Interplay of Proton Transfer, Electron Transfer and Proton-Coupled Electron Transfer in Transition Metal Mediated Nitrogen Fixation Citation Matson, Benjamin David (2018) Interplay of Proton Transfer, Electron Transfer and Proton-Coupled Electron Transfer in Transition Metal Mediated Nitrogen Fixation. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/Z9MS3R0Z. https://resolver.caltech.edu/CaltechTHESIS:02232018-152758526 Abstract Mitigation of the hydrogen evolution reaction (HER) is a key challenge in selective small molecule reduction catalysis, including the nitrogen (N 2 ) reduction reactions (N 2 RR) using H + /e - currency. Here we explore, via DFT calculations, three iron model systems, P 3 E Fe (E = B, Si, C), known to mediate both N 2 RR and HER, but with different selectivity depending on the identity of the auxiliary ligand. It is shown that the respective efficiencies of these systems for N 2 RR trend with the predicted N–H bonds strengths of two putative hydrazido intermediates of the proposed catalytic cycle, P 3 E Fe(NNH 2 ) + and P 3 E Fe(NNH 2 ). Bimolecular proton-coupled electron transfer (PCET) from intermediates with weak N–H bonds is posited as a major source of H2 instead of more traditional scenarios that proceed via metal hydride intermediates and proton transfer/electron transfer (PT/ET) pathways. Studies on our most efficient molecular iron catalyst, [P 3 B Fe] + , reveal that the interaction of acid and reductant, Cp* 2 Co, is critical to achieve high efficiency for NH 3 , leading to the demonstration of electrocatalytic N 2 RR. Stoichiometric reactivity shows that Cp* 2 Co is required to observe productive N–H bond formation with anilinium triflate acids under catalytic conditions. A study of substituted anilinium triflate acids demonstrates a strong correlation between p K a and the efficiency for NH 3 , which DFT studies attribute to the kinetics and thermodynamics of Cp* 2 Co protonation. These results contribute to the growing body of evidence suggesting that metallocenes should be considered as more than single electron transfer reagents in the proton-coupled reduction of small molecule substrates and that ring-functionalized metallocenes, believed to be intermediates on the background HER pathway, can play a critical role in productive bond-forming steps. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: Inorganic Chemistry; Theoretical Chemistry; Computational Chemistry Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Peters, Jonas C. Group: Resnick Sustainability Institute Thesis Committee: Agapie, Theodor (chair) Miller, Thomas F. Hsieh-Wilson, Linda C. Peters, Jonas C. Defense Date: 9 February 2018 Non-Caltech Author Email: bdmatson (AT) gmail.com Record Number: CaltechTHESIS:02232018-152758526 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:02232018-152758526 DOI: 10.7907/Z9MS3R0Z Related URLs: URL URL Type Description http://dx.doi.org/10.1021/acscatal.7b03068 DOI Chapter 1 -- published material. http://dx.doi.org/10.1021/acscentsci.7b00014 DOI Chapter 2 -- published material. ORCID: Author ORCID Matson, Benjamin David 0000-0001-5733-0893 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 10731 Collection: CaltechTHESIS Deposited By: Benjamin Matson Deposited On: 09 Mar 2018 17:27 Last Modified: 08 Nov 2023 00:44 Thesis Files Preview PDF - Final Version See Usage Policy. 3MB Repository Staff Only: item control page

Interplay of Proton Transfer, Electron Transfer and Proton-Coupled Electron Transfer in Transition Metal Mediated Nitrogen Fixation Thesis by Benjamin David Matson In Partial Fulfillment of the Requirements for the degree of Doctor of Philosophy CALIFORNIA INSTITUTE OF TECHNOLOGY Pasadena, California 2018 (Defended February 9th, 2018) ii  2018 Benjamin David Matson iii ACKNOWLEDGEMENTS My time here at Caltech has been one of the most challenging yet rewarding periods of my life and there are so many people who have supported me on this journey. First off, I need to thank my advisor, Jonas, for taking me into his lab and allowing me to grow as a scientist. Jonas gave me the freedom to research new areas for our group and allowed me to really identify the types of chemistry I was truly passionate about. While the road has not always been smooth, Jonas was there to redirect me and get me to where I am today. Next I’d like to thank my coworkers in the Peters lab. Several students in front of me, John Anderson, Jon Rittle, Sid Creutz, and Dan Suess, paved the way for the N2 chemistry that has defined my thesis. During my time here, I have had the distinct pleasure of collaborating with many great chemists and friends. Most notably, Matthew Chalkley, Trevor Del Castillo and I have had the pleasure of working on several fruitful projects. They are truly exceptional chemists whose experimental expertise has guided me greatly along my Ph.D. In addition to N2 chemistry, I have had the opportunity to work on denitrification chemistry with Tanvi Ratani. Tanvi provided numerous experimental breakthroughs really took a difficult project to the next level and I am happy to have played a (small) part in that. In addition to scientific breakthroughs, Matt, Tanvi, Gaël Ung and Miles Johnson have played an invaluable role in ‘team morale’. Their confidence and support has helped me through many bad days in the lab and has kept to grounded and focused. Outside the lab, I have had the chance to do amazing things with truly great people. Paul Walton, Jenna Bush, Noah Duffy, Josh Buss, Mark Nesbit and Kareem Hannoun have iv joined me on many hiking and backpacking trips to see some of the most beautiful places in the country and have been amazing friends all along the way. Paul, Mark, Trevor and Beau Prichett also provided a great outlet in the form of our band, The 818s. Regularly described as “a participant” in some local festivals, my time with the band has been incredibly rewarding and tons of fun. To make up for the nerdy-ness of playing in a band, I have also had the chance to make great friends while playing Dungeons and Dragons at Caltech. Matt Davis, Richard Mossesso, Elizabeth Bernhardt, Nick Cowper, Noah and Trevor have made this quintessentially geeky game one of the most fun experiences of my time at Caltech. Finally, I would like to thank my family for their incredible support over the last 5 years. My parents and my sister, Jennifer, have always been there for me during this time. Their humor and love has kept me going through many of the difficult periods I’ve worked through over the years. I also want to thank my grandma who, unfortunately, did not get to see me make it all the way through my Ph.D. She was always so proud of me and provided constant support over the years. To all my family and friends, I thank them for sticking by my through the years. Even when I was preoccupied, too busy to keep in touch or just grumpy, they have always been by my side. I am truly lucky to be surrounded by so many wonderful people. v ABSTRACT Mitigation of the hydrogen evolution reaction (HER) is a key challenge in selective small molecule reduction catalysis, including the nitrogen (N2) reduction reactions (N2RR) using H+/e- currency. Here we explore, via DFT calculations, three iron model systems, P3EFe (E = B, Si, C), known to mediate both N2RR and HER, but with different selectivity depending on the identity of the auxiliary ligand. It is shown that the respective efficiencies of these systems for N2RR trend with the predicted N–H bonds strengths of two putative hydrazido intermediates of the proposed catalytic cycle, P3EFe(NNH2)+ and P3EFe(NNH2). Bimolecular proton-coupled electron transfer (PCET) from intermediates with weak N–H bonds is posited as a major source of H2 instead of more traditional scenarios that proceed via metal hydride intermediates and proton transfer/electron transfer (PT/ET) pathways. Studies on our most efficient molecular iron catalyst, [P3BFe]+, reveal that the interaction of acid and reductant, Cp*2Co, is critical to achieve high efficiency for NH3, leading to the demonstration of electrocatalytic N2RR. Stoichiometric reactivity shows that Cp*2Co is required to observe productive N–H bond formation with anilinium triflate acids under catalytic conditions. A study of substituted anilinium triflate acids demonstrates a strong correlation between pKa and the efficiency for NH3, which DFT studies attribute to the kinetics and thermodynamics of Cp*2Co protonation. These results contribute to the growing body of evidence suggesting that metallocenes should be considered as more than single electron transfer reagents in the proton-coupled reduction of small molecule substrates and that ring-functionalized metallocenes, believed to be intermediates on the background HER pathway, can play a critical role in productive bond-forming steps. vi PUBLISHED CONTENT AND CONTRIBUTIONS 1. Matson, B. D.; Peters, J. C. Fe-mediated HER vs N2RR: Factors that Determine Selectivity in P3EFe(N2) (E = B, Si, C) Catalyst Model Systems. ACS Catalysis, 2018, 8, 1448-1455. https://pubs.acs.org/doi/abs/10.1021/acscatal.7b03068 I performed all calculations and wrote the manuscript for this work. 2. Chalkley, M. J.; Del Castillo, T. J.; Matson, B. D.; Roddy, J. P.; Peters, J. C. Catalytic N2-to-NH3 Conversion by Fe at Lower Driving Force: A Proposed Role for MetalloceneMediated PCET. ACS Central Science, 2017, 3, 217-223. http://pubs.acs.org/doi/full/10.1021/acscentsci.7b00014 I performed all calculations, assisted in the analysis of catalytic and spectroscopic data, and wrote portions of this manuscript vii TABLE OF CONTENTS Acknowledgements............................................................................................................iii Abstract…………...............................................................................................................v Published Content and Contributions..................................................................................vi Table of Contents…………...............................................................................................vii Detailed Table of Contents................................................................................................vii List of Figures...................................................................................................................xiii List of Tables....................................................................................................................xvii List of Abbreviations......................................................................................................xx Chapter 1. Introduction.........................................................................................................1 Chapter 2. Fe-mediated HER vs N2RR: Factors that Determine Selectivity in P3EFe(N2) (E = B, Si, C) Catalyst Model Systems..............................................................14 Chapter 3. Catalytic N2-to-NH3 Conversion by Fe at Lower Driving Force: A Propose Role for Metallocene-Mediated PCET......................................................................43 Chapter 4. Fe-Mediated Nitrogen Fixation with a Metallocene Mediator: Exploring pKa Effects and Demonstrating Electrocatalysis.....................................................63 Chapter 5. Predicting BDFE Values Using DFT.…………………………………… ...…89 Appendix 1. Supplementary Data for Chapter 2...............................................................100 Appendix 2. Supplementary Data for Chapter 3……………...........................................115 Appendix 3. Supplementary Data for Chapter 4…………………………………….… viii DETAILED TABLE OF CONTENTS Chapter 1. Introduction ........................................................................................................1 1.1. Opening Remarks..............................................................................................2 1.2. N2RR Using Soluble Transition Metal Catalysts...............................................5 1.3. Selectivity for N2RR vs HER.............................................................................7 1.4. Proton-Coupled Electron Transfer in N2RR......................................................9 1.5. Method for Predicting E-H Bonds Strengths Using DFT………………...…..11 1.6. Conclusions……………………………………………………...…………..11 1.5. Cited References..............................................................................................12 Chapter 2. Fe-mediated HER vs N2RR: Factors that Determine Selectivity in P3EFe(N2) (E = B, Si, C) Catalyst Model Systems.............................................................14 2.1. Introduction ....................................................................................................15 2.2. Computational Methods..................................................................................17 2.3 Results and Discussion.....................................................................................19 2.3.1. DFT Support for Slow Fe Protonation and Fe–NxHy Formation.......21 2.3.2 Calculation of BDFEN-H Values for Fe–NxHy Intermediates.............24 2.3.3. Calculated Reduction Kinetics of P3EFe(NNH2)+.............................27 2.3.4. Calculated PCET Reactions..............................................................29 2.3.5. Wiberg Bond Indices of P3BFe(NxHy) Species…….....…………….33 ix 2.3. Conclusions.....................................................................................................37 2.5. References.......................................................................................................38 Chapter 3. Catalytic N2-to-NH3 Conversion by Fe at Lower Driving Force: A Proposed Role for Metallocene-Mediated PCET.............................................................43 3.1. Introduction ....................................................................................................44 3.2. Results ............................................................................................................46 3.2.1. Catalysis Using [P3BFe][BArF4], Cp*2Co and [RnNH(4-n)][OTf].....46 3.2.2. Fe Speciation Under Turnover Conditions.......................................49 3.2.3. DFT Predicted pKa’s and BDFEs......................................................51 3.3. Discussion.......................................................................................................55 3.4. Conclusions.....................................................................................................56 3.5.References........................................................................................................57 Chapter 4. Chapter 4. Fe-Mediated Nitrogen Fixation with a Metallocene Mediator: Exploring pKa Effects and Demonstrating Electrocatalysis…………………..63 4.1. Introduction………………………………………………………………….64 4.2. Results and Discussion………………………………………………………67 4.2.1. pKa Studies…………………………………………………….…..67 4.2.2. Computational Studies…………………………...………………..72 4.2.2. Electrolysis Studies…………………………………………..……76 4.3. Conclusion…………………………………………………………….……..83 x 4.4. References…………………………………………………………….……..92 Chapter 5. Predicting BDFE Values Using DFT……………………………………..…..89 5.1. Introduction………………………………………………………………….90 5.2. Calibration of Gas-Phase BDFEE-H Values……………………………….….91 5.3. Solution Phase BDFEE-H Values………………………………..……..……96 5.4. Transition Metal Bound BDFEE-H Values………………………..…...…….97 5.5. Conclusions…………………………………………………….....………..98 5.6. References……………………………………………………….....………98 Appendix 1. Supplementary Data for Chapter 2...............................................................100 A1.1. General Computational Details…………………………………..….……101 A1.2. Fe-H Bond Formation………………………………..…………..……….102 A1.3. BDFE Calculations…………………………………………..…...………102 A1.4. Approximation of P3EFe(NNHy) Radius………………………….....……104 A1.5. Calculated Reorganization Energies…………………………………..….105 A1.6. Determination of the Work Function………………………………..……107 A1.7. Summary of Wiberg Indices………………………………….....………..108 A1.8. Comparison of Calculated to Known Experimental Values…………..…112 A1.9. References……………………………………......………..……………..113 Appendix 2. Supplementary Data for Chapter 3...............................................................115 xi A2.1. Experimental Details..................................................................................116 A2.1.1. General Considerations...............................................................116 A2.1.2. Physical Methods.........................................................................116 A2.1.3. Mossbauer Spectroscopy.............................................................117 A2.1.4. Ammonia and Hydrazine Quantification.....................................117 A2.1.5. EPR Spectrscopy.........................................................................119 A2.1.6. Computational Methods..............................................................119 A2.2. Synthetic Details.........................................................................................120 A2.2.1. General Procedure for the Synthesis of the Acids........................120 A2.3. Ammonia Production and Quantification Studies.......................................120 A2.3.1. Standard NH3 Generation Reaction Procedure............................120 A2.3.2. Ammonia Production Studies with [Ph215NH2][OTf]..................123 A2.4. NH3 Generation Reaction With Periodic Substrate Reloading....................124 A2.5. Time-Resolved H2 Quantification of Background Acid and Cp*2Co..........125 A2.6. Mossbauer Spectra......................................................................................126 A2.6.1. Rapid-Freeze Mossbauer of P3BFe+.............................................126 A2.6.2. Rapid-Freeze Mossbauer of P3BFe+ and Reductant......................127 A2.6.3. Details of Individual RFQ Mossbauer Spectra.............................130 A2.7. EPR Spectra................................................................................................136 A2.7.1. Preparation of Rapid-Freeze-Quench EPR of Catalytic Mixture.136 A2.7.2. Preparation of the Reaction of P3BFe+ with Reductant.................137 A2.7.3. Preparation of Cp*2Co, P3BFe+ and P3BFe(N2)- Samples..............138 xii A2.8. Details on DFT Estimates of pKa and BDEN-H............................................143 A2.8.1. Computational Estimate of pKa in Et2O.......................................143 A2.8.2. Computational Estimates of BDEN-H...........................................143 A2.9. References..................................................................................................146 Appendix 3. Supplementary Data for Chapter 4...............................................................157 A3.1. Experimental Details……………………………………………………..149 A3.1.1. General Considerations…………………………………….......149 A3.1.2. Gas Chromotography……………………………………..…….150 A3.1.3. Mossbauer Spectroscopy…………………………………….…150 A3.1.4. Ammonia Quantification…………………………………….…151 A3.1.5. Computational Methods…………………………………….….151 A3.2. Synthetic Details………………………………………………….………152 A3.2.1. General Procedure for the Synthesis of Anilinium Triflates…….152 A3.3. Ammonia Generation Details……………………………………….……153 A3.3.1. Standard NH3 Generation Procedure………………………...….153 A3.4. H2 Monitoring Details…………………………………………………….155 A3.4.1. Standard Background Reaction Procedure…………..…………155 A3.4.2. H2 Evolution Kinetics…………………………………………..156 A3.5. Mossbauer Spectroscopy……………………………………………..…..159 A3.5.1. General Procedure for Freeze-Quench Mossbauer……………..159 A3.6. EPR Spectroscopy……………………………………..…………………161 xiii A3.6.1. General Procedure for EPR Spectroscopy……………..………161 A3.6.2. Comment of Stoichiometric Reactivity……………….…..…….162 A3.7. Acid Quench of P3BFeN2-……………………………………..…………..165 A3.7.1. Standard Acid Quench Procedure………………………...…….165 A3.8. Solubility Measurement…………………………………………..………166 A3.8.1. Procedure for Measuring Solubility of Cp*2Co……………...…166 A3.8.2. Procedure for Measuring Solubility of [2,6-ClPhNH3][OT]…...…167 A3.9. Controlled Potential Electrolysis Data……………………………...…….168 A3.10. Computational Details…………………………………………..………174 A3.10.1 Calculation of Acid Dissociation Constants……………...…….174 A3.10.2. Determination of PT, ET and PCET Barriers……………...…..174 A3.10.3. BDFE Calculations…………………………………..………..177 A3.11. X-Ray Photoelectro Spectroscopy Details……………………...……….180 A3.12. pKa Determination Strategy……………………………………………..185 A3.13. References…………………………………………………………..…..186 xiv LIST OF FIGURES Chapter 1. Introduction.........................................................................................................1 Figure 1.1 FeMoco and Biological Nitrogenase……………….…………………..3 Figure 1.2. Overview of P3EFe Catalysts……………..……………………………6 Figure 1.3. Overview of P3EFe Catalyzed HER……………………………………8 Figure 1.4. Proposed Structure of Cp*Co(η4-C5Me5H)+……………………….…..8 Chapter 2. Fe-mediated HER vs N2RR: Factors that Determine Selectivity in P3EFe(N2) (E = B, Si, C) Catalyst Model Systems..............................................................14 Figure 2.1. Schematic Depiction of N2RR/HER Iron Catalysts Studied.................16 Figure 2.2. Formation of P3EFe(NNH2)+.................................................................20 Figure 2.3. BDFEN-H Values of Selected Fe(NxHy) Species....................................26 Figure 2.4. Calculated Free-Energy Changes for Selected PCET Reactions..........30 Figure 2.5. Mechanistic Overview for N2RR and HER..........................................32 Figure 2.6. Wiberg Bond Indices for Fe(NxHy) Species..........................................37 Chapter 3. Catalytic N2-to-NH3 Conversion by Fe at Lower Driving Force: A Proposed Role for Metallocene-Mediated PCET.............................................................43 Figure 3.1. Summary of conditions used for N2-to-NH3 conversion by P3BFe+.......45 Figure 3.2. Mössbauer of a freeze-quenched catalytic reaction..............................51 Figure 3.3. Calculated free-energy changes for the protonation of Cp*2Co............53 xv Chapter 4. Chapter 4. Fe-Mediated Nitrogen Fixation with a Metallocene Mediator: Exploring pKa Effects and Demonstrating Electrocatalysis…………………..63 Figure 4.1. Electron Efficiencies for P3BFe and Nitrogenase…….