally supported. However, the fate of the hydrazine complex u nder catalytic conditions is still ambiguous. Future studies on an authentically synthesized Ru(II) - N 2 complex would provide invaluable insights to the regeneration of the amine complex from the hydrazine complex. 163 APPENDIX 164 A 5.1 . Electronic absorption spectra in NM (0.8% V)/THF Figure A5.1.1 Absorption spectrum of 8.78 10 - 5 M 2a in NM(0.8%)/THF. Figure A5.1.2 Absorption spectrum of 9.70 10 - 5 M 2b in NM(0.8%)/THF. 165 Figure A5.1.3 Absorption spectrum of 9.00 10 - 5 M 2e in NM(0.8%)/THF. 166 REFERENCES 167 REFERENCES (1) Concepcion, J. J.; Tsai, M. - K.; Muckerman, J. T.; Meyer, T. J. J. Am. Chem. Soc. 2010 , 132 (14), 1545 1557. (2) Wasylenko, D. J.; Ganesamoorthy, C.; Kolvisto, B. D.; Henderson, M. A.; Berllnguette, C. P. Inorg. Chem. 2010 , 49 (5), 2202 2209. (3) Kamdar, J. M.; Grotjahn, D. B. Molecules 2019 , 24 (3), 494 518. (4) Meyer, T. J.; Thompson, M. S. J. Am. Chem. Soc. 19 81 , 103 , 5579 5581. (5) Coia, G. M.; White, P. S.; Meyer, T. J.; Wink, D. A.; Keefer, L. K.; Davis, W. M. J. Am. Chem. Soc. 1994 , 116 (4), 3649 3650. (6) Habibzadeh, F.; Miller, S. L.; Hamann, T. W.; Smith, M. R. Proc. Natl. Acad. Sci. 2019 , 3 (l), 2 6. (7) Cao, X.; Wang, B.; Su, Q. J. Electroanal. Chem. 1993 , 361 (1 2), 211 214. (8) Antoniadou, S.; Jannakoudakis, A. D.; Theodoridou, E. Synth. Met. 1989 , 30 , 295 304. (9) Michlmayr, M.; Sawyer, D. T. J. Electroanal. Chem. 1969 , 23 , 375 385. 168 169 PRELIMINARY ELECTROC HEMICAL STUDIES OF O THER RUTHENIUM POLYPYRIDY L AMINE CATALYSTS 170 In order to prepare ruthenium complexes with E 1/2 potentials more negative than [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 , 1 a , more ruthenium polypyridyl amine complexes were prepared . In these complexes, the same dmabpy ligands was used as in [Ru(trpy)(dmabpy)NH 3 ](PF 6 ) 2 , 2 a , but various electron donating groups were placed on different position s on the trpy moiety. Three complexes were synthesized and t he ir catalytic behavior towards oxidation of ammonia was studied using cyclic voltammetry methods. 6.1 . [Ru(Me 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 , (3a). 6.1.1 . Synthesis 3a was synthesized in two steps as described below: 6.1.1.1 . (4',4,4'' - trimethyl - 2,2':6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) chloro ruthenium(II) chloride, [Ru(Me 3 trpy)(dmabpy)Cl]Cl. [Ru(Me 3 trpy)(dmabpy)Cl]Cl was synthesized following a procedure described in section 3. 3.1 with 1:1 ethanol : water being used as the solvent. 78% yield. 1 H NMR (500 MHz, DMSO - d 6 ) 9.47 (d, J = 6.7 Hz, 1H), 8.57 (s, 2H), 8.45 (s, 2H), 7.84 (d, J = 2.7 Hz, 1H), 7.55 (dd, J = 8.9, 4.2 Hz, 3H), 7.25 (dd, J = 6.9, 2.7 Hz, 1H), 7.23 7.19 (m, 2H), 6.35 (d, J = 6.8 Hz, 1H), 6.23 (dd, J = 7.0, 2.8 Hz, 1H), 3.29 (s, 6H), 2.93 (s, 6H), 2.72 (s, 3H), 2.45 (s, 6H). 171 6.1.1.2 . ( 4',4,4'' - tr imethyl - 2,2':6',2'' - terpyridine) (4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) ruthenium (II) ammine dihexafluorophosphate, [Ru(Me3trpy )(bdmabpy)NH 3 ](PF 6 ) 2 , (3a). [Ru(Me 3 trpy)(bdmabpy)NH 3 ](PF 6 ) 2 was synthesized following the procedure described for 2a . 85% yield. 1 H NMR (500 MHz, Acetonitrile - d 3 ): 8.65 (m, 1H), 8.31 (s, 2H), 8.19 (s, 2H), 7.64 (dd, J = 11.1, 4.3 Hz, 3H), 7.34 (d, J = 2.9 Hz, 1H), 7.17 (d, J = 5.7 Hz, 2H), 7.11 (dd, J = 6 .7, 2.8 Hz, 1H), 6.44 (dd, J = 6.9, 1.1 Hz, 1H), 6.13 (dd, J = 7.0, 2.8 Hz, 1H), 3.30 (d, J = 1.1 Hz, 6H), 2.96 (d, J = 1.0 Hz, 6H), 2.77 (s, 3H), 2.50 (s, 6H), 1.60 (s, 3H). 6.1.2 . Ammonia Oxidation Using 3a as the Catalyst Cyclic voltammograms of a solution c ontaining 2.70 10 - 3 M [Ru(Me 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 , 3a , in THF ha s one reversible redox process at E 1/2 = 0.12 V versus Fc +/0 in the absence of ammonia. Bubbling NH 3 to this solution cause s the loss of the reversibility of this redox process and an enhanced catalytic current is observed as shown in Figure 6.1 . 