……………….68 Figure 4.2. Kinetics and Thermodynamics of Cp*2Co Protonation………………74 Figure 4.3. Kinetics and Thermodynamics of P3BFe(NNH2) Formation……...…..75 Figure 4.4. Cyclic Voltammograms……………………………….……………..77 Chapter 5. Predicting BDFE Values Using DFT…………………………………...…….89 Figure 5.1. Plots of BDFEcalc vs BDFElit for Known Gas-Phase Values…………..92 Figure 5.2. Normal Probability Plots for Errors in Table 5.1…………………..95 Appendix 1. Supplementary Data for Chapter 4……………………………….……..…100 Figure A1.1. Structures for Fe-H Formation……………………………...……..102 Figure A1.2. Plot of Calculated vs Literature BDFE values……………………..103 Appendix 2. Supplementary Data for Chapter 3...............................................................115 Figure A2.1. 1H NMR Spectrum of [14NH4][Cl]..................................................124 Figure A2.2. Mossbauer Spectrum Collected on [P3B]57Fe...................................129 Figure A2.3. Mossbauer of a 5 Min Reaction Quench: Cp*2Co + P3BFe+.............130 Figure A2.4. Mossbauer of a 5 Min Reaction Quench: Cp*2Cr + P3BFe+..............132 Figure A2.5. Mossbauer of a 5 Min Catalytic Reaction Quench...........................133 xvi Figure A2.6. Mossbauer of a 30 Min Catalytic Reaction Quench.........................135 Figure A2.7. EPR Spectrum of [P3BFe(N2)][Na(12-crown-4)2]............................138 Figure A2.8. EPR Spectrum of [P3BFe][BArF4]....................................................139 Figure A2.9. EPR Spectrum of Cp*2Co................................................................140 Figure A2.10. EPR Spectrum of Cp*2Co + P3BFe+ + Ph2NH2...............................141 Figure A2.11. EPR Spectrum of Cp*2Co + P3BFe+...............................................142 Appendix 3. Supplementary Data for Chapter 4………………………………………...157 Figure A3.1. Catalyzed/Uncatalyzed H2 Formation from [2,6-ClPhNH3][OTf]…..159 Figure A3.2. Freeze Quench Mossbauer Spectrum for Catalytic Reaction…….160 Figure A3.3. EPR Spectrum of a Freeze-Quenched Catalytic Mixture………….163 Figure A3.4. EPR Spectrum of [P3BFeN2][Na(12-c-4)2] + 2,6-ClPhNH3+…...……164 Figure A3.5. BDFEcalc and BDFElit Plotted with Best Fit Line……………..……178 Figure A3.6. XPS of Unexposed Surface……………………………………..…182 Figure A3.7. XPS of Exposed Surface…………………………………..………185 xvii LIST OF TABLES Chapter 2. Fe-mediated HER vs N2RR: Factors that Determine Selectivity in P3EFe(N2) (E = B, Si, C) Catalyst Model Systems.............................................................14 Table 2.1. Calculated parameters for P3EFe(NNH2)+  P3EFe(NNH2)...................29 Chapter 3. Catalytic N2-to-NH3 Conversion by Fe at Lower Driving Force: A Proposed Role for Metallocene-Mediated PCET.............................................................43 Table 3.1. N2-to-NH3 Conversion with P3EM Complexes (M = Fe, Co)…….........48 Table 3.2. Calculated pKa Values and BDEs of Selected Species...........................54 Chapter 4. Chapter 4. Fe-Mediated Nitrogen Fixation with a Metallocene Mediator: Exploring pKa Effects and Demonstrating Electrocatalysis…………………..63 Table 4.1. pKa Values and Catalyst Efficiencies………………………….………69 Table 4.2. Yields and Faradaic Efficiencies of NH3 From CPE Experiments….....79 Chapter 5. Predicting BDFE Values Using DFT…………………………………………89 Table 5.1. Errors Obtained from Figure 5.1………………………………………93 Table 5.2. Solvent Corrections for Organic BDFEE-H………………………….96 Table 5.3. Solvent Corrections for Transition Metal BDFEE-H………………….97 xviii Appendix 1. Supplementary Data for Chapter 4………………………………..……..100 Table A1.1. Summary of BDFEs Used for Calibration…………………….…..103 Table A1.2. Volume and Calculated Radius of P3EFe Species…………...…….104 Table A1.3. Summary of Calculated Reorganization Energies………….…..….106 Table A1.4. Wiberg Bond Indices for P3EFe(N2) Complexes…………..……….108 Table A1.5. Wiberg Bond Indices for P3EFe(NNH) Complexes………..……….108 Table A1.6. Wiberg Bond Indices for P3EFe(NNH2) Complexes…......……...….109 Table A1.7. Wiberg Bond Indices for P3EFe(N(4-OMePh))…………………….110 Table A1.8. Wiberg Bond Indices for C2H4 and C2H5…………………………..111 Table A1.9. Comparison of Calculate to Known Experimental Values………..112 Appendix 2. Supplementary Data for Chapter 3...............................................................115 Table A2.1. NH3 Generation using P3BFe+...........................................................121 Table A2.2. NH3 Generation Using P3SiFeN2........................................................122 Table A2.3. NH3 Generation Using P3BCoN2−......................................................122 Table A2.4. NH3 Generation Using P3SiCoN2.......................................................123 Table A2.5. NH3 Generation Using P3BFe+ With Reloading.................................125 Table A2.6. Time-Resolved Background H2 Quantification................................126 Table A2.7. Fit Parameters for Figure A2.2..........................................................129 Table A2.8. Fit Parameters for Figure A2.3..........................................................131 Table A2.9. Fit Parameters for Figure A2.4..........................................................133 Table A2.10. Fit Parameters for Figure A2.5........................................................134 xix Table A2.11. Fit Parameters for Figure A2.6........................................................136 Table A2.12. Comparison of Mossbauer Parameters............................................142 Table A2.13. Calculated H Values and Literature BDE Values.........................144 Table A2.14. Calculated Entropy for Selected XH and H● Species......................145 Appendix 3. Supplementary Data for Chapter 4…………………………………..…….157 Table A3.1. NMR Quantification for NH3 Generation using P3BFe+…..………..154 Table A3.2. Data for Background H2 Quantification…………………..………..155 Table A3.3. Time Points for Catalyzed H2 Evolution……………………...……157 Table A3.4. Time Points for Uncatalyzed H2 Evolution………………...………157 Table A3.5. Simulation Parameters for Mossbauer in Figure A3.2…………..….161 Table A3.6. Comparative NH3 and H2 Yields for OTf and BArF4 Acid……….…166 Table A3.7. Controlled Potential Electrolysis Data……………………..………171 Table A3.8. Overview of Parameters Used to Calculate Kinetic Barriers….........176 Table A3.9. Data Used to Generate the Plot in Figure A3.5…………….……….178 xx LIST OF ABBREVIATIONS atm Atmosphere Avg Average BArF4 B(3,5-C6H3(CF3)2)4 BDE Bond Dissociation Enthalpy BDFE Bond Dissociation Free-Energy C Constant ca circa calc Calculated CN Cyanide CNHx Generic CN-derived ligand with x H atoms Cp Cyclopentadienyl Cp* Pentamethylcyclopentadienyl CV cyclic voltammogram DFT Density Functional Theory DME 1,2-dimethoxyethane e- Electron EPR eq Electron paramagnetic resonance Equation equiv Equivalents Eo Fc+/0 Reduction potential Ferrocene/Ferrocenium couple xxi FeMoco Iron-Molybdenum cofactor GC gas chromatography G Gauss g Gram gn Electron g-factor GC Gas chromatography GHz Gigahertz H Enthalpy HIPT hexa-isopropyl-terphenyl HOMO Highest-Occupied Molecular Orbital Hz Hertz iPr isopropyl IR Infrared K Kelvin Keq Equilibrium constant L Generic neutral dative ligand LUMO Lowest-Unoccupied Molecular Orbital kcal Kilocalorie M Concentration in molarity max Maximum Me Methyl mg Milligram xxii MHz Megahertz mL Milliliter mM Millimolar mm Millimeter mV Millivolt mmol Millimole MO Molecular orbital mol Mole n generic number nm nanometer NMR Nuclear magnetic resonance NxHy Generic nitrogenous ligand with x N atoms and y H atoms NNHx o Generic nitrogenous ligand with 2 N atoms and x H atoms ortho OTf Triflate (OSO2CF3) Ph Phenyl pKa Acid dissociation constant R Generic organic group RT Room temperature S Entropy P3B (o-iPr2P(C6H4))3BP3C (o-iPr2P(C6H4))3C- xxiii P3Si (o-iPr2P(C6H4))3SiTBA Tetra-n-butyl ammonium tBu tert-Butyl TEMPO 2,2,6,6-tetramethylpiperidine-N-oxide THF Tetrahydrofuran TMS Trimethylsilyl UV Ultraviolot V Volt XRD X-ray diffraction Nα Proximal nitrogen atom of a bound N2 ligand Nβ Terminal nitrogen atom of a bound N2 ligand γ High frequency electromagnetic radiation Q Mössbauer quadrupole splitting δ chemical shift or Mössbauer isomer shift ° Degree °C Degrees Celsius ε Extinction coefficient in units of M-1cm-1 ηx Hapticity of order x λ Wavelength λmax Wavelength of local maximum intensity μ Bridging μA Microamps xxiv νxy Vibrational frequency between atoms x and y Σ Summation σ Sigma symmetry orbital or interaction σ* Sigma symmetry antibonding interaction π Pi symmetry orbital or interaction π* Pi symmetry antibonding interaction τ1/2 Half-life Å Angstrom 12-C-4 12-crown-4 1 H Hydrogen-1 2 H Hydrogen-2 11 B Boron-11 C Carbon-13 13 15 N Nitrogen-15 31 P Phosphorus-31 2-MeTHF 2-Methyl-tetrahydrofuran def2-xxx Basis sets for DFT TPSS, etc. DFT functional 1 Chapter 1. Introduction 2 1.1. Opening Remarks The global nitrogen cycle is a crucial biogeochemical cycle and underpins much of life on Earth.1 In particular, the conversion of dinitrogen (N2) to ammonia (NH3) provides a means by which atmospheric N2, a relatively inert gas that makes up 78% of Earth’s atmosphere, can be converted to a bioavailable form (NH3, NO2- and NO3-).1 The fixation of N2 can occur via non-biological natural processes, as in the splitting of N2 by lightning, or via biological and industrial processes.1-3 The latter two processes have garnered significant interest in chemistry and biology.4 The industrial process by which N2 is fixed, referred to as the Haber-Bosch process, is performed on a massive scale and the drastic increase in the human population in the 20th century is often credited to its use in fertilizer production.1 While the HaberBosch process continues to be a major source of bioavailable nitrogen globally, accounting for up to 80% of nitrogen in the human body, the high temperatures and pressures required has led to interest in developing a more energy efficient process.1 3 Figure 1.1. (Top) Biological nitrogen fixation is catalyzed by the iron-molybdenum cofactor (FeMoco) in nitrogenase enzymes. (Bottom) Possible mechanisms by which a single Fe site can catalyze the reaction. Biological nitrogen fixation has provided ideal inspiration in the development synthetic N2 fixation processes.4 Studies on the nitrogenase enzyme have revealed three major subtypes: iron-molybdenum nitrogenase, vanadium-iron nitrogenase, and all iron nitrogenase.5 The most well studied of these, the molybdenum-iron nitrogenase, contains the iron-molybdenum cofactor (FeMoco; Figure 1.1). In recent years, structure-function studies on FeMoco have led to increased interest in the mechanism by which N2 can be reduced at a single Fe center.7 At a single metal site, two limiting mechanisms have been proposed, referred to as the distal pathway and the alternating pathway (Figure 1.1, bottom). The distal mechanism is characterized by the early release of the first equivalent of NH3, with concomitant formation of a terminal nitride intermediate, Fe(N). In contrast, 4 the alternating mechanism is associated with the late release of the first equivalent of NH3 and is characterized by the formation of diazene, Fe(NHNH) and hydrazine, Fe(NH2NH2) intermediates.7 In addition to these limiting mechanisms, hybrid mechanisms are similarly plausible. A notable example of a hybrid mechanism is the so-called ‘distal-to-alternating’ mechanism, in which Fe(NH2NH2) and/or Fe(NHNH2) are formed without initial formation of an Fe(NHNH) species. In the study of synthetic systems for NH3 formation, designated in this work as the nitrogen reduction reaction (N2RR), the hydrogen evolution reaction (HER) competes for acid and reductant equivalents. Suppression of this unproductive pathway is crucial to increasing the efficiency of a system for N2RR, generally reported as the %NH3 per e− equivalent. As such, elucidation of the mechanistic interplay between N2RR and HER is perhaps the most important factor in the increasing the efficacy of N2RR catalyst systems. Central to mechanistic studies into multi-proton, multi-electron catalyst systems, including N2RR, is the interchange between proton transfer (PT), electron transfer (ET), proton-coupled electron transfer (PCET) and hydride transfer (HT). Accordingly, mechanistic steps for N2RR can be characterized by PT and ET as separate kinetic steps or by the concerted transfer of a proton and one or two electrons (PCET or HT, respectively). Given that N2RR necessarily involves the net addition of 6 H-atom (H●) equivalents, the synchronous and/or asynchronous delivery of H+ and e− to the N2 substrate produces drastic changes in the intermediates produced in the process and, thus, the selectivity of the system. 5 1.2. N2RR Using Soluble Transition Metal Catalysts In this thesis, research efforts aim at elucidating the role of PT, ET and PCET in N2RR are presented. Particular attention is paid to the impact these mechanistic steps have on the overall selectivity of the systems. Over the last decade, our group, and others, have uncovered several methods for transition metal mediated N2RR via alteration of the catalyst, as well as the stoichiometric reagents (acid and reductant).10 The research presented in this thesis is based on the most diverse class of catalysts, namely the P3EFe (E = B, C or Si) based systems (Figure 1.2). 6 Figure 1.2. (Top) Overview of P3EFe catalysts, acids and reductants discussed in this thesis. (Bottom) Proposed mechanisms for each set of reagents (vida infra). In initial work on N2RR by P3EFe(N2)− based catalysis, all three catalysts studied (E = B, C or Si) were shown to be formally catalytic upon the addition of strong reductant, KC8 (Eo < −3.0 V vs Fc+/0; Figure 1.2), and acid, [H(OEt2)2][BArF4] (HBArF4, BArF4 = tetrakis-(3,5-bis(trifluoromethyl)phenyl)borate; pKa < 0.0; Figure 1.2).4d The P3BFe system was found to be the most efficient catalyst, with efficiencies for NH3 up to 45 ± 3%. While still catalytic, the P3CFe and P3SiFe systems were found to be less efficient for N2RR, with efficiencies up to 36 ± 6% and 5 ± 3%, respectively. More recent studies have revealed that a related catalyst, [P3BFe][BArF4], can achieve much higher efficiencies for N2RR (up to 77%) using a decamethylcobaltocene reductant (Cp*2Co; Eo = −2.0 V vs Fc+/0) and a 2,5-dichloroanilinium triflate acid ([2,5-ClPhNH3][OTf]; pKa = 4.1).4e Of particular interest to the studies presented herein is the role that the acid and reductant choice play in the plausibility of PT, ET and PCET based mechanisms. Notably, the KC8/HBArF4 cocktail should be associated with increased favorability of asynchronous PT and ET steps, given the strength of these reagents and the lack of stable intermediates formed upon reaction between them. In contrast, we’ve shown that the use of Cp*2Co/[RPhNH3][OTf] is associated with an increased role of PCET mechanisms due to the formation of a highly active PCET reagent, a protonated Cp*2Co species ([Cp*Co(η4C5Me5H)]+), under catalyst conditions. 7 1.3. Selectivity for N2RR vs HER In chapter 2, the interplay of N2RR and HER in catalysis using KC8 and HBArF4 is explored in depth.11 The use of KC8 and HBArF4 provides the most complete comparison between P3EFe catalysts discussed, as all three systems (E = B, C or Si) are formally catalytic, but with drastically different efficiencies for N2RR and HER. Beyond the differing N2RR efficiencies discussed previously, the efficiencies for HER in P3BFe (44%) and P3SiFe (88%) have been shown to trend in the opposite direction, consistent with HER being a major source of N2RR efficiency loss. Investigating this difference in N2RR and HER efficiencies via density functional theory (DFT) point to the P3EFe(NNH2)+/0 intermediates, and their stability, as key players. Despite the relative lack of PCET reactivity from the KC8/HBArF4 interaction directly, we invoke bimolecular PCET between reactive P3EFe(NxHy) species as the major source of HER on the these scaffolds (Figure 1.3). In particular, while the Fe(NNH2) is predicted to be the only species capable of bimolecular HER on the P3B scaffold, the cationic species, Fe(NNH2)+, on P3C/Si are predicted to be long-lived and highly reactive for bimolecular H2 release (Figure 1.3). 8 Figure 1.3. Overview of predicted HER mechanisms on P3EFe systems, as discussed in chapter 2. In addition to the Fe(NNH2)+/0 species, predicted to be major sources of HER under catalytic conditions (using KC8/HBArF4), the diazenido species, Fe(NNH), are predicted to be extremely reactive intermediates and are invoked as sources of H2 in the stoichiometric oxidation of P3EFe(N2) species.4a,f,g In addition, the bimolecular coupling of Fe(NNH) species is predicted to become relevant in catalytic systems in which protonation reactions are slowed, i.e. with the use of anilinium triflate acids. The solubility of the anilinium acids 9 in Et2O is quite low and, accordingly, is associated with slower PT under turnover conditions. As a result, PCET reactivity from the Cp*2Co/[RPhNH3][OTf] combination is therefore invoked as the source of increased N2RR efficiencies using these reagents. 1.4. Proton-Coupled Electron Transfer in N2RR Chapters 3 and 4 are focused on our discovery that a protonated decamethylcobaltocene species, [Cp*Co(η4-C5Me5H)]+, likely serves as PCET reagent under conditions in which Cp*2Co serves the reductant and [RPhNH3][OTf] serves as the acid source.4e More specifically, chapter 3 outlines our initial discovery that N2RR using P3BFe can be accomplished with increased efficiency using reagents which substantially lower the overpotential of the system (ca. 100 kcal/mol lower). To rationalize this counterintuitive observation, we invoke the formation of a protonated metallocene species, [Cp*Co(η4-C5Me5H)]+, and suggest that it serves as a PCET reagent under turnover conditions (Figure 1.4). 10 Figure 1.4. (top) Proposed structure of the [Cp*Co(η4-C5Me5H)]+ intermediate discussed in chapters 3 and 4. (bottom) Proposed mechanism in which [Cp*Co(η4-C5Me5H)]+ acts as a PCET reagent. In chapter 4, we expand upon this discovery via a thorough study on the effect of pKa on the efficiency for N2RR vs HER. By alteration of the electronics of the aryl ring in [RPhNH3][OTf] acids, catalytic turnover is achieved over 8 pKa units using Cp*2Co as the reductant. Further, the efficiency for N2RR is shown to increase with acid strength and HER is shown to concomitantly decrease. Using DFT, we suggest that the rate of Cp*2Co protonation as the likely source of this pKa effect. As the final piece of evidence for the potential role of [Cp*Co(η4-C5Me5H)]+ in N2RR, catalytic yields of NH3 are achieved 11 electrochemically using a P3BFe catalyst, diphenylammonium acid [Ph2NH2]+ and a Cp*2Co co-catalyst. Under electrolytic conditions we suggest that Cp*2Co may be acting as a PCET mediator, as well as an ET shuttle. 1.5. Method for Predicting E-H Bonds Strengths Using DFT The final chapter of this thesis discusses efforts to increase the accuracy of DFT predicted bond dissociation free-energies (BDFEE-H) of transition metal bound E–H bonds. The importance of BDFEE-H predictions are apparent in all chapters in this work and in chapter 5, the accuracy of these calculations are discussed as a function of DFT functional and solvent of interest. Most notably it is reported that all functionals tested show similar accuracy for gas-phase BDFEE-H predictions, assuming proper calibration with known literature values. It is further shown that reproduction of solvated BDFEE-H is significantly less accurate than gas-phase values, but that values in several common organic solvents, DMSO, MeCN and C6H6 can be reproduced with acceptable accuracy. 1.6. Conclusions In sum, the following chapters will outline research efforts aimed at elucidating the mechanism by which P3EFe (E = B, C or Si) catalyzes HER and N2RR. Particular attention is paid to how the choice of reductant and acid source can dictate the rates of PT, ET and PCET. Notably, bimolecular coupling of reactive P3EFe intermediates via PCET is presented as a likely mechanism for competing HER. Further, evidence for the formation of a reactive PCET reagent, Cp*Co(η4-C5Me5H), is presented as a background HER 12 intermediate that can be intercepted by P3BFe species to undergo productive N–H bond formation. The following chapters rely heavily on the use of DFT for efficient prediction of BDFEE-H values, pKa values, Eo values and the kinetics associated with each of these reaction. The calculation of known values highlights the accuracy and utility of these calculations’ predictive power. Thus, on a larger scale the research presented highlights the value of DFT calculations in elucidating the mechanism by which multi-proton, multielectron reactions precede as well as key factors that can be used to inform the design of future, more efficient systems. 1.7. References 1 Smil, V. Enriching the Earth. Cambridge: MIT Press, 2001. 2 Drapcho, D. L.; Sisterson, D.; Kumar, R. Atmospheric Environment (1967) 1983, 729– 734. 3 (a) Howard, J. B.; Rees, D. C. Proc. Nat. Acad. Sci. 2006, 103, 17088-17093. (b) Howard, J. B.; Rees, D. C. Chem. Rev. 1996, 96, 2965−2982. (c) Hoffman, B. M.; Lukoyanov, D.; Yang, Z.-Y.; Dean, D. R.; Seefeldt, L. C. Chem. Rev. 2014, 114, 4041−4062. 4 For pertinent Fe systems: (a) Anderson, J. S.; Rittle, J.; Peters, J. C. Nature 2013, 501, 84−87. (b) Creutz, S. E.; Peters, J. C. J. Am. Chem. Soc. 2014, 136, 1105−1115. (c) Ung, G.; Peters, J. C. Angew. Chem., Int. Ed. 2015, 54, 532−535. (d) Del Castillo, T. J.; Thompson, N. B.; Peters, J. C. J. Am. Chem. Soc. 2016, 138, 5341–5350. (e) Chalkley, M. J.; Del Castillo, T. J.; Matson, B. D.; Roddy, J. P.; Peters, J. C. ACS Central Science 2017, 13 3, 217-223. (f) Anderson, J. S.; Cutsail, G. E.; Rittle, J.; Connor, B. A.; Gunderson, W. A.; Zhang, L.; Hoffman, B. M.; Peters, J. C. J. Am. Chem. Soc. 2015, 137, 7803–7809. (g) Rittle, J.; Peters, J. C. J. Am. Chem. Soc. 2016, 138, 4243–4248. (h) Rittle, J.; Peters, J. C. J. Am. Chem. Soc. 2017, 139, 3161–3170. For Mo systems: (i) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Matsuo, Y.; Tanaka, H.; Ishii, K.; Yoshizawa, K.; Nishibayashi, Y. Nat. Commun. 2016, 7, 12181. (j) Yandulov, D. V.; Schrock, R. R. Science 2003, 301, 76−78. (k) Arashiba, K.; Miyake, Y.; Nishibayashi, Y. Nat. Chem. 2011, 3, 120−125. (l) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Tanaka, H.; Kamaru, N.; Yoshizawa, K.; Nishibayashi, Y. J. Am. Chem. Soc. 2014, 136, 9719−9731. (m) Arashiba, K.; Kinoshita, E.; Kuriyama, S.; Eizawa, A.; Nakajima, K.; Tanaka, H.; Yoshizawa, K.; Nishibayashi, Y. J. Am. Chem. Soc. 2015, 137, 5666−5669. For Co systems: (n) Del Castillo, T. J.; Thompson, N. B.; Suess, D. L. M.; Ung, G.; Peters, J. C. Inorg. Chem. 2015, 54, 9256–9262. (o) Kuriyama, S.; Arashiba, K.; Tanaka, H.; Y. Matsuo, Y.; Nakajima, K.; Yoshizawa, K.; Nishibayashi, Y. Angew. Chem. Int. Ed. 2016, 55, 14291–14295. 5 (a) Hu, Y.; Ribbe, M. W. Methods Mol. Biol. 2011, 766, 3–7. (b) Burgess, B. K.; Lowe, D. J. Chem. Rev. 1996, 96, 2983–3011. (c) Seefeldt, L. C.; Hoffman, B. M.; Dean, D. R. Annu. Rev. Biochem. 2009, 78, 701–722. (d) Hoffman, B. M.; Dean, D. R.; Seefeldt, L. C. Acc. Chem. Res. 2009, 42, 609–619. (e) Seefeldt, L. C.; Hoffman, B. M.; Dean, D. R. Curr. Opin. Chem. Biol. 2012, 16, 19–25. 7 (a) Chatt, J.; Dilworth, J. R.; Richards, R. L. Chem. Rev. 1978, 78, 589-625. (b) Schrock, R. R. Acc. Chem. Res. 2005, 38, 955-962. (c) Nishibayashi, Y. Inorg. Chem. 2015, 54, 9234−9247. (d) van der Ham, C. J. M.; Koper, M. T. M.; Hetterscheid, D. G. H. Chem. 14 Soc. Rev. 2014, 43, 5183−5191. (e) Shaver, M. P.; Fryzuk, M. D. Adv. Synth. Catal. 2003, 345, 1061−1076. (f) MacLeod, K. C.; Holland, P. L. Nat. Chem. 2013, 5, 559−565. Chapter 2. Fe-mediated HER vs N2RR: Exploring Factors that Contribute to Selectivity in P3EFe(N2) (E = B, Si, C) Catalyst Model Systems 15 Reproduced in part with permission from: Matson, B. D.; Peters, J. C. ACS Catalysis, 2018, 8, 1448-1455. © 2018 American Chemical Society 2.1. Introduction The reduction of nitrogen (N2) to ammonia (NH3) by nitrogenase enzymes (the nitrogen reduction reaction: N2RR) has garnered substantial interest in the synthetic inorganic community for several decades.1 In particular, the structural characterization of the FeMo-cofactor of biological nitrogen fixation,2 and mechanistic uncertainties associated with this process,3 have motivated studies of synthetic (primarily Mo and Fe) model systems that mediate N2RR in the presence of proton and electron equivalents in organic solvent.4-6 The mechanisms of these systems are at various stages of understanding. Experimental4-6 and theoretical (predominantly Mo)7 studies have been undertaken to provide insight. 16 Single-site iron model complexes, such as P3BFe(N2)− and P3BFe+ (Figure 1), catalyze N2RR under a variety of conditions and driving forces, with reported selectivities (to date) for NH3 generation as high as 72% based on reductant consumed.4e In addition, conditions have been reported under which P3CFe(N2)− and P3SiFe(N2)− also catalyze N2RR to varying degrees, with the P3SiFe-system being far more efficient at the Hydrogen evolution reaction (HER) than N2RR compared to P3BFe and P3CFe.4d,8 We are naturally interested in understanding the mechanism/s by which catalysis in these respective systems occurs, and in exploring alternative systems that might function similarly. Of interest to the present study is the interplay between efficiency for the N2RR and HER on the P3BFe scaffold and its isostructural congeners P3SiFe and P3CFe. In particular, can we elucidate some of the salient factors that dictate overall product selectivity for NH3 versus H2 in these respective systems? Figure 2.1. Schematic depiction of N2RR/HER iron catalysts studied herein to explore key factors dictating product selectivity. 17 Herein we use DFT calculations to explore this question. We examine the comparative feasibility of HER via proton-coupled electron transfer (PCET)9 from several putative Fe(NxHy) early intermediates, using electronic structure calculations coupled with predicted N–H bond strengths, thermodynamic driving forces, and electron-transfer (ET) kinetics as mechanistic probes. Acknowledging the likelihood that numerous and potentially competing factors may be at play, the formation, electronic structure, and reactivity of a key common intermediate, Fe(NNH2)+, is highlighted to be an important factor in the divergent selectivity profile of P3BFe (and P3CFe) relative to the P3SiFe system. 2.1. Computational Methods All calculations were performed using dispersion corrected density functional theory (DFT-D3) using Grimmes dispersion correction. 10 All calculations were done using the full P3EFe scaffold with the TPSS functional11 and a def2-TZVP basis set on transition metals and a def2-SVP basis set on all other atoms.12 All stationary point geometries were optimized using NWChem 6.313 or Orca 3.0.3.14 To ensure consistency in grid size, all reported single point and thermodynamic energies were performed using Orca 3.0.3. Frequency calculations were used to confirm the presence of true minima and to obtain gas phase free-energy values at 195 K (Ggas). Solvation corrections were performed using the COSMO-SMD continuum model.15 The solvation free energy was approximated using gas phase and solvated single point energies (Gsolv ≈ Esoln - Egas). Finally, the free-energy of the solvated species at 195 K was 18 calculated using the gas-phase free-energy and the solvation free-energy (Gsoln,195K = Ggas,195K + Gsolv).16 The accuracy of the described computational methodology was measured by comparison to several experimental benchmarks of interest. In addition to ensuring good agreement between computed and crystallographically determined structural data, experimentally determined bond dissociation enthalpies (BDFEN-H) of the compounds P3SiFe(CNH)+, P3SiFe(CNH), P3SiFe(CNMeH)+, P3SiFe(CNMeH) and P3SiFe(NNMeH)+ could be faithfully reproduced within ±2 kcal/mol (See SI for full description).4h As a further point of calibration, the calculated singlet-triplet energy gap and the redox potentials of P3BFe(NNMe2) and P3SiFe(NNMe2)+ are in good agreement with the experimentally determined values (within ±1.5 kcal/mol, and ±3 kcal/mol (±130 mV vs Fc+/0), respectively ; see Appendix 1).4gh,17 Reduction kinetics were calculated using the standard Marcus equation relating activation barrier with driving force and total reorganization energy (λtot = λis + λos).18 The inner-sphere reorganization energy for electron transfer (λis,ET) was estimated assuming non-adiabatic behavior and by calculating the difference between the single point energies of the relevant species in its ground state and the corresponding single point energy of this ground state in the oxidized or reduced geometry (Eq. 1). λis,ET = [E(Feoxox) – E(Feoxred)] + [E(Feredred) – E(Feredox)] (2.1) The outer-sphere reorganization energy was calculated by assuming a barrier of 1.0 kcal/mol for the reduction of P3BFe(NNH2)+ followed by calculation of λtot using this 19 barrier and λis, as calculated by Eq 1. A continuum solvation model was used to support this λos value (See SI for full description).18 Reduction barriers for P3C/SiFe(NNH2)+ were subsequently calculated relative to P3BFe(NNH2)+. 2.2. Results and Discussion To set the stage for the present study, previously reported catalytic N2-to-NH3 conversion studies by P3EFe (E = B, C, and Appendix A) under an atmosphere of N2 at −78 °C in Et2O, using KC8 and [H(OEt2)2][BArF4] (HBArF4, BArF4 = tetrakis-(3,5bis(trifluoromethyl)phenyl)borate) as the reductant and acid source,4a,d established P3BFe as the most efficient catalyst for N2RR; the highest reported efficiency for this system (under these conditions) was 45 ± 3% (48 equiv acid; 58 equiv reductant). For comparison, the P3SiFe system provided a conversion efficiency of only 5 ± 3%. The P3CFe catalyst system was reasonably active at 36 ± 6% (note: ~25% lower substrate loading was used for this P3CFe value4d). Measurement of HER activity established P3SiFe(N2)− (88% per added acid equiv) as a significantly more efficient HER catalyst than P3BFe(N2)− (40% per added acid equiv) under analogous conditions.4d N2RR catalysis by P3EFe (E = B, Appendix A) has also been studied in the presence of milder reagents (e.g., Cp*2Co and [H2NPH2][OTf] or [H3NPh][OTf]); under these conditions only the P3BFe system is catalytically active. 20 Figure 2.2. (top) Previous experimental work showing the formation of P3EFe(NNH2)+ (E = B or Appendix A) via protonation with excess acid.4f,g (bottom) Calculated free energy changes (in kcal/mol) for the formation of P3EFe(NNH2)+ via P3EFe(NNH) (E = B or Appendix 1). Previous studies of the P3BFe and P3SiFe systems have also explored the generation and characterization of early stage intermediates of the N2RR catalysis.4e,f,g Most salient, low temperature protonation of P3EFe(N2)− (E = B, Appendix A) with excess HBArF4 affords the doubly protonated P3EFe(NNH2)+ species (Figure 2).4f,g As expected, corresponding DFT calculations (this work) are consistent with thermodynamically favored formation of P3EFe(NNH) via proton transfer (Figure 2); another favorable proton transfer forms P3EFe(NNH2)+. 21 2.2.1. DFT Support for Slow Fe Protonation and Fast Fe-NxHy Formation Although metal Hydride (M–H) species are most typically invoked as intermediates of transition-metal catalyzed HER,19 we do not think Fe–H species are the primary players in H2 formation by the present systems. Several experimental observations are consistent with this idea. Foremost among them is that low temperature addition of stoichiometric acid (e.g., HBArF4) to any of the anions, P3EFe(N2)−, causes overall oxidation to their corresponding neutral products, P3EFe(N2), along with release of 0.5 equiv H2.4a,d This is notewortHy because for E = Si or C the diamagnetic Hydride products, P3EFe(N2)(H), are very stable species and are formed during catalysis as end products.4b,d We posit that reactive P3EFe(NxHy) intermediates instead undergo net bimolecular HAT reactions to liberate H2 via NxHy-ligand-mediated steps (vida infra). While iron Hydrides (Fe–H) can tie up the population of active catalyst, in our view they are unlikely to be intermediates of the dominant HER pathway. To speak to this hypothesis computationally, we focus on one acid source, HBArF4, as it has been the subject of the most extensive comparative study.4d The solid-state empirical formula of HBArF4 reveals the presence of two ethers per HBArF4 ([(Et2O)2H][BArF4]).20 To determine the preferred solution-state structure of this acid, optimizations were performed in which a Et2OH+ species was provided with 0, 1 or 2 explicit Et2O molecules with which to Hydrogen bond. We found that [(Et2O)2H]+ was lowest in free-energy, with [(Et2O)3H]+ and [Et2OH]+ higher in energy by +7.0 and +8.2 kcal/mol, respectively. 22 The structure of HBArF4 is particularly crucial for Fe protonation, as a preequilibrium formation of the [Et2OH]+ appears to be required, as evidenced by relaxed surface scans. The need for dissociation of Et2O prior to Fe protonation provides a lower bound on the barrier of +8.2 kcal/mol. The requirement of [Et2OH]+ as the active acid, as opposed to [(Et2O)2H]+, is presumably steric in origin and may speak, in part, to the importance of bulky isopropyl substituents in these catalysts. Our lab recently reported that a structurally related P3SiOs(N2)− complex is an active catalyst N2RR.21 In contrast to the P3EFe(N2)− catalysts, stoichiometric HBArF4 addition can protonate at the metal, generating Os–H species that are not catalytically active for N2RR. Steric access to the larger Os center is presumably less restricted than it is for Fe. The structure of HBArF4 is particularly crucial for Fe protonation, as a preequilibrium formation of the [Et2OH]+ appears to be required, as evidenced by relaxed surface scans. The need for dissociation of Et2O prior to Fe protonation provides a lower bound on the barrier of +8.2 kcal/mol. The requirement of [Et2OH]+ as the active acid, as opposed to [(Et2O)2H]+, is presumably steric in origin and may speak, in part, to the importance of bulky isopropyl-phosphino substituents in these catalysts. Our lab recently reported that a structurally related P3SiOs(N2)- complex is an active catalyst for N2RR. 21 In contrast to the P3EFe(N2)− catalysts, stoichiometric HBArF4 addition can protonate at the metal, generating Os-H species that are not catalytically active for N2RR. Steric access to the larger Os center is presumably less restricted than it is for Fe. 23 The steric profile of the Fe(N2) unit suggests that functionalization of the β–N should not be subject to the same pre-equilibrium. This is consistent with relaxed surface scans, which show that the N2 unit can be protonated in a concerted, low energy step in which an Et2O molecule is favorably displaced by the nucleophilic β N-atom. Subsequent proton transfers yield Fe(NNH) with a low kinetic barrier (0.5–1.0 kcal/mol). Fe–H formation is thermodynamically favored for all three scaffolds. We therefore presume that the dominant source of HER for these systems is not via Fe–H formation, but that Hydride species are formed over the course of catalysis as thermodynamic products. We presume that both HER and N2RR, under the conditions explored in this work, are operating under kinetic control. In subsequent results and discussion, thermodynamics are assumed to be relevant within the context of kinetic parameters. In addition to restricting our analysis to a single acid, HBArF4, we focus on KC8 as a reductant for several reasons. Most salient is that KC8 is the only reductant that has been shown to produce catalytic yields of NH3 for all scaffolds considered. This observation is attributed to the requirement of Fe(N2)− formation during catalysis. While P3BFe(N2)− can be formed with weaker reductants, namely Cp*2Co, the more reducing P3Si/CFe(N2)− is believed to be inaccessible under these conditions. Additionally, it has been noted that, when using KC8 and HBArF4, HER proceeds with similar initial rates on P3SiFe and P3BFe scaffolds,4d possibly due to Fe(N2) reduction being a common rate limiting step. Despite the need to restrict the scope of this study to a specific catalysis cocktail, many of the conclusions should extend to other conditions reported for N2RR catalysis 24 using P3EFe (and related) complexes. In particular, the BDFEN-H values reported herein are acid and reductant independent and hence provide insight into the anticipated stability and reactivity profiles of key early intermediates of N2RR. 2.2.2. Calculation of BDFEN-H Values for Fe–NxHy Intermediates Early stage intermediates of the type Fe(NNH) and Fe(NNH2) are expected to be highly reactive;4e,h thermochemical calculations reveal the presence of extremely weak N– H bonds in these systems, as shown by their calculated bond dissociation enthalpies (BDFEN–H; Figure 2.3). In particular, as yet unobservable P3EFe(NNH) intermediates are predicted to have extremely weak N–H bonds (< 40 kcal/mol), and should therefore be subject to rapid bimolecular loss of H2 and generation of P3EFe(N2). By contrast, the BDFEN-H values of candidate P3EFe(NxHy) intermediates that are further downstream (e.g., Fe(N2H4), Fe(NH), Fe(NH2)) are predicted to be significantly larger (Figure 2.3). This notion is consistent with the solution stability of characterized examples of such downstream intermediates, contrasting the high degree of solution instability of earlier intermediates. Of particular interest herein is that the BDFEN–H values for the P3SiFe(NNH2)n+ (n = 0, 1) system are lower than those for P3B/C, for a given overall charge. As discussed later, these different BDFEN–H values are rooted in the different valence electron counts, and hence electronic structures, of the respective P3EFe-systems. 25 For additional context, it is useful to consider reported BDFEN–H data for a related P3SiFe(CN)-system. The relevant P3SiFe(CNH) species, isoelectronic with P3BFe(NNH), is calculated to have a weak BDFEN–H of 43.5 kcal/mol, in close agreement to that of 41.4 kcal/mol determined experimentally.4h Accordingly, P3SiFe(CNH) loses 0.5 equiv H2 rapidly in solution to afford P3SiFe(CN). In contrast, its oxidized cation, P3SiFe(CNH)+, has a much higher BDFEN–H (61.8 kcal/mol (calc); 61.9 kcal/mol (exp)); this species is stable to H2 loss in solution and can be isolated and structurally characterized. Considering these collected data and observations, and additional data discussed below, we presume that the earliest N2RR intermediates in P3EFe-systems are very important for determining N2RR versus HER selectivity; they engage in bimolecular H2evolving reactions that compete with productive N2RR. We next consider aspects of the H–H bond-forming steps in these early P3EFe(NxHy) intermediates in more detail. 26 Figure 2.3. BDFEN–H values (in kcal/mol) for selected P3EFe(NxHy) species.22 P3EFe(NNH) species are plausible candidates to consider with respect to selectivity since bimolecular H2-evolving reactions can presumably result from their extremely weak N−H bonds (Figure 2.3; 31-17 kcal/mol). P3SiFe(NNH), with a BDFEN–H estimated to be 8.2 kcal/mol lower than for P3BFe(NNH), might be reasonably expected to liberate H2 more readily, thereby attenuating its N2RR efficiency. However, the BDFEN–H for P3CFe(NNH) is calculated to be even lower (17.3 kcal/mol) than for P3SiFe(NNH) (23 kcal/mol), despite the fact that P3CFe(N2)− is appreciably more efficient for N2RR. Hence, a trend is not evident on the basis of the Fe(NNH) intermediates, at least as related to their relative 27 BDFEN–H values. Fe(NNH) intermediates are readily protonated to form Fe(NNH2)+ species in solution at low temperature (Figure 2.2). This likewise suggests that Fe(NNH) intermediates are unlikely to be primarily responsible for HER under catalytic conditions when a large excess of acid is present.4f,g P3EFe(NNH2) BDFEN–H values provide a more tractable trend: the respective calculated values are 38.2 kcal/mol for P3BFe, 34.4 kcal/mol for P3CFe, and 22.9 kcal/mol for P3SiFe; the P3EFe–NNH2 species that exhibits the most efficient N2RR activity exhibits the strongest N-H bond, and the least efficient exhibits the weakest (Figure 2.3). 2.2.3. Calculated Reduction Kinetics of P3EFe(NNH2)+ To gain further insight into the respective role P3EFe(NNH2)+/0 (E = B, Si, C) species might play in dictating product selectivity, P3EFe(NNH2)+ reduction kinetics were derived using the standard Marcus equation relating the driving force and total reorganization energy with the ET activation barrier.18 Comparison of the optimized Fe(NNH2) and Fe(NNH2)+ redox pairs reveals significant differences in their respective reduction potentials and innersphere reorganization energies (λis,ET). The P3BFe(NNH2)+ species is predicted to have a considerably more positive reduction potential (−1.2 V vs Fc+/0) than P3SiFe(NNH2)+ (−1.9 V; Table 2.1), resulting from their different valence electronic counts and electronic structures (see below). Given their dramatic difference in reduction potentials, the barrier for reduction (G*) is expected to sharply increase in moving from B to Si. Relative reduction barrier calculations, 28 assuming G* = 1.0 kcal/mol for the reduction of P3BFe(NNH2)+, predict activation barriers that are 4–5 times higher in energy for the reduction of P3C/SiFe(NNH2)+ versus P3BFe(NNH2)+ (Table 2.1). While the reduction of all three species should be more than readily accomplished by the strong reductant KC8, P3C/SiFe(NNH2)+ species are predicted to be significantly longer lived than the P3BFe(NNH2)+ congener. The P3BFe(NNH2)+ intermediate is predicted to have a lower propensity for H2liberating PCET reactivity, and is also predicted to be reduced much more rapidly. Since the reaction of two P3BFe(NNH2) molecules is a more probable source of H2 on this scaffold, the efficiency for N2RR on P3BFe should be strongly coupled to the rate at which P3BFe(NNH2) can be productively consumed (i.e., protonated to form a P3BFe(NHNH2)+ or P3BFe(NNH3)+). Mechanistic experiments to address these scenarios are ongoing. For example, a recent study has provided experimental evidence that P 3BFe(NNH2) is protonated by acid at low temperature to liberate P3BFe(N)+ and NH3, presumably via P3BFe(NNH3)+. We conclude that facile reduction of P3BFe(NNH2)+ to P3BFe(NNH2), relative to that for P3SiFe(NNH2)+ and P3CFe(NNH2)+, is one important factor in determining its comparative efficiency for N2RR. As further elaborated below, long-lived P3EFe(NNH2)+ intermediates can, via bimolecular PCET pathways, instead lead to unproductive HER. This HER activity, however, is dependent on both a long-lived P3EFe(NNH2)+ intermediate, and the presence of a highly reactive PCET reagent, such as a P 3EFe(NNH2) species. We have previously postulated that P3EFe(NNH2) formation is required for the release of the 29 first equivalent of NH3 and thus suggest that this species may be a crucial intermediate in both HER and N2RR.4fg,17 Table 2.1. Calculated thermodynamic and kinetic parameters for P3EFe(NNH2)+ a P3EFe(NNH2)a P3EFe(NNH2)+ + e-  P3EFe(NNH2) a Eo (vs Fc+/0) λis,ET G*rel krelb E =B −1.2 V 23 1.0 1 E = Si −1.9 V 30 4.4 2x10-4 E=C −2.0 V 30 5.2 2x10-5 Energies are in kcal/mol, unless noted otherwise. bG*rel values were calculated assuming a P3BFe(NNH2)+ reduction barrier of 1.0 kcal/mol. krel ≡ exp[(G*B−G*E)/kbT] where T = 195 K. 2.2.4. Calculated PCET Reactions The differences in N–H bond strengths and relative rates of P3EFe(NNH2)+ reduction, with corresponding implications for product selectivity, are further highlighted by calculating the thermodynamic and kinetic parameters for several PCET reactions of interest (Figure 2.4ABC). In particular, comparative driving forces were calculated for unproductive bimolecular PCET reactions that generate H2 between P3EFe(NNH2)n+ (n = 0,1; E = B, Si, C) and P3EFe(NNH2). Consistent with the calculated BDFEN–H values (Figure 2.3), the P3SiFe, and to a lesser extent the P3CFe, system shows a higher propensity 30 to undergo PCET to liberate H2 and the corresponding reduced Fe–NNHy species. This is especially apparent in the reaction between two P3EFe(NNH2)+ species, and in the crossreaction between an P3EFe(NNH2)+ cation and a neutral P3EFe(NNH2) species. In the former case, two P3EFe(NNH2)+ (E = Si, C) species react in a very favorable step to form 0.5 equiv H2 and P3EFe(NNH)+ (ΔGcalc = −17.5 kcal/mol and −16.5, respectively; Figure 2.4A). The reaction barrier is expected to be dominated in this case by the work required to bring two cationic species together in solution (~5 kcal/mol; see APPENDIX A), highlighting the reactive Nature of P3C/SiFe(NNH2)+. In contrast, P3BFe(NNH2)+ shows a correspondingly uphill PCET reaction (ΔGcalc = +3.1 kcal/mol) in its self-combination to liberate H2 and P3BFe(NNH)+;23 P3BFe(NNH2)+ is also much more readily reduced to P3BFe(NNH2) (Table 2.1). Figure 2.4. Calculated free energy changes (ΔGcalc; in kcal/mol; 195 K) for several putative PCET reactions that evolve H2. 