1 . To evaluate the catalytic nature of the current after addition of ammonia, normalized catalytic currents were plotted and compared. As seen in Figure 6.1 . 2 , the catalytic current is enhanced at lower scan rates as well as the potential of the current onset is also moved to more negative. 172 Figure 6.1 . 1 Top: CVs of 3 a in the absence of NH 3 (sat d) in THF. Scan rates : 0.05 , 0.1 , 0.25 and 0.5 Vs - 1 . Bottom: The catalytic current (black) in a solution 2.7 mM 3a in THF. Dotted green line is the non - catalytic NH 3 oxidation. S can rate is 0.1 Vs - 1 and the onset of the catalyti c current is - 0.1 V versus Fc +/0 . 173 Figure 6.1 . 2 Top : Catalytic currents obtained in solutions of 2.7 mM 3 a and NH 3 (sat d) in THF and their dependence on the scan rate. Bottom : the catalytic currents normalized for scan rate. 174 6.2 . [Ru( t Bu 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 , (4a) . 6.2.1 . Synthesis 4 a was synthesized in two steps as described below: 6.2.1.1. (4',4,4'' - tri - tert - butyl - 2,2':6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) chloro ruthenium (II) chloride, [Ru( t Bu 3 trpy)(dmabpy)Cl]Cl. [Ru( t Bu 3 trpy)(bdmabpy)Cl]Cl was synthesized following a procedure described in section 3. 3.1 . 86% yield. 1 H NMR (500 MHz, DMSO - d 6 ) J = 6.7 Hz, 1H), 8.77 (s, 2H), 8.67 (d, J = 2.1 Hz, 2H), 7.84 (d, J = 2.7 Hz, 1H), 7.57 (dd, J = 12.7, 4.4 Hz, 3H), 7.40 ( dd, J = 6.0, 2.1 Hz, 2H), 7.26 (dd, J = 6.7, 2.7 Hz, 1H), 6.38 (d, J = 6.8 Hz, 1H), 6.31 6.17 (m, 1H), 3.30 (s, 6H), 2.94 (s, 6H), 1.63 (s, 9H), 1.34 (s, 18H). 6.2.1.2. ( 4',4,4'' - tri - tert - butyl - 2,2':6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) ruthenium (II) ammine dihexafluorophosphate, (4a). 175 [Ru( t Bu 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 was synthesized using the same procedure that was followed for 2a . Yield 83%. 1 H NMR (500 MHz, Acetonitrile - d 3 ) 8.37 (dd, J = 2.2, 0.7 Hz, 2H), 7.71 (dd, J = 6.0, 0.6 Hz, 2H), 7.64 (d, J = 2.7 Hz, 1H), 7.36 7.33 (m, 3H), 7.12 (dd, J = 6.7, 2.8 Hz, 1H), 6.44 (d, J = 6.8 Hz, 1H), 6.14 (dd, J = 6.9, 2.8 Hz, 1H), 3.31 (s, 6H), 2.96 (s, 6H), 1.64 (s, 9H), 1.59 (s, 3H), 1.37 (s, 18H). 6.2.2 . Ammonia Oxidation Using 4a as the Catalyst One reversible redox process is obse rved in cyclic voltammograms at E 1/2 = - 0.03 V versus Fc +/0 in of a solution containing 2.51 10 - 3 M [Ru( t Bu 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 , 4a , in THF the absence of ammonia ( Figure 6.2 . 1 , top ). Saturation of the solution with NH 3 changes the electrochemical response to a purely anodic catalytic current as shown in Figure 6.2 . 1 , bottom . The dependence of the catalytic currents to the scan rate is shown in Figure 6.2 . 2 . 176 Figure 6.2 . 1 CVs of 4 a in the absence ( top ) and presence (bottom) of NH 3 (sat d) in THF. Scan rates for the CVs on the top : 0.05, 0.1, 0.25, 0.5 and 0.8 Vs - 1 . On the bottom , the scan rate is 0.1 Vs - 1 and the onset of the catalytic current is - 0.15 V versus Fc +/0 . 177 Figure 6.2 . 2 Top : Catalytic currents obtained in solutions of 2.51 mM 4 a and NH 3 (sat d) in THF and their dependence on the scan rate. Bottom : the catalytic currents normalized for scan rate. 178 6.3 . [Ru(dmaptrpy)(dmabpy)NH 3 ](PF 6 ) 2 , (5a). 