31 The bimolecular reaction between cationic P3EFe(NNH2)+ with P3EFe(NNH2) to produce H2 and the corresponding P3EFe(NNH)+ and P3EFe(NNH) byproducts is predicted to be favorable for all three systems (Figure 2.4C). However, the P3C/SiFe systems proceed with far more driving force than the P3BFe system. Favorable driving forces are also predicted for all three systems in self reactions of P3EFe(NNH2) to produce H2 and P3EFe(NNH), but again the P3C/SiFe systems proceed with far more driving force (Figure 2.4B). While the bimolecular reaction of P3EFe(NNH2) with itself is therefore a presumed source of H2 for each system, in sum the P3C/SiFe systems are more likely, under each of the considered bimolecular reactions, to liberate H2, in accord with their efficiency for HER versus N2RR relative to the P3BFe system. Given that the reduction of P3C/SiFe(NNH2)+ is predicted to be comparatively slow, one might expect such a species to build-up as an intermediate. This possibility warrants future experimental studies aimed at in situ detection. At the present stage, we can suggest that a high (relative) concentration of P3C/SiFe(NNH2)+, and a high predicted propensity for HER via reaction of this species with either itself or P3C/SiFe(NNH2), leads to unproductive PCET steps that evolve H2 as competitive with downstream N2 reduction steps that lead to N2RR. This is one important factor in determining selectivity. 32 Figure 2.5. Overview of predicted bimolecular HER and N2RR pathways for P3EFe(NNHy) species and pertinent BDFEN–H values. Since the P3BFe(NNH2)+ intermediate is predicted to have a lower propensity for H2-liberating PCET reactivity, and is also predicted to be reduced much more rapidly, the reaction of two P3BFe(NNH2) molecules is a more probable source of H2 for this scaffold; the efficiency for N2RR on P3BFe should therefore be related to the rate at which P3BFe(NNH2) can be productively consumed (i.e., protonated to form a P3BFe(NHNH2)+ or P3BFe(NNH3)+). Mechanistic experiments to address these scenarios are ongoing. For 33 example, a recent study has shown that P3BFe(NNH2) can be protonated by strong acid at low temperature to liberate P3BFe(N)+ and NH3, presumably via P3BFe(NNH3)+.17 While the P3CFe scaffold provides a less definitive comparison, the calculated BDFEN–H values and H2-evolving PCET thermodynamics suggest that the dominant source of HER on the P3C/SiFe scaffolds may be the reaction between Fe(NNH2) and Fe(NNH2)+. The highly reducing nature of P3CFe(NNH2)+, as for the P3Si scaffold, suggests it should be comparatively long-lived, and thus more likely to undergo PCET with P3CFe(NNH2). The similarity between P3CFe and P3SiFe in their thermodynamics for the reaction between two Fe(NNH2)+ species (Figure 2.4A) does not correlate with their disparate %NH3 efficiencies. Substantial differences in their predicted thermodynamics for the reaction between Fe(NNH2) and Fe(NNH2)+ (Figure 2.4C) are more in line with the observed trend. This type of bimolecular reactivity may be an important source of HER on the P 3C/SiFe scaffolds (Figure 2.5). 2.2.5. Wiberg Bond Indices of P3EFe(NxHy) Species We next examine how each P3E auxiliary, and the corresponding P3EFe(NNHy) valence at iron, confers variability in bonding to, and the electronic structure of, the NNHy ligand, as a means of further considering corresponding reactivity differences of P3EFe(NNHy) species. Wiberg bond indices provide a means to examine how the localized bonding between various atoms, expressed as a bond index,24 changes as a function of the NNHy 34 reduction state (i.e., NNH to NNH2). We have suggested elsewhere that the relative flexibility of the P3B ligand, owing to a weak and dative Fe-B interaction, may allow for stabilization of Fe−NNHy intermediates where Fe–N π-bonding is accompanied by pyramidalization at the Fe center, and a corresponding lengthening of the Fe–B distance.4a,17,25 The P3Si ligand is expected to give rise to a more shared, covalent Fe–Si interaction, irrespective of the NNHy reduction state, and the P3C system may be expected to fall in the middle of these extremes.4b Changes in the respective bond indices of these frameworks have been determined between pairs of P3EFe(NNH) and P3EFe(NNH2) species (E = B, C, Appendix A), related by formal addition of an H-atom to the former. Interestingly, the N–H bond indices are essentially invariant across all complexes studied, indicating that differences in BDFEN–H are mostly dependent on the relative bonding through the E–Fe–N–N manifold.26 The most salient data, reproduced in Figure 2.6, are the total Wiberg bond indices for Fe−Nα, Fe−Nβ, Fe−E, N−N and N−H. The total Fe−N–N bond order, ∑(Fe−N−N), is also provided, as is the net difference in the BDFEN-H value, for each pair on moving from Fe(NNH) to Fe(NNH2). As expected, the Fe–E bond order weakens slightly from Fe(NNH) to Fe(NNH2) for E = B, and stays constant for both Si and C. The respective change at Fe-Nα is also informative. For the B system, a significant increase is observed (1.6 to 1.9), reflecting a build-up in pi-bonding in P3BFe(NNH2), akin to low-spin (pseudotetrahedral) iron imides 35 of the type P3BFe(NR). For comparison, a previously characterized P3BFe(NR) species (R = 4-OMe-Ph) is predicted to have an Fe–N bond order of 1.8 (see Appendix A). By contrast, the Fe–Nα index for Si is sharply attenuated (from 1.6 to 1.2), reflecting a corresponding decrease in pi bonding. While this difference must partly reflect a less flexible Fe-Si interaction, it also reflects the electronic structure resulting from an extra electron in the frontier orbitals of the 2E {Fe-Si}7 system relative to 1A {Fe-B}6. Interestingly, P3BFe(NNH2) is pyramidalized at Nβ whereas Nβ is planar for P3SiFe(NNH2). This observation can again be rationalized by the assignment of a low-spin iron “imidelike” electronic structure to {Fe-B}6 P3BFe(NNH2), but not for {Fe-Si}7 P3SiFe(NNH2), where substantial spin leaks onto the NNH2 subunit (19% on P3SiFe(NNH2)). The C system provides an interesting further comparison, with spin leakage onto the NNH2 unit falling between these two extremes (12% on P3CFe(NNH2)). An increase in the Fe-Nα index occurs from P3CFe(NNH) to P3CFe(NNH2) (1.2 to 1.4), but Nβ is predicted to remain planar. There also appears to be a strong trend between the degree of change in the total Fe−N−N bond order (∑(Fe−N−N)) and the BDFEN-H; The B and C systems show little change in ∑(Fe−N−N), with a corresponding significant increase in BDFEN-H from Fe(NNH) to Fe(NNH2) (7.0 and 17.9 kcal/mol, respectively). However, the P3CFe system starts at a much weaker BDFEN-H of 17.3 kcal/mol for P3CFe(NNH) (compared to 31.2 kcal/mol for P3BFe). This observation is consistent with their total ∑(Fe−N−N) values (3.8 for B and 2.9 for C). Thus, the comparative stability of P3CFe(NNH2), with its much higher BDFEN-H relative that in P3CFe(NNH), appears to reflect a higher degree of instability in 36 P3CFe(NNH) (relative to the same comparison for E = B). This idea is further supported by Wiberg bond indices of the P3EFe(N2) species, which show a total bond order of 4.0 across the Fe–N–N unit for all three scaffolds (Figure 2.6). In sharp contrast, the P3SiFe system has a relatively high ∑(Fe−N−N) value in P3SiFe(NNH), but this value decreases dramatically in P3SiFe(NNH2). There is correspondingly very little change in the BDFEN-H, reflecting a comparatively very weak N-H bond in P3SiFe(NNH2). The instability of P3SiFe(NNH2), with an electronic structure that places substantial unpaired spin on NNH2 owing to the {Fe-Si}7 configuration, presumably contributes to the cathodically shifted reduction potential predicted for P3SiFe(NNH2)+ relative to P3BFe(NNH2)+, and also its propensity for facile PCET to liberate H2. In sharp contrast, the P3SiFe system has a relatively high ∑(Fe−N−N) value in P3SiFe(NNH), but this value decreases dramatically in P3SiFe(NNH2). There is correspondingly very little change in the BDFEN-H, reflecting a comparatively very weak N-H bond in P3SiFe(NNH2). The instability of P3SiFe(NNH2), with an electronic structure that places substantial unpaired spin on NNH2 owing to the {Fe-Si}7 configuration, presumably contributes to the cathodically shifted reduction potential predicted for P3SiFe(NNH2)+ relative to P3BFe(NNH2)+, and also its propensity for facile PCET to liberate H2. 37 Figure 2.6. Selected total Wiberg bond indices for P3EFe(N2), P3EFe(NNH) and P3EFe(NNH2) species, along with the total Fe−N–N bond order, ∑ (Fe−N−N). BDFEN-H values are reported in kcal/mol. 2.3. Conclusions Exploring the chemical basis for N2RR versus HER selectivity for a molecular catalyst is important to future catalyst design. The DFT study described herein suggests 38 that PCET reactions involving P3EFe(NNH2)n+ species likely play an important role in the efficiency of N2-to-NH3 conversion catalysis by P3EFe model systems. These calculations enable predictions qualitatively consistent with previous stoichiometric and catalytic experiments. The comparative stability of P3EFe(NNH2)n+ intermediates, as predicted by calibrated BDFEN–H values and redox potentials, emerges as one of the important factors in determining selectivity for N2RR versus HER in these systems. Corresponding Wiberg bond indices intimate P3B as an especially well-equipped ligand for supporting N2RR at Fe, due to its high degree of flexibility and the valence electron count it confers to Fe in the reduced intermediate P3BFe(NNH2). Our study suggests that increasing the rate at which an P3EFe(NNH2) intermediate is productively consumed so as to avoid bimolecular HER, possibly via rapid PCET reagents, may be a promising route to increasing efficiency for NH3 production. Looking beyond these iron model systems, our study underscores the potential utility of DFT-predicted BDFEN-H determinations towards the rational design of catalysts for N2RR. Intermediates with weak N–H bonds (e.g., M(NNH) and M(NNH2)) are highlighted as important sources of H2 production via bimolecular PCET. Such a scenario is distinct from HER activity via more traditional metal-Hydride intermediates. 2.4. References 1 (a) Chatt, J.; Dilworth, J. R.; Richards, R. L. Chem. Rev. 1978, 78, 589-625. (b) Schrock, R. R. Acc. Chem. Res. 2005, 38, 955-962. (c) Nishibayashi, Y. Inorg. Chem. 2015, 54, 39 9234−9247. (d) van der Ham, C. J. M.; Koper, M. T. M.; Hetterscheid, D. G. H. Chem. Soc. Rev. 2014, 43, 5183−5191. (e) Shaver, M. P.; Fryzuk, M. D. Adv. Synth. Catal. 2003, 345, 1061−1076. (f) MacLeod, K. C.; Holland, P. L. Nat. Chem. 2013, 5, 559−565. 2 (a) Howard, J. B.; Rees, D. C. Proc. Nat. Acad. Sci. 2006, 103, 17088-17093. (b) Howard, J. B.; Rees, D. C. Chem. Rev. 1996, 96, 2965−2982. 3 Hoffman, B. M.; Lukoyanov, D.; Yang, Z.-Y.; Dean, D. R.; Seefeldt, L. C. Chem. Rev. 2014, 114, 4041−4062. 4 For pertinent Fe systems: (a) Anderson, J. S.; Rittle, J.; Peters, J. C. Nature 2013, 501, 84−87. (b) Creutz, S. E.; Peters, J. C. J. Am. Chem. Soc. 2014, 136, 1105−1115. (c) Ung, G.; Peters, J. C. Angew. Chem., Int. Ed. 2015, 54, 532−535. (d) Del Castillo, T. J.; Thompson, N. B.; Peters, J. C. J. Am. Chem. Soc. 2016, 138, 5341–5350. (e) Chalkley, M. J.; Del Castillo, T. J.; Matson, B. D.; Roddy, J. P.; Peters, J. C. ACS Central Science 2017, 3, 217-223. (f) Anderson, J. S.; Cutsail, G. E.; Rittle, J.; Connor, B. A.; Gunderson, W. A.; Zhang, L.; Hoffman, B. M.; Peters, J. C. J. Am. Chem. Soc. 2015, 137, 7803–7809. (g) Rittle, J.; Peters, J. C. J. Am. Chem. Soc. 2016, 138, 4243–4248. (h) Rittle, J.; Peters, J. C. J. Am. Chem. Soc. 2017, 139, 3161–3170. 5 For Mo systems: (a) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Matsuo, Y.; Tanaka, H.; Ishii, K.; Yoshizawa, K.; Nishibayashi, Y. Nat. Commun. 2016, 7, 12181. (b) Yandulov, D. V.; Schrock, R. R. Science 2003, 301, 76−78. (c) Arashiba, K.; Miyake, Y.; Nishibayashi, Y. Nat. Chem. 2011, 3, 120−125. (d) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Tanaka, H.; Kamaru, N.; Yoshizawa, K.; Nishibayashi, Y. J. Am. Chem. Soc. 2014, 40 136, 9719−9731. (e) Arashiba, K.; Kinoshita, E.; Kuriyama, S.; Eizawa, A.; Nakajima, K.; Tanaka, H.; Yoshizawa, K.; Nishibayashi, Y. J. Am. Chem. Soc. 2015, 137, 5666−5669. 6 For Co systems: (a) Del Castillo, T. J.; Thompson, N. B.; Suess, D. L. M.; Ung, G.; Peters, J. C. Inorg. Chem. 2015, 54, 9256–9262. (b) Kuriyama, S.; Arashiba, K.; Tanaka, H.; Y. Matsuo, Y.; Nakajima, K.; Yoshizawa, K.; Nishibayashi, Y. Angew. Chem. Int. Ed. 2016, 55, 14291–14295. 7 (a) Reiher, M.; Le Guennic, B.; Kirchner, B. Inorg. Chem. 2005, 44, 9640–9642. (c) Studt, F.; Tuczek, F. Angew. Chem. Int. Ed. 2005, 44, 5639–5642. (d) Studt, F.; Tuczek, F. J. Comput. Chem. 2006, 27, 1278–1291. (e) Ujjal Gogoi, U.; Kanti Guha, A.; Phukan, A. K. Chem. Eur. J. 2013, 19, 11077–11089. 8 Side-by-side comparisons of catalytic N2RR by P3EFe (E = B, C, and Appendix A) using potassium graphite (KC8) and [(Et2O)2H][BArF4] (HBArF4, BArF4 = tetrakis-(3,5bis(trifluorometHyl)phenyl)borate) have revealed P3BFe to be the most efficient catalyst, with reported efficiencies up to 37%. N2RR catalysis by P3CFe or P3SiFe is less efficient under these conditions, with reported efficiencies of 33% and 4%, respectively. Studies focused on HER catalysis have shown P3SiFe(N2)- (88% per H+) to be significantly more efficient than P3BFe(N2)- (40% per H+) under analogous conditions. See 4d. 9 In this study, we do not distinguish between the terms H-atom transfer (HAT) vs PCET. For consistency, we refer to PCET throughout. For discussion of terminology see: (a) Warren, J. J.; Tronic, T. A.; Mayer, J. M. Chem. Rev. 2010, 110, 6961–7001. (b) HammesSchiffer, S. J. Am. Chem. Soc. 2015, 137, 8860−8871. 10 Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104. 41 11 Tao, J. M.; Perdew, J. P.; Staroverov, V. N.; Scuseria, G. E. Phys. Rev. Lett. 2003, 91, 146401. 12 Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297–3305. 13 Valiev, M.; Bylaska, E. J.; Govind, N.; Kowalski, K.; Straatsma, T. P.; Van Dam, H. J. J.; Wang, D.; Nieplocha, J.; Apra, E.; Windus, T. L.; de Jong, W. A. Comput. Phys. Commun. 2010, 181, 1477–1489. 14 Neese, F. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2012, 2, 73-78. 15 (a) Klamt, A.; Schüürmann, G. J. Chem. Soc. Perkin Trans. 2. 1993, 2, 799–805. (b) Marten, B.; Kim, K.; Cortis, C.; Friesner, R. A.; Murphy, R. B.; Ringnalda, M. N.; Sitkoff, D.; Honig, B. J. Phys. Chem. 1996, 100, 11775–11788. 16 (a) Ribeiro, R. F.; Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B 2011, 115, 14556–14562. (b) Wang, T.; Brudvig, G.; Batista, V. S. J. Chem. Theory Comput. 2010, 6, 755–760. (c) Marten, B.; Kim, K.; Cortis, C.; Friesner, R. A.; MurpHy, R. B.; Ringnalda, M. N.; Sitkoff, D.; Honig, B. J. Phys. Chem. 1996, 100, 11775–11788. 17 Thompson, N.B.; Green, M. T.; Peters, J.C. J. Am. Chem. Soc., 2017, 139, 16105-16108. 18 Marcus, R. A. J. Chem. Phys. 1956, 24, 966–978. 19 (a) Dempsey, J. L.; Brunschwig, B. S.; Winkler, J. R.; Gray, H. B. Acc. Chem. Res. 2009, 42, 1995–2004. (b) Lewandowska-Andralojc, A.; Baine, T.; Zhao, X.; Muckerman, J. T.; Fujita, E.; Polyansky, D. E. Inorg. Chem. 2015, 54, 4310–4321. (c) Yang, J. Y.; Smith, S. E.; Liu, T.; Dougherty, W. G.; Hoffert, W. A.; Kassel, W. S.; DuBois, M. R.; DuBois, D. L.; Bullock, R. M. J. Am. Chem. Soc. 2013, 135, 9700–9712. 20 Brookhart, M.; Grant, B.; Volpe Jr., A. F. Organometallics 1992, 11, 3920–3922. 42 21 Fajardo, J.; Peters, J. C. J. Am. Chem. Soc. 2017, 139, 16105–16108. 22 We have previously reported bond dissociation enthalpies (BDEN-H) for P3BFe(NNH) and P3BFe(NNH2) (see ref 4e). Here we reported BDFEN-H values as they have more theoretical justification in the absence of experimental knowledge of the entropy change associated with H· loss. 23 While this discussion may seem at odds with the enhanced stability of P 3SiFe(NNH2)+ relative to P3BFe(NNH2)+, other factors are presumably responsible in solution, such as the more facile reduction of P3BFe(NNH2)+ relative to P3SiFe(NNH2)+. 24 Wiberg, K. B. Tetrahedron, 1968, 24, 1083-1096. 25 Moret, M-.E.; Peters, J. C. J. Am. Chem. Soc., 2011, 133, 18118-18121. 26 Similarly, the C2H5 radical is predicted to have a very low BDFEC-H (34 kcal/mol) when compared to C2H4 (100 kcal/mol), but the Wiberg bond indices for their respective C–H bonds do not change appreciably (See Appendix A) 43 Chapter 3. Catalytic N2-to-NH3 Conversion by Fe at Lower Driving Force: A Proposed Role for Metallocene-Mediated PCET 44 Reproduced in part with permission from: Chalkley, M. J.; Del Castillo, T. J.; Matson, B. D.; Roddy, J. P.; Peters, J. C. ACS Central Science, 2017, 3, 217-223. © 2017 American Chemical Society 3.1. Introduction The reduction of N2 to NH3 is critical for life and is performed on a massive scale both industrially and biologically.1 The high stability of the N≡N triple bond necessitates catalysts and high-energy reagents/conditions to achieve the desired transformation.2 Synthetic studies of catalytic N2-to-NH3 conversion by model complexes are of interest to constrain hypotheses concerning the mechanism/s of biological (or industrial) N2-fixation and to map fundamental catalyst design principles for multi-electron reductive transformations.3 Interest in Fe model systems that catalyze N2-to-NH3 conversion has grown in part due to the postulate that one or more Fe centers in the FeMo-cofactor of FeMo-nitrogenase may serve as the site of N2 binding and activation during key bond-breaking and -making steps.4 Previous examples of synthetic molecular Fe catalysts that mediate N2-to-NH3 conversion operate with high driving force, relying on a very strong acid (pKa ca. 0) and reductant (E° < −3.0 V vs Fc+/0).5 In contrast, several Mo catalysts have been shown to facilitate N2-toNH3 conversion with significantly lower driving force.6 There is thus interest in exploring 45 the viability of Fe-mediated catalytic N2-to-NH3 conversion under less forcing conditions from a practical perspective, and to continue assessing these systems as functional models of biological nitrogenases. Figure 3.1. Summary of conditions used for catalytic N2-to-NH3 conversion by P3BFe+ highlighting the estimated enthalpic driving force (ΔΔHf).7 In this chapter, it is demonstrated that catalytic conversion of N2 to NH3 by P3BFe+ (P3B = tris(o-diisopropylphosphinophenyl)borane) can be achieved with a significantly lower driving force by coupling Cp*2Co with [Ph2NH2]+ or [PhNH3]+ (Figure 3.1). Such conditions additionally afford unusually high selectivity and catalytic turnover for NH3.8 Moreover, it is noted that the use of milder reagents as reductant and acid engenders a higher effective bond dissociation enthalpy (BDFE; eq 3.1).7a,9 This may in turn afford access to proton-coupled electron transfer (PCET) pathways (e.g., FeN2 + H·  FeN2H) distinct from electron transfer (ET)/proton transfer (PT) pathways, thus enhancing overall catalytic efficiency. BDFEeffective = 1.37(pKa) + 23.06(E0) + CG (3.1) 46 Theoretical considerations, including DFT calculations, are discussed that suggest the viability of a decamethylcobaltocene-mediated PCET pathway in this system; by extension we suggest metallocene-mediated (e.g., Cp*2Cr) PCET pathways may be operative in previously studied Mo and Fe N2-fixing systems that use metallocene reductants.6,8 3.2. Results 3.2.1 Catalysis Using [P3BFe][BArF4], Cp*2Co and [RnNH(4-n)][OTf] Various observations of P3BFe complexes in the presence of acids and reductants suggested that this system might be capable of N2-to-NH3 conversion with lower driving force than that originally reported. Accordingly, we had observed that the treatment of P3BFeN2- with KC8 and weaker acids (pKa > 0) led to greater than stoichiometric NH3 formation (e.g., under unoptimized conditions [2,6-dimethylanilinium][OTf] afforded 2.1 equiv NH3 per Fe).10 Similarly, the treatment of P3BFeN2- with [H(OEt2)2][BArF4] (HBArF4, BArF4 = tetrakis(3,5-bis(trifluoromethyl)phenyl)borate) and weaker reductants led to modest yields of NH3. For example, under unoptimized conditions we had observed that decamethylcobaltocene (Cp*2Co) and HBArF4 afforded 0.6 equiv NH3 per Fe.10,11 Most recently, an apparent catalytic response was observed during a cyclic voltammetry experiment at the P3BFeN20/- couple (−2.1 V vs Fc+/0) upon addition of excess HBArF4 under an N2 atmosphere. Electrolytic NH3 generation by P3BFe+ was observed at −2.4 V vs Fc+/0 in Et2O,11 and Na/Hg (−2.4 V vs Fc+/0 in THF)7b could instead be used for N2-to-NH3 conversion catalysis (albeit less selectively and with low turnover). Finally, mixing P3BFe+ with Cp*2Co in Et2O at −78 °C under N2 generates some P3BFeN2- as observed by X-band 47 EPR and Mössbauer spectroscopy (See Appendix 2), suggesting that Cp*2Co is in principle a sufficiently strong reductant to trigger catalysis by P3BFe+. Treatment of P3BFe+ with Cp*2Co and [Ph2NH2][OTf], [Ph2NH2][BArF4], or [PhNH3][OTf] in Et2O at −78 °C under an N2 atmosphere affords catalytic yields of NH3 (Table 3.1). Notably, the highest selectivity for NH3 obtained among this series (72% at standard substrate loading; Entry 1) is significantly improved compared to all previously described (molecular) Fe catalysts for N2-to-NH3 conversion.8,12 Tripling the initial substrate loading (Entry 2) nearly triples the NH3 production with only modest loss in efficiency for NH3 (63%). Preliminary attempts to further increase the initial substrate loading have led to substantially decreased efficiency (Entry 3). However, substrate reloading experiments (Entries 4 and 5) maintain greater than 50% efficiency for NH3 overall; a turnover number for NH3 generation via two reloadings has been achieved as high as 89 (84 ± 8; Entry 5). This is the highest turnover number yet reported for a (molecular) N2-to-NH3 conversion catalyst under any conditions.13 The use of the more soluble acid [Ph2NH2][BArF4] (Entry 6) provides significantly lower, but still catalytic, yields of NH3. This more soluble acid presumably increases background reactivity with Cp*2Co (See Appendix 2). Perhaps more significantly, [PhNH3][OTf] is a considerably weaker acid than [Ph2NH2][OTf] (Figure 3.1), but still provides substantial catalytic yields of NH3 (Entries 7 and 8) and at efficiencies that compare well with those obtained previously using HBArF4 and KC8 despite a difference in driving force of nearly 100 kcal/mol.11 48 Table 3.1. N2-to-NH3 Conversion with P3EM Complexes (M = Fe, Co) Catalyst Cp*2Co Acid Equiv. % Yield (equiv) (equiv) NH3/Fe (NH3/e-) 1 P3BFe+ 54 108b 12.8 ± 0.5 72 ± 3 2 P3BFe+ 162 322b 34 ± 1 63 ± 2 3 P3BFe+ 322 638b 26.7 ± 0.9 25 ± 1 4a P3BFe+ [162]x2 [322]x2b 56 ± 9 52 ± 9 5a P3BFe+ [162]x3 [322]x3b 84 ± 8 52 ± 5 6 P3BFe+ 54 108c 8±1 42 ± 6 7 P3BFe+ 54 108d 7±1 38 ± 7 8 P3BFe+ 162 322d 16 ± 3 29 ± 4 9 P3SiFeN2 54 108b 1.2 ± 0.1 6±1 10 P3BCoN2- 54 108b 1.1 ± 0.4 6±2 11 P3SiCoN2 54 108b 0±0 0±0 The catalyst, acid, Cp*2Co, and Et2O were sealed in a vessel at −196 °C under an N2 atmosphere followed by warming to −78 °C and stirring. Yields are reported as an average of at least 2 runs; for individual experiments See Appendix 2. aFor these experiments the reaction was allowed to proceed for 3 hours at −78 °C before cooling to −196 °C and furnished with additional substrate and solvent b[Ph2NH2][OTf]. c[Ph2NH2][BArF4]. d [PhNH3][OTf]. We also screened several related phosphine-ligated Fe–N2 and Co–N2 complexes14 under the new standard reaction conditions with [Ph2NH2][OTf] and Cp*2Co (Entries 9–11) but 49 found that none of these other systems were competent catalysts. While we anticipate other catalyst systems for N2-to-NH3 conversion may yet be found that function under the conditions described herein,8 certain features of the P3BFe system correlate with unusually productive catalysis.14b Also significant is that when P3BFe+ is loaded with 322 equiv [Ph2NH2][OTf] and 162 equiv Cp*2Co in Et2O at −78 °C, modest levels of N2H4 are detected (< 1 equiv per Fe; see Appendix 2). We had previously reported that catalytic N2 reduction with KC8 and HBArF4 yielded no detectable hydrazine, but observed that if hydrazine was added at the outset of a catalytic run, it was consumed.5a When 5 equiv of N2H4 were added at the beginning of a catalytic run (again with 322 equiv [Ph2NH2][OTf] and 162 equiv Cp*2Co), only 0.22 equiv of N2H4 (4.4% recovery) remained after workup. This result indicates that liberated hydrazine can also be reduced or disproportionated under the present conditions. That N2H4 is detected to any extent in the absence of initially added N2H4 under these conditions indicates that a late N–N cleavage mechanism to produce NH3 (e.g., alternating or hybrid cross-over) is accessible.3b,15 Whether such a pathway is kinetically dominant is as yet unclear.11,16 3.2.2. Fe Speciation under Turnover Conditions The P3BFe speciation under turnover conditions was probed via freeze-quench Mössbauer spectroscopy.11 The Mössbauer spectrum of a catalytic reaction mixture after five minutes of reaction time (Figure 3.2) reveals the presence of multiple species featuring wellresolved sets of quadrupole doublets. The spectrum is satisfactorily simulated with P3BFeN2 (δ = 0.55 mm/sec, ΔEQ = 3.24 mm/sec, 32%; Figure 3.2 green), P3BFeN2- (δ = 50 0.40 mm/sec, ΔEQ = 0.98 mm/sec, 26%; Figure 2 blue),11,17 an unknown P3BFe species (δ = 0.42 mm/sec, ΔEQ = 1.84 mm/sec, 18%; Figure 3.2 yellow), and a final species that is modeled with δ = 0.96 mm/sec and ΔEQ = 3.10 mm/sec (24%; Figure 3.2 orange). The broad nature of this last signal and its overlap with other features in the spectrum prevents its precise assignment, but its high isomer shift and large quadrupole splitting are suggestive of a tetrahedral, S = 2 Fe(II) complex.18 The Mössbauer spectrum of a catalytic reaction mixture after 30 minutes was also analyzed (See Appendix 2). The spectrum still shows P3BFeN2 (53%), the same unknown P3BFe species (18%), and again a tetrahedral, high-spin Fe(II) component (22%). However, P3BFe+ is now present (δ = 0.75 mm/sec, ΔEQ = 2.55 mm/sec, 8%) and P3BFeN2- is no longer observed. The reloading experiments described above provide strong evidence that “P3BFe” species represent an “active catalyst” population; interpretation of the relative speciation via spectroscopy should hence bear on the mechanism of the overall catalysis. 51 Figure 3.2. Mössbauer spectrum at 80 K with 50 mT applied parallel field of a freezequenched catalytic reaction (54 equiv Cp*2Co, 108 equiv [Ph2NH2][OTf], 1 equiv P3B[57Fe]+) after five minutes of reaction time. The appearance of a presumed high-spin (S = 2), tetrahedral Fe(II) species during catalysis (ca. 25%) might arise via dechelation of a phosphine arm. This species could represent an off-path state, or a downstream deactivation product. Interestingly, under the present catalytic conditions we do not observe the borohydrido-hydrido species P3B(μ-H)Fe(H)(L) (L = N2 or H2); this species was postulated to be an off-path resting state during N2-to-NH3 conversion catalysis using HBArF4 and KC8 and was the major component observed at early times (ca. 60% at 5 min).11 It therefore appears that a larger fraction of the “P3BFe” species are in a catalytically on-path state at early reaction times under these new catalytic conditions. Additionally, the presence of a significant degree of P3BFeN2- (Figure 2) at an early time point is distinct from conditions with HBArF4 and KC8.11 This observation is consistent with the notion that protonation of P3BFeN2- is slowed under the present conditions, likely as a result of the insolubility of the triflate salt [Ph2NH2][OTf] and its attenuated acidity relative to HBArF4.7c-d,19 Clearly, differences in the rates of key elementary steps under the new conditions described here may lead to new mechanistic scenarios for N2-to-NH3 conversion. 3.2.3. DFT Predicted pKa’s and BDFEs The improved catalytic efficiency at significantly lower driving force warrants additional consideration. When using HBArF4 and KC8 we have previously suggested that protonation 52 of P3BFeN2-, which itself can be generated by reduction of P3BFeN2, to produce P3BFeN=NH is a critical first step; P3BFe-N=NH can then be trapped by acid to produce spectroscopically observable P3BFe=N-NH2+.16 These steps, shown in eq 3.2a-b, represent an ET-PT pathway. A PT-ET pathway, where P3BFeN2 is sufficiently basic to be protonated to generate P3BFe-N=NH+ as a first step, followed by ET, is also worth considering (eq 3.3a-b). A direct PCET pathway (eq 3.4) where H-atom delivery to P3BFeN2 occurs, thus obviating the need to access either P3BFeN2- or P3BFe-N=NH+, needs also to be considered. P3BFeN2 + e-  P3BFeN2- (3.2a) P3BFeN2- + H+  P3BFe-N=NH (3.2b) P3BFeN2 + H+  P3BFe-N=NH+ (3.3a) P3BFe-N=NH+ + e-  P3BFe-N=NH (3.3b) P3BFeN2 + H·  P3BFe-N=NH (3.4) Initial PT to P3BFeN2 to generate P3BFe-N=NH+ (eq 3.3a) is unlikely under the present conditions due to the high predicted acidity of P3BFe-N=NH+ (pKa = −3.7; estimated via DFT; See Appendix 2); efficient generation of such a species seems implausible for acids whose pKa’s are calculated at 1.4 (Ph2NH2+) and 6.8 (PhNH3+) in Et2O (Table 3.2). We note that [Ph2NH2][OTf] does not react productively with P3BFeN2 at -78 °C in Et2O, as analyzed by Mössbauer spectroscopy. Focusing instead on the PCET pathway (eq 3.4), the DFT-calculated BDEN-H for P3BFeN=NH (35 kcal/mol; Table 2; See Appendix 2 for details)20 is larger than the effective BDE9 of either Cp*2Co/Ph2NH2+ or Cp*2Co/PhNH3+ (25 and 31 kcal/mol, respectively). This suggests that PCET (eq 3.4) is plausible on thermodynamic grounds. Given that we have employed Cp*2Co in this study, and that this and also Cp2Co and Cp*2Cr have been 53 effective in other N2-fixing molecular catalyst systems,6,8 we have explored via DFT several putative metallocene-derived PCET reagents. Based on the analysis we describe below, we propose that protonated metallocenes may serve as discrete and highly active H· sources for PCET. Figure 3.3. (A) Calculated free-energy changes for the protonation of Cp*2Co. (B) DFT optimized structure of endo-Cp*Co(η4-C5Me5H)+ (methyl protons omitted for clarity). (C) The unfavorable reduction of 2,6-lutidinium by Cp*2Cr with the calculated free energy change. (D) The favorable protonation of Cp*2Cr by lutidinium with the calculated free energy change. Accordingly, we find that the formation of endo- and exo-Cp*Co(η4-C5Me5H)+ are predicted to be thermodynamically favorable via protonation of Cp*2Co by either Ph2NH2+ or PhNH3+ (−21 and −13 kcal/mol, respectively; Figure 3.3A).21 We have calculated the BDEC-H’s for both endo- and exo-Cp*Co(η4-C5Me5H)+ as 31 kcal/mol (Figure 3.3B; Table 54 2), indicating that they should be among the strongest PCET reagents accessible in this catalyst cocktail. Indeed, they would be among the strongest PCET reagents known.9 Table 3.2. Calculated pKa Values and BDEs of Selected Speciesa Species pKa BDEb Ph2NH2+ 1.4c - PhNH3+ 6.8 - Lutidinium 14.5 - 16.8 31 16.8 31 17.3 37 12.1 30 P3BFe-N=NH 38.7 35 P3BFe=N-NH2+ 14.4 51 P3BFe=N-NH2 - 47 [HIPTN3N]Mo-N=NH - 51 endo-Cp*Co(η4+ C5Me5H) exo-Cp*Co(η4C5Me5H)+ endo-Cp*Cr(η4C5Me5H)+ exo-Cp*Cr(η4C5Me5H)+ a Calculations were performed using the M06-L22 functional with a def2-TZVP basis set on Fe and a def2-SVP basis set on all other atoms23 (See Appendix 2). bIn kcal/mol. cpKa values were calculated in Et2O and reported relative to (Et2O)2H+. 55 We have also calculated the N–H bond strengths (Table 2) of several early stage candidate intermediates, including the aforementioned P3BFe-N=NH (35 kcal/mol), P3BFe=N-NH2+ (51 kcal/mol), and P3BFe=N-NH2 (47 kcal/mol). We conclude that PCET from Cp*Co(η4C5Me5H)+ to generate intermediates of these types is thermodynamically favorable in each case. To generate the first and most challenging intermediate (eq 3.5), the enthalpic driving force for PCET is estimated at ~4 kcal/mol (ΔGcalc = −9 kcal/mol). This driving force, and hence the plausibility of PCET steps, increases sharply as further downstream Fe-NxHy intermediates are considered.24 P3BFeN2 + Cp*Co(η4-C5Me5H)+  P3BFe-N=NH + Cp*2Co+ (3.5) Independent studies of H2 evolution from cobaltocene have invoked a protonated cobaltocene intermediate.25,26 The observation of a background H2 evolution reaction (HER) when employing metallocene reductants, but in the absence of an N2-to-NH3 conversion catalyst, suggests that metallocene protonation is kinetically competent.6c,27 3.3. Discussion Given the prevalence of metallocene reductants in N2-to-NH3 (or -N2H4) conversion,6,8 especially for the well-studied Mo catalyst systems, it is worth considering metallocenemediated PCET more generally. For instance, a role for ET/PT steps (or conversely PT/ET) in N2-to-NH3 conversion catalyzed by [HIPTN3N]Mo (HIPTN3N = [(3,5-(2,4,6i Pr3C6H2)2C6H3NCH2CH2)3N]3-, a bulky triamidoamine ligand) has been frequently posited.27 But PCET steps may play a critical role, too. In the latter context, we note reports from Schrock and coworkers that have shown both acid and reductant are required to 56 observe productive reactivity with [HIPTN3N]MoN2. These observations are consistent with PCET to generate [HIPTN3N]Mo-N=NH.28e A PCET scenario has been discussed in this general context of N2-to-NH3 conversion, where a lutidinyl radical intermediate formed via ET from Cp*2Cr has been suggested as a PCET reagent that can be generated in situ.27,29 However, our own calculations predict that the lutidinyl radical should not be accessible with Cp*2Cr as the reductant (∆Gcalc = +10 kcal/mol; Figure 3C).30 We instead propose protonation of Cp*2Cr by the lutidinium acid as far more plausible (∆Gcalc = −5.3 kcal/mol; Figure 3D) to generate a highly reactive decamethylchromocene-derived PCET reagent. While N–H bond strengths have not been experimentally determined for the [HIPTN3N]Mo-system, using available published data we deduce the N–H bond of [HIPTN3N]Mo-N=NH to be ca. 49 kcal/mol and we calculate it via DFT (truncated HIPTN3N; See Appendix 2) as 51 kcal/mol.31 The BDEN-H for this Mo diazenido species is hence much larger than we predict for P3BFe-N=NH (35 kcal/mol), perhaps accounting for its higher stability.28e A PCET reaction between endo-Cp*Cr(η4-C5Me5H)+ (BDEcalc = 37 kcal/mol) and [HIPTN3N]MoN2 to generate [HIPTN3N]Mo-N=NH and Cp*2Cr+ would be highly exergonic. Furthermore, we predict a similarly weak BDEC-H for Cp-protonated cobaltocene, CpCo(η4-C5H6)+ (BDEcalc = 35 kcal/mol). These considerations are consistent with the reported rapid formation of [HIPTN3N]Mo-N=NH using either Cp*2Cr or Cp2Co in the presence of lutidinium acid.32 3.4 Conclusions To close, we have demonstrated catalytic N2-to-NH3 conversion by P3BFe+ at a much lower driving force (nearly 100 kcal/mol) than originally reported via combination of a weaker 57 reductant (Cp*2Co) and acid ([Ph2NH2][OTf] or [Ph3NH][OTf]). Significantly improved efficiency for NH3 formation is observed (up to 72% at standard substrate loading), and by reloading additional substrate at low temperature the highest turnover number yet observed for any synthetic molecular catalyst (84 ± 8 equiv NH3 per Fe) has been achieved. Freezequench Mössbauer spectroscopy under turnover conditions reveals differences in the speciation of P3BFe compared to previous studies with HBArF4 and KC8, suggesting changes in the rates of key elementary steps. Using DFT calculations we have considered the viability of a decamethylcobaltocene-mediated PCET pathway as an alternative to previously formulated ET-PT and PT-ET pathways. Based on our calculations, we propose that protonated metallocenes should serve as discrete, very reactive PCET reagents in N 2to-NH3 conversion catalysis. Indeed, the achievement of high efficiency for N2-to-NH3 conversion by both P3BFe and various Mo catalysts that benefit from metallocene reductants raises the intriguing possibility that metallocene-based PCET reactivity is a potentially widespread and overlooked mechanism. Efforts are underway to further explore such pathways. 3.5. References 1 Smil, V. Enriching the Earth. Cambridge: MIT Press, 2001. 2 van der Ham, C. J.; Koper, M. T.; Hetterscheid, D. G. Chem. Soc. Rev. 2014, 43, 5183. 3 (a) Shaver, M. P.; Fryzuk, M. Adv. Synth. Catal. 2003, 345, 1061. (b) Macleod, K. C.; Holland, P. L. Nat. Chem. 2013, 5, 559. 4 Hoffman, B. M.; Lukoyanov, D.; Yang, D. Y.; Dean, D. R.; Seefeldt, L. C. Chem. Rev. 2014, 114, 4041. 58 5 (a) Anderson, J. S.; Rittle, J.; Peters, J. C. Nature 2013, 501, 84. (b) Creutz, S.; Peters, J. C. J. Am. Chem. Soc. 2014, 136, 1105. (c) Ung, G.; Peters, J. C. Angew. Chem. Int. Ed.2015, 54, 532. (d) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Matsuo, Y.; Tanaka, H.; Ishii, K.; Yoshizawa, K.; Nishibayashi, Y. Nat. Commun. 2016, 7, 12181. 6 (a) Yandulov, D. V.; Schrock, R. R. Science 2003, 301, 76. (b) Arashiba, K.; Miyake, Y.; Nishibayashi, Y. Nat. Chem. 2011, 3, 120. (c) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Tanaka, H.; Kamaru, N.; Yoshizawa, K.; Nishibayashi, Y. J. Am. Chem. Soc. 2014, 136, 9719. (d) Arashiba, K.; Kinoshita, E.; Kuriyama, S.; Eizawa, A.; Nakajima, K.; Tanaka, H.; Yoshizawa, K.; Nishibayashi, Y. J. Am. Chem. Soc. 2015, 137, 5666. 7 The enthalpic driving force (ΔΔHf) has been estimated here by taking 3*(BDEH·−BDEeffective), where BDE is bond dissociation enthalpy. This allows for an evaluation of the driving force for a given reaction with respect to that for a hypothetical N2to-NH3 conversion catalyst that uses H2 as the proton and electron source. This is achieved by using Bordwell’s equation (with the assumption that S(X·) = S(XH); See Appendix 2) and literature values for pKa, redox potential, the enthalpy of reaction for H+ + e-  H· (CH = 66 kcal/mol in THF), and the energy of H· in THF (52 kcal/mol): (a) Bordwell, F. G.; Cheng, J.P.; Harrelson, Jr., J. A. J. Am. Chem. Soc. 1988, 110, 1229. (b) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877. (c) Garido, G.; Rosés, M.; Ràfols, C.; Bosch, E. J. Sol. Chem. 2008, 37, 689. (d) Kaljurand, I.; Kütt, A.; Sooväli, L.; Rodima, T.; Mäemets, V.; Leito, I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019. (e) Cappellani, E. P.; Drouin, S. D.; Jia, G.; Maltby, P. A.; Morris, R. H.; Schweitzer, C. T. J. Am. Chem. Soc. 1994, 116, 3375. 8 While initiating our studies we became aware of a phosphine-supported Fe system that catalyzes N2-to-N2H4 conversion using Cp*2Co and [Ph2NH2][OTf] with efficiency as high 59 as 72% for e- delivery to N2: Hill, P. J.; Doyle L. R.; Crawford A. D.; Myers W. K.; Ashley, A. E. J. Am. Chem. Soc. 2016, 138, 13521. 8 Warren, J. J.; Tronic, T. A.; Mayer, J. M. Chem. Rev. 2010, 110, 6961. 9 Anderson, J. S. Catalytic conversion of nitrogen to ammonia by an iron model complex. Ph.D. Thesis, California Institute of Technology, September 2013. 10 Del Castillo, T. J.; Thompson, N. B.; Peters, J. C. J. Am. Chem. Soc. 2016, 138, 5341. 11 Previously reported molecular Fe catalysts for N2-to-NH3 conversion utilize KC8 and HBArF4 and achieve NH3 selectivities ≤ 45% with respect to their limiting reagent (see ref 5) at a similar reductant loading. Lower selectivities are observed with higher loading (see refs Error! Bookmark not defined.a and Error! Bookmark not defined.). 12 In catalytic runs performed with labeled [Ph215NH2][OTf] under an atmosphere of natural abundance 14N2 the production of exclusively 14NH3 is observed, demonstrating that the NH3 formed during catalysis is derived from N2 and not degradation of the acid (See Appendix 2). 13 (a) Whited, M. T.; Mankad, N. P.; Lee, Y.; Oblad, P. F.; Peters, J. C. Inorg. Chem. 2009, 48, 2507. (b) Del Castillo, T. J.; Thompson, N. B.; Suess, D. L.; Ung, G.; Peters, J. C. Inorg. Chem. 2015, 54, 9256. 14 Rittle, J.; Peters, J. C. J. Am. Chem. Soc. 2016, 138, 4243. 15 Anderson, J. S.; Cutsail, 3rd, G. E.; Rittle, J.; Connor, B. A.; Gunderson, W. A.; Zhang, L.; Hoffman, B. M.; Peters, J. C. J. Am. Chem. Soc. 2015, 137, 7803. 16 The presence of P3BFeN2- was confirmed by freeze-quench EPR spectroscopy experiments (See Appendix 2). The asymmetry observed in the Mössbauer lineshapes is characteristic of this species. A redox equilibrium between P3BFeN20/- and Cp*2Co+/0 is also 60 observed in the reaction of P3BFe+ with excess Cp*2Co in the absence of acid (See Appendix 2). 17 The distinct properties of tetrahedral, high spin Fe(II) leads to high isomer shifts (0.9- 1.3) and large quadrupole splittings (> 2.5) that are characteristic of these types of species: (a) E. Münck, in Physical Methods in Bioinorganic Chemistry: Spectroscopy and Magnetism (Ed.: L. Que Jr.), University Science Books, Sausalito, CA, 2000, pp. 287-320. (b) Daifuku, S. L.; Kneebone, J. L.; Snyder, B. E.; Neidig, M. L. J. Am. Chem. Soc. 2015, 137, 11432. 18 Hamashima, Y.; Somei, H.; Shimura, Y.; Tamura, T.; Sodeoka, M. Org. Lett. 2004, 6, 1861. 19 Experimental BDEN-H’s for related species (P3SiFe-C=NH+, P3SiFe-C=NH, P3SiFe- C=N(Me)H+, P3SiFe-C=N(Me)H, and P3SiFe-N=N(Me)H+) have been measured and are in good agreement with the BDEN-H values calculated using the DFT methods described in this work (See Appendix 2 for full details): Rittle, J.; Peters, J. C. manuscript submitted for publication. 20 Note: Efforts to instead optimize a [Cp*2Co-H]+ species led to hydride transfer to the ring system. 21 Zhao Y.; Truhlar D. G. J. Chem. Phys. 2006, 125, 194101. 22 Weigend F.; Ahlrichs R. Phys. Chem. Chem. Phys. 2005, 7, 3297. 23 Studies have shown that the Marcus cross-relation holds quite well for many PCET reactions. This is indicative of a substantial correlation between thermodynamic driving force and reaction kinetics; it is, however, unclear whether the proposed reactivity would demonstrate such behavior: (a) Roth, J. P.; Yoder, J. C.; Won, T.–J.; Mayer, J. M. Science, 61 2001, 294, 2524. (b) Mayer, J. M.; Rhile, I. J. Biochim. Biophys. Acta, Bioenerg. 2004, 1655, 51. (c) Hammes-Schiffer, S. Acc. Chem. Res. 2001, 34, 273. 24 Koelle, U.; Infelta, P. P.; Grätzel, M. Inorg. Chem. 1988, 27, 879. 25 For recent studies relevant to protonated metallocenes in the context of HER see: (a) Pitman, C. L.; Finster, O. N. L.; Miller, A. J. M. Chem. Commun. 2016, 52, 9105. (b) Aguirre Quintana, L. M.; Johnson, S. I.; Corona, S. L.; Villatoro, W.; Goddard, 3rd, W. A.; Takase, M. K.; VanderVelde, D. G.; Winkler, J. R.; Gray, H. B.; Blakemore, J. D. Proc. Natl. Acad. Sci. U.S.A. 2016, 113, 6409. 26 Munisamy, T.; Schrock, R. R. Dalton Trans. 2012, 41, 130. 27 (a) Studt, F.; Tuczek, F. Angew. Chem. Int. Ed. 2005, 44, 5639. (b) Reiher, M.; Le Guennic, B.; Kirchner, B. Inorg. Chem. 2005, 44, 9640. (c) Studt, F.; Tuczek, F. J. Comput. Chem. 2006, 27, 1278. (d) Thimm, W.; Gradert, C.; Broda, H.; Wennmohs, F.; Neese, F.; Tuczek, F. Inorg. Chem. 2015, 54, 9248. (e) Yandulov, D. V.; Schrock, R. R. Inorg. Chem. 2005, 44, 1103. 28 Pappas, I.; Chirik, P. J. J. Am. Chem. Soc. 2016, 138, 13379. 