6.3.1 . Synthesis 6.3.1.1. (4 - N,N - dimethylaminophenyl) - 2,2',6',2'' - terpyridine, dmaptrpy. The dmaptrpy ligand was synthesized initially using a one - step literature method: 1 2 - acetyl pyridine (0.25 g, 2.06 mmol, 2 equiv.) and 4 - dimethylaminobenzaldehyde (0.15 g, 1.00 mmol, 1 equiv.) were added to a round bottom flask containing a solution consist ing of 0.12 g (2.14 mmol, 2 equiv.) KOH and 0.15 mL concentrated NH 4 OH(aq) (2.17 mmol, 2 equiv.) in 200 mL ethanol. The solution was refluxed for 30h. The yellow solid product was collected by filtration and was wa shed with ethanol (5 10 mL) and diethyl ether (5 10 mL) and dried under vacuum overnight. Recrystallization from chloroform - methanol solutions yielded 0.11 0 g of bright yellow crystals (31% yield). 1 H NMR (500 MHz, methylene chloride - d 2 ) 8.71 (m, 4H), 8.67 (dt, J = 8.0, 1.1 Hz, 2H), 7.88 (dd, J = 7.2, 1.3 Hz, 2H), 7.87 7.84 (m, 2H), 7.36 (ddd, J = 7.5, 4.8, 1.2 Hz, 2H), 6.87 6.83 (m, 2H), 3.04 (s, 6H). Scheme 6 . 1 Synthesis of the dmaptrpy ligand. 179 6.3.1.2 . (( 4 - N,N - dimethylaminophenyl) - 2,2',6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) chloro ruthenium (II) chloride, [Ru(dmaptrpy)(dmabpy)Cl]Cl. [Ru(dmaptrpy)(dmabpy)Cl]Cl was synthesized following a procedure described in section 3.1.2 . 6.3.1.3 . ((4 - N ,N - dimethylaminophenyl ) - 2,2',6',2'' - terpyridine)(4,4´ - bis(N,N - dimethylamino) - 2,2´ - bipyridyl) ruthenium(II) ammine dihexafluo rophosphate, (5a). [Ru(dmaptrpy)(dmabpy)NH 3 ](PF 6 ) 2 was synthesized following the procedure described for 2a . Yield 88%. 1 H NMR (500 MHz, CD 2 Cl 2 - d 2 ) 8.1 Hz, 2H), 7.96 7.86 (m, 6H), 7.47 (d, J = 2.8 Hz, 1H), 7.36 (ddd, J = 7.2, 5.5, 1.3 Hz, 2H), 7.19 (d, J = 2.8 Hz, 1H), 7.13 (dd, J = 6.7, 2.8 Hz, 1H), 6.96 6.93 (m, 2H), 6.58 (d, J = 6.9 Hz, 1H), 6.12 (dd, J = 6.9, 2.8 Hz, 1H), 3.34 (s, 6H), 3.13 (s, 6H), 2.99 ( s, 6H), 2.12 (s, 3H). 180 6.3.2 . A mmonia Oxidation Using 5a as the Catalyst Unfortunately, the solubility of [Ru(dmaptrpy)(dmabpy)NH 3 ](PF 6 ) 2 , 5a , is very low in THF and DCM. Thus, the cyclic voltammetry experiments were conducted in NM. Addition of 4.05 10 - 3 M 5a to solutions of NH 3 in NM, shows a dramatic negative shift in the onset potential for the NH 3 oxidation current ( Figure 6.3 . 1 ). Figure 6.3 . 1 The CVs obtained in solutions: 4.05 10 - 3 M 5 a in NM (red), 4.05 mM 5 a in the presence of NH 3 in NM (black) and a solution of NH 3 When the catalytic currents were normalized for the scan rate, the dependence appears different than what was previously observed for other catalysts in THF ( Figure 6.3 . 2 , bottom ). It seems that while the onset of the catalytic current is moving to more favorable potentials with slower scan rates, the magnitude of the plateau current does not show a significant increase, which could be an indication of a different catalytic mechan ism controlling the currents. 181 Figure 6.3 . 2 Top : The CVs of the catalytic ammonia oxidation in NM in the presence of 4.05 10 - 3 M 5 a , with different scan rates. Bottom : The normalized currents shown for the same set of data. The magnitudes of the normalized currents remain relatively constant at different scan rates. The normalized currents obtained for the other complexes in NM, do not look similar to what were observed for 5a . For instance, the shi ft in the onset of the catalytic currents obtained for 3a in NM is also very negative compared to THF. However, the normalized catalytic currents behave similarly in both THF and NM ( Figure 6.3 . 3 ), suggesting that the difference in Figure 6.3 . 2 182 ( bottom ) and Figure 6.3 . 3 (bottom, right) are related to the chemistry of the catalysis when 5 a is used as the catalyst. Figure 6.3 . 3 Top : CVs for 1.68 10 - 3 M 3 a in NM in the absence (red) and presence (black) of NH 3 . Scan rate 0.1 Vs - 1 . Bottom: CVs of the catalytic ammonia oxidation in NM in the presence of 1.68 10 - 3 M 3a , with different scan rates (left). The normalized currents shown for the same set of data (r ight). 