29 Although our calculations for a hypothetical lutidinyl radical predict a weak N–H bond (BDEN-H ~ 35 kcal/mol), the oxidation potential of this species is calculated to be −1.89 V vs Fc+/0 in THF (See Appendix 2). Experimental determination of this reduction potential for calibration has been contentious; however, our calculated reduction potential is similar to that previously calculated for pyridinium in aqueous media (−1.37 V vs SCE): (a) Yan, Y.; Zeitler, E. L.; Gu, J.; Hu, Y.; Bocarsley, A. B. J. Am. Chem. Soc. 2013, 135, 14020. (b) Keith, J. A.; Carter, E. A. J. Am. Chem. Soc. 2012, 134, 7580. 62 30 It has been reported that [HIPTN3N]MoN2-/[HIPTN3N]Mo-N=NH is in equilibrium with DBU/DBUH+ (DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene; pKa = 18.5 in THF; see refs 7c-d). Taken with the reported reduction potential of [HIPTN3N]MoN2 (E1/2 = −1.81 V vs Fc+/0 in THF, see ref 28e), the experimental BDE can be approximated with the Bordwell equation and the enthalpy of reaction for H+ + e-  H· (see ref 7). 31 In addition to lutidinium salts, [Et3NH][OTf] has been shown to affect the formation of [HIPTN3N]Mo-N=NH from [HIPTN3N]MoN2 in the presence of metallocene reductants (see ref 28e). 63 Chapter 4. Fe-Mediated Nitrogen Fixation with a Metallocene Mediator: Exploring pKa Effects and Demonstrating Electrocatalysis 64 4.1 Introduction There has been substantial recent progress in the development of soluble, welldefined molecular catalysts for N2-to-NH3 conversion, commonly referred to as the nitrogen reduction reaction (N2RR).1 Nevertheless, a significant and unmet challenge is to develop molecular catalysts, and conditions, compatible with electrocatalytic N2RR. Progress in this area could have both fundamental and practical benefits, including access to informative in situ mechanistic studies via electrochemical techniques, and an electrochemical means to translate solar or otherwise derived chemical currency (H+/e-) into NH3. The latter goal, which has been the subject of numerous studies using heterogeneous catalysts, is key to the long-term delivery of new ammonia synthesis technologies for fertilizer and/or fuel.2 Many soluble coordination complexes are now known that electrocatalytically mediate the hydrogen evolution reaction (HER),3 the carbon dioxide reduction reaction (CO2RR),4 and the oxygen reduction reaction (ORR).5 The study of such systems has matured at a rapid pace in recent years, coinciding with expanded research efforts towards solar-derived fuel systems. In this context, it is noteworthy how little corresponding progress has been made towards the discovery of soluble molecular catalysts that mediate electrocatalytic N2RR. To our knowledge, only two prior systems address this topic directly.6-8 Pickett and coworkers reported, more than three decades ago, that a Chatt-type tungsten-hydrazido complex (W=NNH2) could be electrochemically reduced to release ammonia (and trace hydrazine), along with some amount of a reduced W–N2 product; the latter species serves as the source of the W=NNH2 species (via its protonation by acid).6a 65 By cycling through such a process, an electrochemical, but not an electrocatalytic, synthesis of ammonia was demonstrated. Indeed, efforts to demonstrate electrocatalysis with this and related system instead led to substoichiometric NH3 yields.6c An obvious limitation to progress in electrochemical N2RR by molecular systems concerns the small number of synthetic N2RR catalysts that have been available for study; it is only in the past five years that sufficiently robust catalyst systems have been identified to motivate such studies. In addition, the conditions that have to date been employed to mediate N2RR have typically included non-polar solvents, such as heptane, toluene, and diethyl ether (Et2O), that are not particularly well-suited to electrochemical studies owing to their low conductivity.1 Nevertheless, several recent developments, including by our lab, point to the likelihood that iron (and perhaps other) molecular coordination complexes may be able to mediate electrocatalytic N2RR in organic solvent. Specifically, our lab recently reported that a tris(phosphine)borane iron complex, P3BFe+, that is competent for catalytic N2RR, can also mediate electrolytic N2-to-NH3 conversion,6d with the available data (including this study) pointing to bona fide electrocatalysis in Et2O. Focusing on the P3BFe+ system we have studied, a development relevant to the current study was its recently demonstrated compatibility with reagents milder than those originally employed.1c Thus decamethylcobaltocene (Cp*2Co) and diphenyl ammonium acid (Ph2NH2+) are effective; these reagents give rise to fast, and also quite selective (>70% vs HER) N2RR catalysis at low temperature and pressure in ethereal solvent. In addition, based on preliminary spectroscopic evidence and density functional theory (DFT) predictions, it appears that a protonated metallocene species, Cp*(η4-C5Me5H)Co+, may be 66 an important intermediate of N2RR catalysis under such conditions. Indeed, we have suggested that Cp*(η4-C5Me5H)Co+ may be an effective proton-coupled-electron-transfer (PCET) donor (BDEC–H(calc) = 31 kcal/mol), thereby mediating net H-atom transfers to generate N−H bonds during N2RR. The presence of a metallocene mediator might, therefore, enhance N2RR during electrocatalysis.9 We present here a study of the effect of pKa on the selectivity of P3BFe+ for N2RR vs HER. By using substituted anilinium acids, we are able to vary the acidity over 9 orders of magnitude and find that the selectivity is highly correlated with the pKa. In our efforts to investigate the origin of the observed pKa effect, we found to our surprise that the catalytically relevant acids were unable to faciltate productive early-stage N–H bond formation in stoichiometric reactions and therefore hypothesized that the formation of a protonated metallocene species Cp*(η4-C5Me5H)Co+ indeed played a critical role in N–H bond-forming reactions either via PCET or PT during N2RR catalysis. DFT studies support this hypothesis and also establish that the observed pKa effect can be explained by the varying ability of the acids to protonate Cp*2Co. The critical role of this protonated metallocene intermediate in N–H bond forming reactions led us to test the effect of Cp*2Co+ as an additive in the electrolytic synthesis of NH3 by P3BFe+. We found that the addition of co-catalytic Cp*2Co+ enhances both yield of NH3 and Faradaic efficiency (FE), and thus furnishes the first unequivocal demonstration of electrocatalytic N2RR with a soluble, molecular coordination complex. 67 4.2. Results and Discussion 4.2.1 pKa studies. In our recent study on the ability of P3BFe+ to perform N2RR with Cp*2Co as the chemical reductant, we found that there was a marked difference in efficiency for NH3 demonstrated by diphenylammonium triflate ([Ph2NH2][OTf]) and anilinium triflate ([PhNH3][OTf]).Error! Bookmark not defined. In that study we posited that this difference could a rise from a variety of sources including the differential solubility, sterics, or pKa’s of these acids. To better investigate this last possibility we explored the efficiency of the catalysis by quantifying the NH3 and H2 produced when using substituted anilinium acids with different pKa values (Table 4.1). The table is organized in descending acid strength, from [4-OMePhNH3][OTf] as the weakest acid to the perchlorinated derivative ([perCl PhNH3][OTf]) as the strongest. Importantly, good total electron yields (85.8 ± 3.3) were obtained in all cases. As can be seen from the table, the NH3 efficiencies are found to be strongly correlated with pKa.10 In particular, a comparison of the efficiency for NH3 with the pKa of the anilinium acid used gives rise to four distinct activity regimes (Table 4.1). Firstly, a regime that is completely inactive for N2RR, but active for HER, is defined by the weakest acid [4OMe PhNH3][OTf] (pKa = 8.9).11 A gradual increase in observed NH3 yields, coupled with a decreased H2 yield, comprises a second regime in which the acid is strengthened from [PhNH3][OTf] (pKa = 7.8), to [2,6-MePhNH3][OTf] (pKa = 6.8), to [2-ClPhNH3][OTf] (pKa = 5.6). Yet stronger acids, [2,5-ClPhNH3][OTf] (pKa = 4.1), [2,6-ClPhNH3][OTf] (pKa = 3.4), and [2,4,6-ClPhNH3][OTf] (pKa = 2.1), provide the third, most active N2RR regime in which the H2 yields are nearly invariant.12 The highest selectivity for N2RR (~78%) was observed 68 using [2,5-ClPhNH3][OTf] as the acid. A fourth regime of very low N2RR activity is encountered with [per-ClPhNH3][OTf] (pKa = 1.3) as the acid. We suspect this last acid undergoes unproductive reduction via ET, thereby short-circuiting N2RR. Intriguingly the observed behavior is remarkably similar to that for nitrogenase which is the only other N2RR system for which such pKa effects have been well-studied (Figure 4.4.1).13,14 Figure 4.1. (top) Percentage of electrons being used to form NH3 and H2 at different pH values by the FeMo-nitrogenase in A. vinelandii. Reprinted with permission from Pham, D. N.; Burgess, B. K. Biochemistry 1993, 32, 13725. Copyright 1993 American Chemical 69 Society. (bottom) Percentage of electrons being used to form NH3 and H2 at different pKa values by P3BFe+. In our previous study on Cp*2Co mediated N2RR with P3BFe+, we had identified that P3BFeN2− forms under catalytic conditions. Earlier studies on the reactivity of P3BFeN2− with excess of soluble acids such as HOTf or [H(OEt2)2][BArF4] (HBArF4, BArF4 = tetrakis(3,5-bis(trifluoromethyl)phenyl)borate)) at low temperature in Et2O had determined that P3BFe=NNH2+ was rapidly formed.15 Furthermore, recent computational work from our group posits that under catalytic conditions with a soluble acid the disparate efficiency for N2RR demonstrated by P3EFe catalysts (E = B, C, Si) is determined by the rate of formation and consumption of early N2RR intermediates (i.e. P3EFe=NNH and P3EFe=NNH2+/0).16 Thus we were interested in the reactivity of these anilinium triflate acids with P3BFeN2− and hypothesized that they may show differential efficiency in the formation of P3BFe=NNH2+. Table 4.1. Literature10 and calculateda,b pKa values and efficiencies observed in catalytic N2-to-NH3 conversion pKaexp pKacalc pKdcalc NH3/Fe %NH3/e- %H2/e-c [4-OMePhNH3][OTf] 8.9 9.6 15.7 0.04 ± .01 0.2 ± 0.1 89.1 ± 0.2 [PhNH3][OTf] 7.8 7.7 13.8 7.3 ± 0.1 40.4 ± 0.5 48.6 ± 0.7 [2,6-MePhNH3][OTf] 6.8 7.3 13.2 8.6 ± 0.7 47.5 ± 4.0 37.8 ± 0.2 N/A 9.2 11.8 ― ― 5.6 5.6 6.0 10.7 ± 0.1 53.9 ± 0.4 26.1 ± 1.9 [Cp*(exo-η4― C5Me5H)Co][OTf] [2-ClPhNH3][OTf] 70 [2,5-ClPhNH3][OTf] 4.1 4.0 5.0 13.9 ± 0.7 [2,6-ClPhNH3][OTf] 3.4 3.4 3.4 13.8 ± 0 .9 76.7 ± 4.9 12.6 ± 2.5 [2,4,6-ClPhNH3][OTf] 2.1 2.7 1.8 12.8 ± 0.4 70.9 ± 2.2 12.0 ± 0.8 [per-ClPhNH3][OTf] 1.3 0.8 0.4 3.6 ± 0.1 19.9 ± 0.5 63.5 ± 1.1 77.5 ± 3.8 10.5 ± 1.1 [4-OMePhNH3][OTf] = 4-methoxyanilinium triflate, [PhNH3][OTf] = anilinium triflate, [2,6Me PhNH3][OTf] = 2,6-dimethylanilinium triflate, [2-ClPhNH3][OTf] = 2-chloroanilinium triflate, [2,5-ClPhNH3][OTf] = 2,5-dichloroanilinium triflate, [2,6-ClPhNH3][OTf] = 2,6dichloroanilinium triflate, [2,4,6-ClPhNH3][OTf] = 2,4,6-trichloroanilinium triflate, [perCl PhNH3][OTf] = 2,3,4,5,6-pentachloroanilinium triflate. aAcidities calculated at 298 K in THF and referenced to the known literature value for [2,6-ClPhNH3][OTf]. bAll species calculated as the ion-paired OTf− species in Et2O at 195 K and referenced to the known literature value for [2,6-ClPhNH3][OTf] in THF. To our surprise the freeze-quench EPR spectrum of the reaction of excess [2,6PhNH3][OTf] (high efficiency regime) at −78 °C in Et2O with P3BFeN2− did not reveal Cl any P3BFe=NNH2+. In accord with this result, freeze-quench Mössbauer experiments show only the formation of oxidized Fe products, namely P3BFeN2 and P3BFe+. Finally, analysis of such reactions for NH3 or N2H4 after warming led to the observation of no fixed-N products. The observation of exclusively oxidation rather than productive N−H bond formation is reminiscent of experiments in which only 1 equiv of a soluble acid source (HBArF4 or HOTf) is added to P3BFeN2−. In these cases, we have proposed that the unstable P3BFe=NNH is formed and without excess acid to trap it as P3BFe=NNH2+ it decays with the loss of 1/2 an equivalent of H2 to form P3BFeN2. 71 Although large excesses of the triflate acids (25 eq) were employed in these experiments to mimic catalytic conditons, the low solubility of these anilinium triflate acids under the catalytically relevant conditions (Et2O, -78 °C) likely leads to a situation in which the formed P3BFe=NNH is not sufficienctly rapidly captured by acid and hence decays with loss of H2. Further evidence in support of this hypothesis comes from experiments employing [2,6-ClPhNH3][BArF4] as a soluble source of acid. In this case, freeze-quench EPR of the reaction between P3BFeN2− and 25 eq of acid leads to the observation of P3BFe=NNH2+ and the quantification of fixed-N products upon warming (0.20 ± 0.04 eq. NH3 per Fe). While internally consistent, these observations must be reconciled with the seemingly contradictory observation of highly efficient N2RR when [2,6-ClPhNH3][OTf] and other anilinium triflate acids are employed under catalytic conditions. In fact, we have previously observed that [Ph2NH2][OTf] leads to superior efficiencies for NH3 than observed with [Ph2NH2][BArF4] (72 ± 3% and 42 ± 6% respectively). An obvious difference between the stoichiometric reactions described above and the catalytic reaction is the presence of Cp*2Co in the latter. We had previously postulated that Cp*2Co could be protonated under the catalytic reaction conditions to form Cp*(η4-C5Me5H)Co+. This species could then play a role in N−H bond forming reactions. The results herein suggest that such a mechanism is not only plausible but likely necessary to explain the observed catalytic results with anilinium triflate acids. Given the effect of pKa on the efficiency for N2RR, we hypothesized that this effect might arise from the thermodynamics or kinetics of Cp*2Co protonation by the different anilinium acids. 72 4.2.2. Computational Studies To investigate the kinetics and thermodynamics of Cp*2Co protonation by anilinium triflate acids we employed DFT. DFT-D317calculations were performed at the TPSS/def2-TZVP(Fe); def2-SVP18 level of theory that our group has previously successfully employed in studies of this system.19 We calculated the free energy of H+ exchange (ΔGa) for all of the used acids (one example in eq 4.1) and Cp*(exo-η4C5Me5H)Co+ in Et2O at 298 K. These free energies were then used to determine the pKa by including a term to adjust them to the literature pKa value for [2,6-ClPhNH3] [OTf] at 298 K in THF (eq 4.2). PhNH2 + 2,6-ClPhNH3+  PhNH3+ + 2,6-ClPhNH2 (eq 4.1) pKa(PhNH3+) = −ΔGa/(2.303×RT) + pKa(2,6-ClPhNH3+) (eq 4.2) However, because we believe that the variable triflate hydrogen bonding effects (0.5–10 kcal/mol) are likely important under the catalytic conditions (low temperature and low polarity solvent), we additionally calculated the free energy for net HOTf exchange reactions (ΔGd) at 195 K in Et2O (one example in eq 4.3). The free energy of these reactions can then be used to determine a pKd, which were also referenced to the pKa value for [2,6Cl PhNH3][OTf] at 298 K in THF for ease of comparison (eq 4.4). Going forward we will use the pKd values in our discussion, but using the pKa values would not substantively alter the conclusions. PhNH2 + [2,6-ClPhNH3][OTf]  [PhNH3][OTf] + 2,6-ClPhNH2 (eq 4.3) pKd([PhNH3][OTf]) = −ΔGd/(2.303×RT) + pKa(2,6-ClPhNH3+) (eq 4.4) 73 Calculations of the pKd of all of the relevant species (Table 4.1) leads to the observation that the pKd of [Cp*(η4-C5Me5H)Co][OTf] (pKdcalc = 11.8; Table 4.1), falls within the range of acids studied (0.4 ≤ pKdcalc ≤15.7; Table 4.1) suggesting there would be a significant acid dependence on the kinetics and thermodynamics of Cp*2Co protonation. To better elucidate the differences in Cp*2Co protonation between the acids, we investigated in detail the kinetics for three acids, [2,6-ClPhNH3][OTf] (high selectivity; pKdcalc = 3.4), [2,6-MePhNH3][OTf] (modest selectivity; pKdcalc = 13.2), and [4OMe PhNH3][OTf] (poor selectivity; pKdcalc = 15.8). Transition states were readily located for all three acids and as can be seen in Figure 4.2, protonation of Cp*2Co is found to have only a moderate barrier in all three cases ([4-OMePhNH3][OTf], [2,6-MePhNH3][OTf] and [2,6PhNH3][OTf]: ΔG‡ = +4.5 kcal/mol, +3.8 kcal/mol and +1.3 kcal/mol, respectively). This Cl suggests that Cp*2Co protonation is kinetically accessible in all cases, in agreement with the experimental observation of background HER with all of these acids (see SI). The small differences in rate, and the large variance in the equilibrium constant defined in eq 4.5 (Keq, Figure 4.2) illustrates the significant difference in the population of protonated metallocene for these different acids. K eq = [ RPhNH2 −Cp∗ (𝜂 4 −𝐶5 𝑀𝑒5 H)Co−OTf] [OTf−H3 N RPh−Cp∗2 Co] (eq 4.5) The low solubility of the anilinium triflate acids and the low catalyst concentration leads to a situation in which the interaction between the acid and the Cp*2Co significantly affects the kinetics of productive N−H bond formation. As such, the difference in [Cp*(η4C5Me5H)Co][OTf] population and formation rate is the origin of the observed pKa effect, rather than differences in rates involving the direct interaction of a P3BFe species with acid. 74 Figure 4.2. The kinetics and thermodynamics of protonation of Cp*2Co for three acids from different catalytic efficiency regimes. As we discussed in our previous study, [Cp*(η4-C5Me5H)Co][OTf] is a strong PCET donor.20 This reaction can occur either in a synchronous fashion akin to HAT or in an asynchronous fashion that approaches a PT-ET reaction.21 We believe that this reagent is likely effective in a variety of N−H bond forming reactions under the catalytic conditions. However, the experimental results suggest that this reagent likely plays a critical role in trapping the unstable P3BFe=NNH (Figure 4.3) before it can decompose. Indeed, both a synchronous PCET reaction (ΔGPCET = −17.3 kcal/mol; eq 4.6), and the two individual steps of an asynchronous PCET reaction (ΔGPT = −5.7 kcal/mol, ΔGET = −11.6 kcal/mol; eq 4.7-4.8) are found to be thermodynamically favorable. P3BFe=NNH + [Cp*(η4-C5Me5H)Co][OTf]  P3BFe=NNH2 + [Cp*2Co][OTf] (eq 4.6) P3BFe=NNH + [Cp*(η4-C5Me5H)Co][OTf]  P3BFe=NNH2+ + Cp*2Co (eq 4.7) P3BFe=NNH2+ + Cp*2Co  P3BFe=NNH2 + [Cp*2Co][OTf] (eq 4.8) 75 To evaluate the kinetics of these reactions the Marcus theory expressions22 and the Hammes-Schiffer method23 were used to approximate relative rates of bimolecular ET and PCET. We find that there is a slight kinetic preference for the fully synchronous PCET reaction (krelPCET ~ 2x103 – 4.5x103 kcal/mol) compared to the fully asynchronous PT-ET reaction (krelET ≡ 1 M-1s-1; Figure 4.3) but both reaction mechanisms appear viable.24 Figure 4.3. The calculated thermodynamics and kinetics of the proposed synchronous PCET and asynchronous PCET (PT–ET) reaction between P3BFeNNH and [Cp*(η4C5Me5H)Co][OTf]. This leads to the conclusion that the efficiency for NH3 formation in this system is coupled to the kinetics and/or thermodynamics of the reaction between the acid and reductant. As the protonation of the reductant is also the first step on the background HER reaction25 this conclusion is somewhat counterintuitive. The fact that an HER intermediate 76 can be intercepted and used for productive N2RR steps is a potentially important design principle in the catalysis. Furthermore, this expanded role for metallocenes opens up an exciting new avenue of research into these reagents and the potentially multifaceted role that they play in the proton-coupled reduction of a variety of small molecule substrates. In N2RR systems, efforts are often made to suppress the ‘background’ reaction between the acid and reductant.1a-b We were thus particularly curious whether the inclusion of a metallocene co-catalyst could improve the yield and Faradaic efficiency (FE) for NH3 formed in controlled potential electrolysis (CPE) experiments. 4.2.3. Electrolysis studies. In our previous study, we had shown that ~2.2 equiv NH3 (per Fe) could be generated via controlled potential electrolysis (CPE; −2.3 V vs Fc+/0) at a reticulated vitreous carbon working electrode, using P3BFe+ as the (pre)catalyst in the presence of HBArF4 (50 equiv) at −45 °C under an atmosphere of N2. This yield of NH3 corresponds to a ~25% FE which, while modest in terms of overall chemoselectivity, compares favorably to FE’s typically reported for heterogeneous electrocatalysts for N2RR that operate below 100 °C (< 2%).2,26 To further explore the possibility of using P3BFe+ as an electrocatalyst for N2RR, we have since surveyed various conditions to determine whether, from CPE experiments enhanced yields of NH3 could be obtained. For example, screening various applied potentials (ranging from −2.1 to −3.0 V vs Fc+/0), varying the concentrations of P3BFe+ and HBArF4, varying the ratio of acid to catalyst, and varying the rate at which acid was delivered to the system (initial loading, batch-wise addition, reloading, or continuous slow 77 addition) all led to no substantial improvement (<2.5 eq NH3 obtained per Fe). Attempts to vary the ratio of electrode surface area to working compartment solution volume, either by employing smaller cell geometries or using different morphologies of glassy carbon as the working electrode (reticulated porous materials of different pore density or plates of different dimensions) also yielded no substantial improvement of NH3 yield with respect to Fe. The replacement of HBArF4 with 50 equiv of [Ph2NH2][OTf] as the acid led to similar yields of NH3 (Table 4.2, entry 1). To investigate the potential effect of the Cp*2Co+ additive, a systematic cyclic voltammetry study was undertaken. Traces of cyclic voltammograms (Figure 4.4) are provided for [Ph2NH2][OTf] (both panels, gray trace), Cp*2Co+ (panel A, yellow trace), Cp*2Co+ with the addition of ten equiv of [Ph2NH2][OTf] (panel A, green trace), P3BFe+ (panel B, dark blue trace), P3BFe+ with the addition of ten equiv of [Ph2NH2][OTf] (panel B, light blue trace), and P3BFe+ with the addition of five equiv of Cp*2Co+ and ten equiv of [Ph2NH2][OTf] (both panels, red trace). Figure 4.4. A) Cyclic voltammograms of 10 equiv [Ph2NH2][OTf] (gray trace), 5 equiv [Cp*2Co][BArF4] (Cp*2Co+) (yellow trace), 5 equiv Cp*2Co+ with 10 equiv [Ph2NH2][OTf] (green trace), and 1 equiv P3BFe+ with 5 equiv of Cp*2Co+ and 10 equiv [Ph2NH2][OTf] (red trace). B) Cyclic voltammograms of 10 equiv [Ph2NH2][OTf] (gray trace), 1 equiv 78 P3BFe+ (dark blue trace), 1 equiv P3BFe+ with 10 equiv [Ph2NH2][OTf] (light blue trace), and P3BFe+ with 5 equiv of Cp*2Co+ and 10 equiv [Ph2NH2][OTf] (red trace). All spectra are collected in 0.1 M NaBArF4 solution in Et2O at -35 °C using a glassy carbon working electrode, and externally referenced to the Fc+/0 couple. Scan rate is 100 mV/s. CPE studies were undertaken to characterize the reduction products associated with the red trace at ~−2.1 V vs Fc+/0. These studies employed a glassy carbon plate electrode, a Ag/Ag+ reference electrode that was isolated by a CoralPorTM frit and referenced externally to the ferrocene/ferricinium redox couple (Fc+/0), and a solid sodium auxiliary electrode.27 Unless otherwise noted, CPE experiments were performed at −2.1 V versus Fc+/0, again with 0.1 M NaBArF4 as the ether-soluble electrolyte, at −35 °C under an atmosphere of N2, electrolysis was generally continued until the current measurement dropped to 1% of the initial current measured or until 21.5 hours had passed. The Supporting Information provides additional details. CPE experiments were conducted with the inclusion of 0, 1, 5, and 10 equiv of Cp*2Co+ with respect to P3BFe+, using excess [H2NPh2][OTf] as the acid. In the absence of added Cp*2Co+, a significant amount of NH3 was generated (2.4 equiv per Fe), consistent with the previous finding, in the presence of a strong acid, that P3BFe+ can electrolytically mediate N2-to-NH3 conversion.6d We found that inclusion of 1.0 equiv of Cp*2Co+ significantly enhanced the NH3 yield, by a factor of 1.5 and led to improvements in the FE (Table 4.2, entry 2). The data provide a total yield, with respect to both Fe and Co, that confirm modest, but unequivocal, N2RR electrocatalysis. The best NH3 yield we have observed, in a single-run experiment, was 4.4 equiv per Fe. 79 Increasing the amount of added Cp*2Co+ did not affect the NH3 yield (entry 3, 5). However, the addition of a second loading of [Ph2NH2][OTf] following the first electrolysis (entry 4) followed by additional electrolysis, leads to an improved yield of NH3 suggesting that some active catalyst is still present after the first run.6d,9 Notably when a CPE experiment that did not include added Cp*2Co+ was reloaded with additional acid after electrolysis and electrolyzed again the yield of NH3 did not improve above the levels obtained from a single loading of acid (2.2 equiv NH3 per Fe). Table 4.2. Yields and Faradaic Efficiencies of NH3 from CPE Experiments with P3BFe+ Equiv Equiv NH3 Equiv NH3 Entry NH3 FE (%) + 1 Cp*2Co (per Fe) (per Co) 0 2.6 ± 0.4 ― 22 ± 6 80 2 1 4.0 ± 0.6 4.0 ± 0.6 28 ± 5 3 5 4.0 ± 0.6 0.8 ± 0.1 25 ± 3 4a 5 5.5 ± 0.9 1.1 ± 0.2 19 ± 1 5 10 4±1 0.4 ± 0.1 24 ± 7 6b 5 0.9 ± 0.4 0.2 ± 0.1 6±3 7c 5 1.9 ± 0.2 0.4 ± 0.1 10 ± 1 All CPE experiments conducted at −2.1 V vs Fc+/0 with 0.1 M NaBArF4 in Et2O as solvent, cooled to −35 °C under an N2 atmosphere, featuring a glassy carbon plate working electrode, Ag/Ag+ reference couple isolated by a CoralPorTM frit referenced externally to Fc+/0, and a solid sodium auxiliary electrode. Working and auxiliary chambers separated by a sintered glass frit. See SI for further experimental details, controls, and additional data. a After initial electrolysis with 50 equiv [Ph2NH2][OTf], an additional 50 equiv [Ph2NH2][OTf] in 0.1 M NaBArF4 Et2O solution was added to the working chamber, via syringe through a rubber septum, followed by additional CPE at −2.1 V vs Fc+/0. b [PhNH3][OTf] employed as the acid. c[2,6-ClPhNH3][OTf] employed as the acid. CPE of P3BFe+ in the presence of Cp*2Co+ was also explored with other acids, replacing [Ph2NH2][OTf] in these experiments with [2,6-ClPhNH3][OTf] led to lower yields of NH3 and [PhNH3][OTf] led to even lower yields of NH3 (Table 4.1, entries 7 and 6 81 respectively). The lower but nonzero yield of NH3 provided by [PhNH3][OTf] in these CPE experiments is consistent with chemical trials employing various acids (vide supra) and can be rationalized similarly by the relative pKa of the acids (Table 4.1). The intermediate yield of NH3 provided by [2,6-ClPhNH3][OTf] in these CPE experiments is less consistent with a simple pKa consideration, suggesting that additional factors contribute to acid compatibility with these CPE conditions perhaps including the relative stability of the acid or conjugate base to electrolysis. To probe whether electrode-immobilized iron might contribute to the N2RR electrocatalysis, X-ray photo-electron spectroscopy (XPS) was used to study the electrode. After a standard CPE experiment with P3BFe+, 5 equiv of Cp*2Co,+ and 50 equiv [Ph2NH2][OTf], the electrode was removed, washed with fresh 0.1 M NaBArF4 Et2O solution, then fresh Et2O, and the electrode surface was then probed by XPS. A very low coverage of Fe (<0.3 atom % Fe) was detected in the post-electrolysis material; no Fe was detected on a segment of the electrode which was not exposed to the electrolytic solution. This observation implies a modest degree of degradation of P3BFe+ over the course of a 15 hour CPE experiment. Worth noting is that no Co was detected on the post-electrolysis electrode. To test whether the small amount of deposited Fe material might be catalytically active for N2RR, following a standard CPE experiment the electrode was removed from the cold electrolysis solution, washed with fresh 0.1 M NaBArF4 Et2O at −35 °C (the electrode itself was maintained at −35 °C at all times), and then used for an additional CPE experiment, under identical conditions except that P3BFe+ was excluded. This CPE experiment yielded no detectable NH3. The charge passed, and H2 yield, were very similar 82 to a “no P3BFe+” control experiment conducted with a freshly cleaned electrode (See SI for further details). Accordingly, a CPE experiment in the absence of P3BFe+ demonstrated that Cp*2Co+ serves as an effective electrocatalyst for HER with [Ph2NH2][OTf] as the acid source, but does not catalyze the N2RR reaction (0% FE for NH3, 76% FE for H2; see SI). This background HER and the observed catalytic response to the addition of [Ph2NH2][OTf] at the Cp*2Co+/0 couple provides circumstantial evidence for the formation of a protonated decamethylcobaltocene intermediate, Cp*(η4-C5Me5H)Co+ on a timescale similar to that of the N2RR mediated by P3BFe+. To probe the possibility that the sodium auxiliary electrode used in the CPE experiments might play a non-innocent role as a chemical reductant, a standard CPE experiment with P3BFe+, 5 equiv Cp*2Co+, and 50 equiv [Ph2NH2][OTf] was assembled, but was left to stir at −35 °C for 43 hours without an applied potential bias. This experiment yielded 0.3 equiv NH3 (relative to Fe), suggesting that background N2RR due to the sodium auxiliary electrode is very minor. To ensure the NH3 produced was derived from the N2 atmosphere during these electrolysis experiments, as opposed to degradation of the acid used, a standard CPE experiment using P3BFe+, 5 equiv Cp*2Co+, and 50 equiv of [Ph215NH2][OTf] was performed. Only 14NH3 product was detected. We also sought to compare the chemical N2RR catalysis efficiency of the P3BFe+ catalyst under conditions similar to those used for electrocatalysis. Hence, chemical catalysis with P3BFe+, employing Cp*2Co as a reductant and [Ph2NH2][OTf] as the acid, at −35 °C in a 0.1 M NaBArF4 Et2O solution, afforded lower yields of NH3 (1.8 ± 0.7 equiv of NH3 per Fe) than the yields observed via electrolysis with Cp*2Co+ as an additive. The lower yields of NH3 in these chemical trials, compared with our previously reported 83 conditions (12.8 ± 0.5 equiv of NH3 per Fe),Error! Bookmark not defined. may be attributable to i ncreased competitive HER resulting from a more solubilizing medium (0.1 M NaBAr F4 Et2O vs pure Et2O) and a higher temperature (−35 °C vs −78 °C).Error! Bookmark not defined. T hese results illustrate that an electrochemical approach to NH3 formation can improve performance, based on selectivity for N2RR, of a molecular catalyst under comparable conditions. 4.3. Conclusion Herein we report the first pKa studies on a synthetic nitrogenase and find a strong correlation between pKa and N2RR efficiency. Chemical studies revealed that, on their own, the anilinium triflate acids employed in catalysis are unable to form the N−H bonds in early-stage N2RR intermediates. We propose that the insolubility of these acids prevents the sufficiently rapid proton transfer necessary to capture the critical, unstable P3BFe=NNH intermediate. Under catalytic conditions, we believe that the presence of the metallocene reductant (Cp*2Co) is critical, as that species can be protonated to form Cp*(η4C5Me5H)Co+ which in turn plays a key role in N−H bond formation. This leads to the intriguing conclusion that an intermediate on the background HER pathway is actually a critical species in productive N2RR chemistry. We thus investigated the protonation of Cp*2Co by anilinium triflate acids using DFT. This study unveiled that the pKa effect on the N2RR efficiency could be explained by the variation in the kinetics and thermodynamics of Cp*2Co protonation by the different acids. Detailed investigation of the reactivity of Cp*(η4-C5Me5H)Co+ with the critical P3BFe=NNH intermediate revealed that indeed PCET reactivity, either synchronous or 84 asynchronous, was favorable and proceeds with only a small barrier suggesting that this intermediate could indeed be rapidly trapped by this reagent. Although we have highlighted this particular reaction, we believe Cp*(η4-C5Me5H)Co+ may be involved in a variety of N−H bond forming reactions during catalysis. Indeed given the widespread use of metallocene reductants in chemical N2RR, we believe that this type of PCET reactivity may be an overlooked mechanism for N–H bond formation. Intrigued by the idea that the conjugate acid of Cp*2Co was playing an important role in N−H bond formation, we decided to investigate the effect of Cp*2Co+ as a catalytic additive in electrochemical N2RR experiments. Indeed despite the fact that Cp*2Co+ itself only catalyzes HER under the conditions employed for electrocatalysis, we found that its inclusion in CPE experiments containing P3BFe+ and acid led to improvements in the yield of NH3 and the FE for NH3. This system represents the first unambiguous example of electrocatalytic N2RR with a soluble, molecular coordination complex. This discovery opens the door for further studies in this area. Although the yield of NH3 is modest the FE is far superior to almost all known heterogenous electrocatalysts that operate at low temperature, suggesting that such studies could provide important design criteria in the development of electrocatalytic N2RR. 4.4. References 1 (a) Yandulov, D. V.; Schrock, R. R. Science 2003, 301, 76. (b) Arashiba, K.; Miyake, Y.; Nishibayashi, Y. Nat. Chem. 2010, 3, 120. (c) Anderson, J. S.; Rittle, J.; Peters, J. C. Nature 2013, 501, 84. (d) Imayoshi, R.; Tanaka, H.; Matsuo, Y.; Yuki, M.; Nakajima, K.; Yoshizawa, K.; Nishibayashi, Y. Angew. Chem. Int. Ed. 2015, 21, 8905. (e) Ung, G.; 85 Peters, J. C. Angew. Chem. Int. Ed. 2015, 54, 532. (f) Kuriyama, S.; Arashiba, K.; Nakajima, K.; Matsuo, Y.; Tanaka, H.; Ishii, K.; Yoshizawa, K.; Nishibayashi, Y. Nat. Commun. 2016, 7, 12181. (g) Hill, P. J.; Doyle, L. R.; Crawford, A. D.; Myers, W. K.; Ashley, A. E. J. Am. Chem. Soc. 2016, 138, 13521. (h) Buscagan, T. M.; Oyala, P. H.; Peters, J. C. Angew. Chem. Int. Ed. 2017, 56, 6921. (i) Wickramasinghe, L. A.; Ogawa, T.; Schrock, R. R.; Müller, P. J. Am. Chem. Soc. 2017, 139, 9132. (j) Fajardo Jr., J.; Peters, J. C. J. Am. Chem. Soc. 2017, 139, 16105. 2 (a) Shipman, M. A.; Symes, M. D. Catal. Today 2017, 286, 57. (b) Kyriakou, V.; Garagounis, I.; Vasileiou, E.; Vourros, A.; Stoukides, M. Catal. Today 2017, 286, 2. 3 (a) McKone, J. R.; Marinescu, S. C.; Brunschwig, B. S.; Winkler, J. R.; Gray, H. B. Chem. Sci. 2014, 5, 865. (b) Bullock, R. M.; Appel, A. M.; Helm, M. L. Chem. Commun. 2014, 50, 3125. 4 (a) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M. Chem. Soc. Rev. 2009, 38, 89. (b) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois, D. L.; Dupuis, M.; Ferry, J. G.; Fujita, E.; Hille, R.; Kenis, P. J. A.; Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F.; Portis, A. R.; Ragsdale, S. W.; Rauchfuss, T. B.; Reek, J. N. H.; Seefeldt, L. C.; Thauer, R. K.; Waldrop, G. L. Chem. Rev. 2013, 113, 6621. (c) Francke, R.; Schille, B.; Roemelt, M. Chem. Rev. 2018, DOI: 10.1021/acs.chemrev.7b00459. 5 Zhang, W.; Lai, W.; Cao, R. Chem. Rev. 2017, 117, 3717. 6 (a) Pickett, C. J.; Talarmin, J. Nature 1985, 317, 652. (b) Al-Salih, T. I.; Pickett, C. J. J. Chem. Soc. Dalton Trans. 1985, 1255. (c) Pickett, C. J.; Ryder, K. S.; Talarmin, J. J. Chem. Soc. Dalton Trans. 1986, 1453. (d) Del Castillo, T. J.; Thompson, N. B.; Peters, J. C. J. Am. Chem. Soc. 2016, 138, 5341. 86 7 In this context a recent report in which the bioelectrosynthesis of ammonia by nitrogenase is coupled to H2 oxidation is also noteworthy: Milton, R. D.; Cai, R.; Abdellaoui, S.; Leech, D.; De Lacey, A. L.; Pita, M.; Minteer, S. D. Angew. Chem. Int. Ed. 2017, 56, 2680. 8 Very recently there was a report of electrolytic NH3 synthesis by Cp2TiCl2 but although rates and Faradaic efficiencies are discussed no yields of NH3 are reported: Jeong, E.-Y.; Yoo, C.-Y.; Jung, C. H.; Park, J. H.; Park, Y. C.; Kim, J.-N.; Oh, S.-G.; Woo, Y.; Yoon, H. C. ACS Sustainable Chem. Eng. 2017, 5, 9662. 9 Chalkley, M. J.; Del Castillo, T. J.; Matson, B. D.; Roddy, J. P.; Peters, J. C. ACS Cent. Sci. 2017, 3, 217. 10 In some cases the pKa of a particular anilinium acid was already known in THF in which case this value was used. In cases where the pKa was not reported in THF a literature procedure was used to appropriately convert the pKa so that it would be comparable. See SI for details. 11 Consistent with this observation is that efforts to use other weak, non-anilinium acids such as benzylammonium triflate and collidinium triflate also led to no observed NH3 formation. 12 These results are also consistent with our previous observation of [Ph2NH2][OTf] (pKa in THF of 3.2) yielding 72% ± 3 NH3. See reference 9. 13 Pham, D. N.; Burgess, B. K. Biochemistry 1993, 32, 13275. 14 In some other reports on N2RR by molecular catalysts, efficiencies for NH3 have been reported for several acids but typically these acids span only a small pKa range, electron yields are inconsistent, and variations are not explained. 87 15 Anderson, J. S.; Cutsail III, G. E.; Rittle, J.; Connor, B. A.; Gunderson, W. A.; Zhang, L.; Hoffman, B. M.; Peters, J. C. J. Am. Chem. Soc. 2015, 137, 7803. 16 Matson, B. D.; Peters, J. C. ACS Catal. 2018, 8, 1448. 17 Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104. 18 (a) Tao, J.; Perdew, J. P.; Staroverov, V. N.; Scuseria, G. E., Phys. Rev. Lett. 2003, 91, 146401. (b) Weigend, F.; Ahlrichs, R., Phys. Chem. Chem. Phys. 2005, 7, 3297. 19 We have found that this functional and basis set are able to reproduce accurately not only crystallographic details but also known singlet-triplet gaps, reduction potentials, and N−H BDFE’s. For further detail see Reference 16. 20 In this section we present DFT data that includes explicit OTf− interactions. It should be noted that while the absolute thermodynamics and kinetics are dependent on the OTf−, calculations in its absence suggest that the trends discussed here are more broadly applicable with respect to more non-coordinating anions. 21 The reactivity of ring-functionalized Cp rings has been discussed previously in the context of hydride transfer in HER with 4d and 5d metals: (a) Pitman, C. L.; Finster, O. N. L.; Miller, A. J. M. Chem. Commun. 2016, 52, 9105. (b) Quintana, L. M. A.; Johnson, S. I.; Corona, S. L.; Villatoro, W.; Goddard, W. A.; Takase, M. K.; VanderVelde, D. G.; Winkler, J. R.; Gray, H. B.; Blakemore, J. D. Proc. Natl. Acad. Sci. 2016, 113, 6409. (c) Peng, Y.; Ramos-Garcés, M. V.; Lionetti, D.; Blakemore, J. D. Inorg. Chem. 2017, 56, 10824. 22 Marcus, R. A. J. Chem. Phys. 1956, 24, 966–978. 23 Iordanova, N.; Decornez, H.; Hammes-Schiffer, S. J. Am. Chem. Soc. 2001, 123, 3723– 3733. 88 24 We have assumed a PT/ET mechanism in which ET is rate limiting based on significantly lowered reorganization energies and barriers for PT compared to ET. See SI for full description. 25 Koelle, U.; Infelta, P. P.; Graetzel, M., Inorg. Chem. 1988, 27, 879. 26 Very recently there has been a report of electrocatalytic N2RR under ambient conditions in ionic liquids with Fe nanoparticles that achieve FE’s as high as 60%: Li, S.-J.; Bao, D.; Shi, M.-M.; Wulan, B.-R.; Yan, J.-M.; Jiang, Q., Adv. Mater. 2017, 29, 1700001. 27 A sodium auxiliary electrode was employed because oxidation of Et2O or NaBArF4 at an inert electrode lead to prohibitively high compliance voltage requirements. Likewise due to the extensive diffusion between the working and auxiliary chambers the application of a counter redox process which produced an oxidation product which could diffuse to the working electrode and be re-reduced at −2.1 V vs Fc+/0 lead to excessive nonproductive redox cycling between chambers over the course of the lengthy CPE experiments. Using sodium metal as an electrode material provided a suitable solution to these technical challenges as the product of its oxidation (Na+) is stable to the CPE conditions. Concern regarding whether the sodium electrode could serve directly as a reducant in a chemical process to form NH3 under these conditions was addressed in a control experiment described in the text. 89 Chapter 5. Predicting BDFE Values Using DFT 90 5.1. Introduction As demonstrated in the preceding chapters of this thesis, the ability to efficiently and accurately predict the bond dissociation free-energies (BDFE) of highly reactive intermediates can be a powerful tool in investigating the mechanism of multi-proton, multielectron reduction of N2 (the N2 reduction reaction, N2RR).1 Accordingly, there are a variety of studies that have attempted to quantify the strength of catalytically relevant E– H bonds (BDFEE-H) in N2RR intermediates.1,2 These BDFEE-H values can provide crucial information regarding the stability and reactivity of catalytic intermediates, and can provide fundamental insight to guide experimental studies, particularly those which involve proton-coupled electron transfer (PCET). Studies aimed at predicting BDFEN-H bonds strengths in N2RR catalysis have established density functional theory (DFT) as a useful method for the prediction of relative BDFEs. DFT is most powerful, however, when it can be calibrated to experimental values and, as such, the accurate prediction of relative and absolute BDFEE-H values across a wide variety of E–H bonds, both metal bound and free, is highly desirable. In this chapter, a DFT method for the calibration of literature BDFEE-H values is presented which leads to the accurate (within 4 kcal/mol) prediction of both free and metal bound E–H bond strengths in the gas-phase. While the prediction of solvated BDFEE-H values is shown to have only slightly higher overall errors (ca. 5 kcal/mol) for organic E-H bonds, accurate calibration of solution phase metal bound E-H bonds remains less understood due to the relative lack of literature data. Nonetheless, similar empirical corrections appear to show similar 91 promise with a smaller library of literature data in MeCN and THF. In sum, the method outlined in the following presents an efficient method for gas-phase BDFEE-H predictions and suggest simple empirical corrections that can allow for solution phase values within ca. 5 kcal/mol. 5.2. Calibration of Gas-Phase BDFEE-H Values Gas-phase organic molecules provide an ideal starting point from which to begin calibrating a DFT method. This is due to the large library of literature values and the lack of any solvation induced weakening or strengthening of the E-H bond.3 To investigate the efficacy of BDFEE-H prediction for gas-phase values, four common functionals, B3LYP,4 BP86,5 M06-L6 and TPSS,7 were used to calculate known gas-phase BDFEE-H values using a common and efficient basis set (def2-SVP).8 92 Figure 5.1. Plots of BDFEcalc vs BDFElit for known gas-phase literature values using 4 common functionals. As shown in Figure 5.1, all four functionals produced calibration curves with similar parameters (slope from 0.95 to 1.0; intercept from 4 to 6.5; Figure 5.1). Using each calibration equation, errors were calculated for each functional (Table 5.1) and were shown to vary between 2.3 (B3LYP) and 4.6 (M06-L). It is further noted that the hybrid functional, B3LYP, was found to have significantly lowered errors compared to the pure functionals. The ability of any of the these functionals to be effectively reproduce gas-phase BDFEE-H values, however, is a useful observation. In particular, our research has shown large functional dependences on properties of P3EFe species, with TPSS and BP86 providing 93 superior predictions for spin-state energy gaps, reduction potentials and BDFEN-H values than B3LYP and M06-L. Notably, B3LYP and M06-L have been observed to favor highspin states, a factor that is crucial in metal bound BDFEE-H predictions but not relevant in prediction of free, organic E-H bond strengths. In addition to evaluating the overall errors for gas-phase BDFEE-H values, the data in Figure 5.1 and Table 5.1 allow us to confirm that these errors are normally distributed. Normal probability plots of the errors from each functional are shown in Figure 5.2. Notably, all four functionals tested shown a linear relationship and r-values that are greater than the critical value of 0.96 at a 5% confidence level. Accordingly, it is reasonable to assume that upon application of the lines of best fit shown in Figure 5.1, any errors are normally distributed. Table 5.1. Errors Obtained from Figure 5.1. O-H Bonds B3LYP BP86 M06-L TPSS HOO–H −4.1 −4.7 −4.1 −5.6 MeO–H −3.0 −3.4 −3.4 −4.2 EtOO–H −2.0 −2.8 −2.5 −3.6 HO–H −2.7 −0.8 −1.8 −3.3 PhOH −0.8 −0.8 −1.1 −1.6 OO–H −0.3 −0.1 −2.8 −0.1 HC(O)OO–H −0.3 0.4 0.7 −0.6 94 O–H− −1.6 1.3 −1.8 −0.9 C–H Bonds B3LYP BP86 M06−L TPSS Me3C–H −1.5 −2.7 −2.1 −1.6 Me2CH2 −0.8 −1.9 −0.5 −0.6 C6H6 1.2 −0.7 −1.5 0.5 C2H4 0.8 −0.6 −0.7 0.8 C2H6 0.7 −0.1 1.4 1.0 PhCH3 1.1 0.7 1.5 1.5 CH4 1.8 1.5 2.7 2.0 CpH 1.4 2.0 1.2 2.1 N–H Bonds B3LYP BP86 M06−L TPSS Et2NH −0.6 −0.7 −1.3 −1.3 NH2NH2 −0.1 −0.6 −0.3 −1.0 NH3 −0.8 −0.1 −0.2 −1.6 NH4+ 0.5 1.0 1.0 −0.6 PhNH2 3.3 3.4 3.6 2.4 NHNH 2.8 3.7 3.5 3.1 Other E–H Bonds B3LYP BP86 M06−L TPSS PhSH −1.8 −1.2 −0.2 −0.7 MeSH 0.6 1.0 3.3 2.2 95 EtSH 0.7 1.2 2.7 2.2 H2S 1.9 3.2 4.8 4.1 H2 3.5 1.8 −2.2 5.3 MSE 0.0 0.0 0.0 0.0 MUE 2.3 3.7 4.6 4.0 Figure 5.2. Normal probability plots for the errors in Table 5.1. 