6.4 . Conclusions The catalytic activity of three other ruthenium polypyridyl amine complexes was evaluated by cyclic voltammetry. Catalyst 4a shows the lowest onset potential for the catalytic current. 183 Interestingly, the normalized catalytic currents f or complex 5a behave independently from the scan rate, suggesting that a different catalytic mechanism is in operation. Evolution of N 2 and H 2 for these complexes must be performed to further prove the catalytic activity as well as the efficiencies of the catalysts. 184 APPENDIX 185 A6.1 1 H NMR Spectra Figure A 6 .1. 1 1 H NMR spectrum of [Ru(Me 3 trpy)(dmabpy)Cl]Cl in DMSO - d 6 . Top: full spectrum, Bottom: blow - up of the aromatic region. Starred peaks are solvent residuals or impurities. (ppm) (ppm) 186 Figure A 6 .1. 2 1 H NMR spectrum of [Ru(Me 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 in DMSO - d 6 . Top: full spectrum, Bottom: blow - up of the aromatic region. Starred peaks are solvent/impurity related. Figure A 6 .1. 3 1 H NMR spectrum of [Ru( t Bu 3 trpy)(dmabpy)Cl]Cl in DMSO - d 6 . Top: full spectrum, Bottom: blow - up of the aromatic region. Starred peaks ar e solvent residuals. (ppm) 187 Figure A 6 .1. 4 1 H NMR spectrum of [Ru( t Bu 3 trpy)(dmabpy)NH 3 ](PF 6 ) 2 in acetonitrile - d 3 . Top: full spectrum, Bottom: blow - up of the aromatic region. Starred peaks are solvent related. (ppm) 188 Figure A 6 . 1.5 1 H NMR spectrum of (4 - N,N - dimethylaminophenyl) - 2,2',6',2'' - terpyridine in methylene chloride - d 2 . Starred peak is solvent residuals. (ppm) 189 Figure A 6 .1. 6 1 H NMR spectrum of [Ru(dmaptrpy)(dmabpy)NH 3 ](PF 6 ) 2 in CD 2 Cl 2 - d 2 . Starred peaks are solvent and impurity residuals. Some unknown impurities are not labeled. (ppm) 190 REFERENCES 191 REFERENCES (1) Hanan, G. S.; Wang, J. SYNLETT 2005 , No. 8, 1251 1254. 192 CHAPTER 7: CONCLUDING REMARKS A ND FUTURE DIRECTIONS 193 In this work, molecular catalysts based on ruthenium polypyridyl amine complexes were synthesized and studied for their catalytic activity towards homogeneous oxidation of NH 3 to N 2 in non - aqueous media under mild conditions. The catalytic behavior was confirmed via electrochemical studies and the products of the ammonia splitting, N 2, and H 2 , were quantified in the headspace of a solution of NH 3 in THF in the presence of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 after segments of controlled potential electrolysis. Those results confirmed that the onset of the oxidation of NH 3 in THF was reduced by approximately 300 mV upon addition of the catalyst and N 2 and H 2 w ere generated with ratios of approximately 1: 3, with faradaic efficiencies as high as 80%. Since the solution after the electrolysis, only contained the starting [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 catalyst after 180 min of electrolysis, it was concluded that the ca talyst regeneration was fully achieved and a closed catalytic cycle was performing. Isotope labeling experiments suggested that the catalysis is ensued through a hydrazine pathway as no evidence was found to support the formation of N 2 bridged intermediate s. The possibility of heterogeneous catalysis was ruled out after no N 2 and H 2 were detected in control rinse test experiments, as well as an examination of the surface of the electrode after the electrolysis did not show any depositions on the surface. N ext, understanding of the mechanism of the catalysis was pursued. The [Ru(trpy)(bpy)NH 3 ](PF 6 ) 3 intermediate was isolated and was further studied using a variety of techniques including 1 H NMR and electrochemical measurements. It was shown that the [Ru(trpy )(bpy)NH 3 ](PF 6 ) 3 was the first intermediate under catalytic conditions where the applied potential is positive to enable the one - electron oxidation of the Ru(II) center to Ru(III). Spectrophotochemical experiments revealed that this intermediate is unstabl e in the presence of a proton acceptor and undergoes deprotonation to give a [Ru(trpy)(bpy)NH 2 ](PF 6 ) 2 intermediate which was observed in mass spectrometry experiments. This second intermediate is very unstable and yielded half equivalent of 194 [Ru(trpy)(bpy)N H 3 ](PF 6 ) 2 as one product. Based on this observations and reported similar chemistry for analogous aqua complexes, a redox disproportionation pathway was considered to be occurring in which one equivalent of [Ru(trpy)(bpy)NH 2 ](PF 6 ) 2 disproportionated to 0.5 equivalent of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 and 0.5 equivalent of [Ru(trpy)(bpy)NH](PF 6 ) 2 , another intermediate that was not directly detected. However, since the hydrazine pathway for the catalytic cycle was previously established, the generation of the Ru (IV)NH intermediate was still envisioned to be a key step in the reaction mechanism. This was further studied in experiments which were conducted using an authentic Ru(II)N 2 H 4 complex. The cyclic voltammograms of solutions of the hydrazine complex were clo sely comparable to the CVs of solutions of [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 and stoichiometric concentrations of NH 3 in THF. Cyclic voltammograms of the saturated solution of NH 3 in THF in the presence of [Ru(trpy)(bpy)N 2 H 4 ](PF 6 ) 2 exhibited a catalytic current with an onset identical to that obtained when [Ru(trpy)(bpy)NH 3 ](PF 6 ) 2 was used as the catalyst. Putting all the results together, we proposed a mechanistic cycle for the homogenous catalytic ammonia oxidation as shown in Scheme 0 . 1 . 195 Scheme 0 . 1 The proposed catalytic cycle. The formation of the complexes highlighted in red has not yet been directly confirmed. While direct evidence supports the roles of species 2a , 2b , 2c, and 2e , the Ru(IV)NH intermediate ( 2d ) is still the missing puzzle piece. The very high reactivity of 2d and its high expected acidity would be one reason that its detection in the presence of a base was not successful. While it not yet clear that what possible reactions 2d might undergo in the presence of a non - coordinating ba se such as DBU, more experiments are necessary using solvents and based more innocent than the ones employed in this study (nitromethane and DBU). 196 Another approach that seems to be crucial to be taken is to move from hexafluorophosphate salts of the compl exes to more soluble counter anions. For instance, swapping the counter ion from PF 6 - to a more lipophilic anion such as Tetrakis(3,5 - bis(trifluoromethyl)phenyl)borate (BAr F ) would enable us to do more accurate NMR measurements in THF and more importantly, kinetic studies using stopped - flow techniques in DCM. The steps of the catalytic cycle past the formation of the hydrazine intermediate are also still unknown. Independent synthesis of [Ru(trpy)(bpy)N 2 ](PF 6 ) 2 would provide useful approaches to final steps of the cycle. Another important consideration is the possibility of proton - coupled electron transfer steps. Cundari et al. at the University of North Texas are preparing a computational article which discusses the mechanism of the catalytic pathway for ammonia oxidation based on the catalysts 1a and 2a introduced in this study. Those results would be invaluable in terms of setting a guideline towards the electron and proton transfer kinetics and would be complementary to our efforts on establishing a proposed catalytic cycle. With the catalytic cycle being decoded, the work could be directed towards developing more effective catalysts and eventually achieve lower overpotenti als as well as higher faradaic efficacies.