96 5.3. Solution Phase BDFEE-H Values Given the similarity between functional calibration and the accuracy with which TPSS has been shown to treat P3EFe complexes, we turned our attention to its use in solution phase BDFEE-H predictions and metal bound BDFEE-H predictions. Within the context of free, organic BDFEE-H species, a simple empirical formula (eq. 5.1) was shown to produce only slightly higher errors when compared to the gas-phase values. As shown in Table 5.2, the combined errors for BDFEsoln range from 0.2 to 1.1 kcal/mol greater than the baseline error in BDFEgas (4.0 kcal/mol; Table 5.1). BDFEsoln = BDFEgas + Csolv (eq. 5.1) Table 5.2. Solvent correction (Csolv) and total rrors for several E-H bond/solvent combinationsa using TPSS/def2-SVP CDMSO CMeCN CC6H6 {Total Error}b {Total Error} {Total Error} +4.9 kcal/mol +3.9 kcal/mol +8.3 kcal/mol {4.4 kcal/mol) {4.2 kcal/mol} {4.6 kcal/mol} +11.0 kcal/mol +10.1 kcal/mol C-H, H-H bonds O-H, N-H, S-H bonds N. D. {5.1 kcal.mol} a {4.8 kcal/mol} Using TPSS/def2-SVP level of theory. b Error = (√[(BDFEgas)2 + (Csolv)2]) 97 5.4. Transition Metal Bound E–H Bonds Transition metal bound E–H bonds provide additional challenges, in large part due to a smaller library of literature values and the lack of any well-defined values in the gas-phase. Application of the gas-phase calibration (TPSS; Figure 5.1) and eq. 5.1 provides solvent correction terms for transition metal bound BDFEE-H values in THF and MeCN (Table 5.3). It is notable that the correction terms show a significantly decreased magnitude and opposite sign. The decrease in the magnitude of solvent corrections for transition metal species is not surprising, as the large size and comparatively constant ligand entropy in the E–H and E● species leads to decreased relative role of entropic factors. While not investigated in this research, the BDFEE-H values of several transition metal species are known in DMSO and H2O, leading to the possible expansion of the approached described to transition metal systems in polar and/or protic solvents. Table 5.3. Solvent correction (Csolv) and total errors for several transition metal based E-H bond/solvent combinations a Csolv Total Error MeCN –3.2 kcal/mol 4.1 kcal/mol THF –2.5 kcal/mol 4.4 kcal/mol Using TPSS/def2-TZVP(TM); def2-SVP. b Error = (√[(BDFEgas)2 + (Csolv)2]) 98 5.5. Conclusions In conclusion, a simple and efficient method for predicting free and transition metal bound BDFEE-H values is presented. The method describe relies first on producing a calibration curve using literature data for BDFEE-H species in the gas-phase. It has been shown with four common functionals that these calibrations can be accomplished with errors of ca. 4 kcal/mol. Further, it has been shown that an empirical correction for the type of bond and solvent can be added to these gas-phase values with only marginally increased total errors (0.2 – 1.1 kcal/mol; total errors: 4 – 5 kcal/mol). 5.6. References 1 (a) Matson, B. D.; Peters, J. C. ACS Catalysis, 2018, 8, 1448-1455. (b) Chalkley, M. J.; Del Castillo, T. J.; Matson, B. D.; Roddy, J. P.; Peters, J. C. ACS Central Science 2017, 3, 217-223. 2 (a) Rittle, J.; Peters, J. C. J. Am. Chem. Soc. 2017, 139, 3161–3170. (b) Pappas, I.; Chirik, P. J. J. Am. Chem. Soc. 2016, 138, 13379. 3 (a) Warren, J. J.; Tronic, T. A.; Mayer, J. M. Chem. Rev. 2010, 110, 6961–7001. (b) Yandulov, D. V.; Schrock, R. R. Inorg. Chem. 2005, 44, 1103. 4 (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652. (b) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B. 1988, 37, 785-789. (c) Vosko, S. H.; Wilk, L.; Nusair, M.; Can. J. Phys. 1980, 58, 1200-1211. (d) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98, 11623-11627. 99 5 (a) Becke, A. D. Phys. Rev. A. 1988, 38, 3098-30100. (b) Perdew, J. P. Phys. Rev. B. 1986, 33 8822-8824. 6 Zhao, Y.; Schultz, N. E.; Truhlar, D. G. J. Chem. Theory and Comput. 2006, 2, 364-382. 7 Tao, J. M.; Perdew, J. P.; Staroverov, V. N.; Scuseria, G. E. Phys. Rev. Lett. 2003, 91, 146401. 8 Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297–3305. 100 Appendix 1. Supplementary Data for Chapter 2 101 A1.1. General Computational Details All stationary point geometries were calculated using dispersion corrected DFTD31 with a TPSS functional,2 a def2-TZVP basis set on transition metals and a def2-SVP basis set on all other atoms.3 Calculations were performed, in part, using Xtreme Science and Engineering Discovery Environment (XSEDE) resources.4 Calculations were performed on the full P3EFe scaffolds. Geometries were optimized using the NWChem 6.5 package or Orca 3.0.3 package.5 All single point energy, frequency and solvation energy calculations were performed with the Orca 3.0.3 package. Frequency calculations were used to confirm true minima and to determine gas phase free energy values (Ggas). Single point solvation calculations were done using an SMD solvation model6 with diethyl ether solvent and were used to determine solvated internal energy (Esoln). Free energies of solvation were approximated using the difference in gas phase internal energy (Egas) and solvated internal energy (∆Gsolv ≈ Esoln – Egas) and the free energy of a species in solution was then calculated using the gas phase free energy (Ggas) and the free energy of solvation (Gsoln = Ggas + ∆Gsolv). All reduction potentials were calculated referenced to Fc+/0 using the standard Nernst relation G = -nFE0. 102 A1.2. Fe–H Formation Figure A1.1. Structure of P3BFeN2− + (Et2O)2H+ immediately before (top) and after (bottom) dissociation of a Et2O moiety. The relaxed surface scan reveals little change in the P3BFeN2− unit before Et2O dissociation, indicative of the presence of a (Et2O)2H+  (Et2O)H+ + Et2O pre-equilibrium. A1.3. BDFE Calculations Bond dissociation free energies (BDFE) of X–H bonds were calculated in the gasphase using a series of known reference compounds.7 The free-energy difference between the H-atom donor/acceptor pair was calculated based on the thermochemical information provided by frequency calculations after structure optimizations using the procedure described in the general computational section. A linear plot of ΔG vs BDFElit was 103 generated to form a calibration curve (Figure A1.1). BDFE predictions were generated by application of the line of best fit to the calculated ΔG of the unknown species. Errors were calculated by application of the trend line to the calculated free-energies of known species and comparison to their literature BDFE value. Errors are reported as the average of BDFEcalc-BDFElit (mean signed error, MSE = 0.0) and the average of the absolute values of BDFEcalc-BDFElit (mean unsigned error, MUE = 1.3). Figure A1.2. Plot of calculated BDFE vs literature BDFE. Line of Best fit shown with equation along with r2 value. Table A1.1. Summary of BDFEs used for calibration. G (E-H) G (E*) Gcalc BDFElit Error 104 PhNH2 NH2NH2 PhSH PhH C6H6 PhOH NH3 NHNH Me2NH NH4 (+) OOH -287.4 -111.8 -630.2 -271.3 -232.1 -307.2 -56.5 -110.6 -213.6 -56.8 -150.8 -286.7 -111.1 -629.5 -270.7 -231.4 -306.6 -55.8 -110.0 -212.9 -56.1 -150.2 79.8 67.3 70.3 79.0 101.6 74.0 94.0 51.0 81.0 113.0 37.5 81.5 72.6 75.3 81.6 104.7 79.8 99.4 52.6 86.4 116.9 42.7 -2.4 1.2 0.9 -1.5 -0.9 1.7 1.3 -2.6 1.3 0.0 1.0 MUE 1.4 MSE 0.1 A1.4. Approximation of P3EFe(NNHy) Radius The radius of P3EFe(NNHy) was approximated by using the average molar volume of several relevant crystal structures to determine a radius assuming a spherical molecule. Table A1.2. Volume and Calculated Radius of Previous Characterized P3EFe Species from XRD Data Volume (Å3) rcalc (Å) Ref P3SiFe(N2) 881.2 5.9 8 P3SiFe(CN) 1101.9 6.2 9 105 P3SiFe(CNMe) 915.7 6.0 10 P3CFe(N2) 869.3 5.9 11 P3CFe(H)(N2) 869.8 5.9 9 P3BFe(NH2) 866.1 5.9 12 Average 6.0 Å Std Dev 0.1 Å A1.5. Calculated Reorganization Energies The inner-sphere reorganization energy for electron transfer (λis,ET) was estimated assuming non-adiabatic behavior and by calculating the difference between the single point energies of the relevant species in its ground state and the corresponding single point energy of this ground state in the oxidized or reduced geometry. λis,ET = [E(Feoxox) – E(Feoxred)] + [E(Feredred) – E(Feredox)] (Eq. A1.1) Relative reduction barriers were approximated by first defining the barrier for P3BFe(NNH2)+ to be 1.0 kcal/mol. Subsequent back-calculation of λtot yielded solutions of 30.5 kcal/mol and 56.5 kcal/mol, corresponding to the solutions in the inverted and normal regimes, respectively. The reorganization energy leading to the inverted solution would imply very small energies for KC8 and solvent reorganization (λKC8 + λOS = 7.5 kcal/mol). This led us to assume that the reduction steps were in the normal region. To check this assumption, outer-sphere reorganization energy was approximated using a continuum 106 model.13 For electron transfer (λos,ET) reactions the KC8 reductant was modeled as an electrode surface (rKC8 >> rcat). The radius of the P3EFe molecules (rcat; Eq. 2) was approximated using the volumes of several relevant crystal structures. The values for the static and optical dielectric constant (εs and εop) of diethyl ether were taken as the values used in the SMD solvation model. This value was approximated at 33 kcal/mol, consistent with the reductions of interest occurring in the normal region. Accordingly, the total reorganization for P3BFe(NNH2)+ reduction (G* ≡ 1.0 kcal/mol) was assumed to be 56.5 kcal/mol. Perturbation of this value by the differences between λ ISSi/C and λISB lead to the relative barriers shown in Table S3. Table A1.3. Summary of Calculated Reorganization Energiesa Redox Couple λIS,ET λOS + λKC8 G*rel P3BFe(NNH2)+/0 23.0 33.5 1.0b P3SiFe(NNH2)+/0 29.7 33.5 4.4 P3CFe(NNH2)+/0 33.5 5.2 29.7 107 a All energies are in kcal/mol b G* values expressed relative to that of P3BFe(NNH2)+ reduction, defined as 1.0 kcal/mol A1.6. Determination of the Work Function The work required to bring two cationic iron species together was approximated following the methods of Hammes-Schiffer and Mayer (Eq 1).14 𝑤𝑟 = 𝑒 2 𝑍1 𝑍2 𝑓 𝜖𝑜 𝑟 (eq. A1.2) Here Z1 and Z2 are the charges on each complex (Z1 = Z2 = +1) and e is the elementary charge. The distance between iron centers was taken as twice the radius of the P3EFe species (r = 12 Å) and ϵo is the static dielectric constant. The debye screening factor (f) was calculated using eq. 2. 8𝜋𝑁 𝑒 2 𝜇 𝑓 −1 = 1 + 𝑟√1027 𝜖𝐴 𝑘 𝑇 𝑜 𝐵 (eq. A1.3) Where μ is the ionic strength (taken as [Fe] = 1.3 mM) and NA are kB Avogadro’s number and the Boltzmann constant, respectively. The temperature was taken as the standard temperature for catalysis (T =195 K). Substitution of the appropriate values into Eq. 1 and 2 yields wr = 5.2 kcal/mol. 108 A1.7. Summary of Wiberg Indices Table A1.4. Summary of Wiberg Bond Indices for P3EFe(N2) complexes P3BFe Alpha Beta Total P3SiFe Alpha Beta Total P3CFe Alpha Beta Total Fe-N1 0.2 0.2 0.9 Fe-N1 0.2 0.3 1.0 Fe-N1 0.2 0.3 1.0 Fe-N2 0.1 0.1 0.4 Fe-N2 0.1 0.1 0.4 Fe-N2 0.1 0.1 0.4 N-N 0.7 0.6 2.6 N-N 0.6 0.6 2.6 N-N 0.6 0.6 2.5 Fe-B 0.1 0.1 0.4 Fe-Si 0.2 0.2 0.7 Fe-C 0.2 0.2 0.7 Fe-P1 0.2 0.2 0.7 Fe-P1 0.2 0.2 0.8 Fe-P1 0.2 0.2 0.8 Fe-P2 0.2 0.2 0.7 Fe-P2 0.2 0.2 0.8 Fe-P2 0.2 0.2 0.8 Fe-P3 0.2 0.2 0.7 Fe-P3 0.2 0.2 0.7 Fe-P3 0.2 0.2 0.7 Table A1.5. Summary of Wiberg Bond Indices for P3EFe(NNH) complexes P3BFe Alpha Beta Total P3SiFe Alpha Beta Total P3CFe Alpha Beta Total Fe-N1 0.4 0.4 1.6 Fe-N1 0.4 0.4 1.6 Fe-N1 0.3 0.3 1.2 Fe-N2 0.1 0.1 0.4 Fe-N2 0.1 0.1 0.4 Fe-N2 0.0 0.0 0.2 N-N 0.5 0.4 1.8 N-N 0.4 0.4 1.8 N-N 0.4 0.4 1.5 N-H 0.2 0.2 0.8 N-H 0.2 0.2 0.8 N-H 0.2 0.2 0.8 109 Fe-B 0.1 0.1 0.5 Fe-Si 0.2 0.2 0.7 Fe-C 0.1 0.1 0.5 Fe-P1 0.2 0.2 0.7 Fe-P1 0.2 0.2 0.8 Fe-P1 0.2 0.2 0.7 Fe-P2 0.2 0.2 0.8 Fe-P2 0.2 0.2 0.8 Fe-P2 0.2 0.2 0.8 Fe-P3 0.2 0.2 0.8 Fe-P3 0.2 0.2 0.8 Fe-P3 0.2 0.2 0.7 Table A1.6. Summary of Bond Indices for P3EFe(NNH2) complexes P3BFe Alpha Beta Total P3SiFe Alpha Beta Total P3CFe Alpha Beta Total Fe-N1 0.5 0.5 1.9 Fe-N1 0.2 0.4 1.2 Fe-N1 0.3 Fe-N2 0.0 0.0 0.2 Fe-N2 0.0 0.0 0.2 Fe-N2 0.0 N-N 0.3 0.3 1.2 N-N 0.4 0.3 1.4 N-N 0.4 0.3 1.4 N-H 0.2 0.2 0.8 N-H 0.2 0.2 0.8 N-H 0.2 0.2 0.8 N-H 0.2 0.2 0.8 N-H 0.2 0.2 0.8 N-H 0.2 0.2 0.8 0.4 1.4 0.1 110 Fe-B 0.1 0.1 0.4 Fe-Si 0.2 0.2 0.7 Fe-C 0.1 0.1 0.5 Fe-P1 0.2 0.2 0.8 Fe-P1 0.2 0.2 0.8 Fe-P1 0.2 0.2 0.8 Fe-P2 0.2 0.2 0.8 Fe-P2 0.2 0.2 0.8 Fe-P2 0.2 0.2 0.8 Fe-P3 0.2 0.2 0.8 Fe-P3 0.2 0.2 0.8 Fe-P3 0.2 0.2 0.8 Table A1.7. Summary of Wiberg Bond Indices for P3EFe(N(4-OMe-Ph)) P3BFe Total Fe-N1 1.8 N-C 1.2 Fe-B 0.4 Fe-P1 0.8 Fe-P2 0.8 Fe-P3 0.8 111 Table A1.8. Summary of Wiberg Bond Indices for C2H4 and C2H5 C2H4 Alpha Beta Total C2H5 Alpha Beta Total C1-H1 0.94 0.94 1.9 C1-H1 0.24 0.24 0.96 C1-H2 0.94 0.94 0.2 C1-H2 0.24 0.24 0.96 C1-C2 0.94 0.94 1.2 C1-C2 0.23 0.23 0.93 C2-H3 0.94 0.94 0.8 C2-H3 0.23 0.23 0.93 C2-H4 2.05 2.05 0.8 C2-H4 0.23 0.22 0.90 C2-H5 0.27 0.28 1.10 112 A1.8. Comparison of Calculated to Known Experimental Values Table A1.9. Comparing of calculated to experimental values for several parameters of interest. Si + P3 Fe(NNMe2) P3SiFe(NNMe2)+ P3BFe(NNMe2) P3BFe(NNMe2) P3SiFe(CNH) P3SiFe(CNH) P3SiFe(NNMeH)+ Parameter Singlet-Triplet Gap Reduction Potential Singlet-Triplet Gap Reduction Potential BDFEN-H BDFEN-H BDFEN-H Calculated 6.9 kcal/mol Experimental 6.0 Ref 9 -1.81 V vs Fc+/0 -1.73 V vs Fc+/0 9 5.5 kcal/mol 4.0 kcal/mol 15 -1.29 V vs Fc+/0 -1.20 V vs Fc+/0 14 43.5 kcal/mol 61.8 kcal/mol 45.9 kcal/mol 41.4 kcal/mol 61.9 kcal/mol 44.9 kcal/mol 9 9 9 113 A1.9. References 1 Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104. 2 Tao, J. M.; Perdew, J. P.; Staroverov, V. N.; Scuseria, G. E. Phys. Rev. Lett. 2003, 91, 146401. 3 Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297–3305. 4 John Towns, Timothy Cockerill, Maytal Dahan, Ian Foster, Kelly Gaither, Andrew Grimshaw, Victor Hazlewood, Scott Lathrop, Dave Lifka, Gregory D. Peterson, Ralph Roskies, J. Ray Scott, Nancy Wilkins-Diehr, "XSEDE: Accelerating Scientific Discovery", Computing in Science & Engineering, vol.16, no. 5, pp. 62-74, Sept.-Oct. 2014, doi:10.1109/MCSE.2014.80 5 (a) Valiev, M.; Bylaska, E. J.; Govind, N.; Kowalski, K.; Straatsma, T. P.; Van Dam, H. J. J.; Wang, D.; Nieplocha, J.; Apra, E.; Windus, T. L.; de Jong, W. A. Comput. Phys. Commun. 2010, 181, 1477–1489. (b) Neese, F. Wiley interdisciplinary Reviews Computational Molecular Science, 2012, 2:1, 73-78. 6 (a) Klamt, A.; Schüürmann, G. J. Chem. Soc. Perkin Trans. 2. 1993, 2, 799– 805. (b)Marten, B.; Kim, K.; Cortis, C.; Friesner, R. A.; Murphy, R. B.; Ringnalda, M. N.; Sitkoff, D.; Honig, B. J. Phys. Chem. 1996, 100, 11775–11788. 7 Warren, J. J.; Tronic, T. A.; Mayer, J. M. Chem. Rev. 2010, 110, 6961–7001. 114 8 Lee, Y.; Mankad, N. P.; Peters, J. C. Nature Chem. 2010, 2, 558-565. 9 Rittle, J.; Peters, J. C. Angew. Chem. Int. Ed. 2016, 55, 12262 –12265. 10 Rittle, J.; Peters, J. C. J. Am. Chem. Soc. 2017, 139, 3161–3170. 11 Creutz, S. E.; Peters, J. C. J. Am. Chem. Soc. 2014, 136, 1105−1115. 12 Anderson, J. S.; Moret, M-. E.; Peters, J. C. J. Am. Chem. Soc. 2013, 135, 534−537. 13 Marcus, R. A. J. Chem. Phys. 1956, 24, 966–978. 14 (a) Iordanova, N.; Decornez, H.; Hammes-Schiffer, S. J. Am. Chem. Soc. 2001, 123, 3723-3733. (b) Roth, J. P.; Lovell, S.; Mayer, J. M. J. Am. Chem. Soc. 2000, 122, 54865498. 15 Thompson, N. B.; Green, M. T.; Peters, J. C. J. Am. Chem. Soc. 2017, 139, 15312–15315. 115 Appendix 2. Supplementary Data for Chapter 3 116 A2.1. Experimental details A2.1.1 General Considerations: All manipulations were carried out using standard Schlenk or glovebox techniques under an N2 atmosphere. Solvents were deoxygenated and dried by thoroughly sparging with N2 followed by passage through an activated alumina column in a solvent purification system by SG Water, USA LLC. Non-halogenated solvents were tested with sodium benzophenone ketyl in tetrahydrofuran (THF) in order to confirm the absence of oxygen and water. Deuterated solvents were purchased from Cambridge Isotope Laboratories, Inc., degassed, and dried over activated 3-Å molecular sieves prior to use. Cp*2Co,1 [P3BFe][BArF4],2 P3SiFeN2,3 [P3BCoN2][Na(12-crown-4)2],4 P3SiCoN2,5 [P3BFeN2][Na(12-crown-4)2],6 and [Ph215NH2][OTf]7 were prepared according to literature procedures. Ph15NH2 was obtained from Sigma-Aldrich, Inc. degassed, and dried over activated 3-Å molecular sieves prior to use. All other reagents were purchased from commercial vendors and used without further purification unless otherwise stated. Diethyl ether (Et2O) used in the experiments herein was stirred over Na/K (≥ 2 hours) and filtered or vacuum-transferred before use unless otherwise stated. A2.1.2 Physical Methods: 1 H chemical shifts are reported in ppm relative to tetramethylsilane, using 1H resonances from residual solvent as internal standards. IR measurements were obtained as 117 solutions or thin films formed by evaporation of solutions using a Bruker Alpha Platinum ATR spectrometer with OPUS software (solution IR collected in a cell with KBr windows and a 1 mm pathlength). H2 was quantified on an Agilent 7890A gas chromatograph (HPPLOT U, 30 m, 0.32 mm ID; 30 °C isothermal; nitrogen carrier gas) using a thermal conductivity detector. A2.1.3 Mössbauer Spectroscopy: Mössbauer spectra were recorded on a spectrometer from SEE Co. (Edina, MN) operating in the constant acceleration mode in a transmission geometry. The sample was kept in an SVT-400 cryostat form Janis (Wilmington, MA). The quoted isomer shifts are relative to the centroid of the spectrum of a metallic foil of α-Fe at room temperature (RT). Solution samples were transferred to a sample cup and chilled to 77 K inside of the glovebox, and unless noted otherwise, quickly removed from the glovebox and immersed in liquid N2 until mounted in the cryostat. Data analysis was performed using version 4 of the program WMOSS (www.wmoss.org) and quadrupole doublets were fit to Lorentzian lineshapes. See discussion below for detailed notes on the fitting procedure. A2.1.4 Ammonia and Hydrazine Quantification: Reaction mixtures are cooled to 77 K and allowed to freeze. The reaction vessel is then opened to atmosphere and to the frozen solution is slowly added an excess (with respect to acid) solution of a NaOtBu solution in MeOH (0.25 mM) over 1-2 minutes. This 118 solution is allowed to freeze, then the headspace of the tube is evacuated and the tube is sealed. The tube is then allowed to warm to RT and stirred at RT for at least 10 minutes. An additional Schlenk tube is charged with HCl (3 mL of a 2.0 M solution in Et2O, 6 mmol) to serve as a collection flask. The volatiles of the reaction mixture are vacuum transferred at RT into this collection flask. After completion of the vacuum transfer, the collection flask is sealed and warmed to RT. Solvent is removed in vacuo, and the remaining residue is dissolved in H2O (1 mL). An aliquot of this solution (10–100 μL) is then analyzed for the presence of NH3 (present as NH4Cl) by the indophenol method.8 A further aliquot of this solution is analyzed for the presence of N2H4 (present as N2H5Cl) by a standard colorimetric method.9 Quantification is performed with UV−vis spectroscopy by analyzing absorbance at 635 nm. In this case of runs with [PhNH3][OTf] we found that aniline in the form of anilinium chloride was present in the receiving vessels. The anilinium chloride interfered with the indophenol and hydrazine detection method. Therefore, quantification for NH3 was performed by extracting the solid residue into 1 mL of DMSO-d6 that has 20 mmol of trimethoxybenzene as an internal standard. Integration of the 1H NMR peak observed for NH4 was then integrated against the two peaks of trimethoxybenzene to quantify the ammonium present. This 1H NMR detection method was also used to differentiate [14NH4][Cl] and [15NH4][Cl] produced in the control reactions conducted with [15NPh2H2][OTf]. 119 A2.1.5 EPR Spectroscopy X-band EPR spectra were obtained on a Bruker EMX spectrometer. Samples were collected at powers ranging from 6-7 mW with modulation amplitudes of 2.00 G, modulation frequencies of 100.00 kHz, over a range of 2450 to 2900 Gauss. Spectra were baseline corrected using the algorithm in SpinCount.10 EPR spectra were modeled using the easyspin program.11 A2.1.6 Computational Methods All stationary point geometries were calculated using DFT with an M06-L functional,12 a def2-TZVP13 basis set on transition metals (Stuttgart ECP14 was used on Mo atoms) and a def2-SVP13 basis set on all other atoms. Calculations were performed, in part, using Xtreme Science and Engineering Discovery Environment (XSEDE) resources.15 Calculations were performed on the full P3EFe scaffolds. Calculations on the [HIPTN3N]Mo system were performed on a truncated scaffold in which the isopropyl groups were removed (i.e. [{3,5-(C6H4)2C6H3NCH2CH2}3N]3–). Geometries were optimized using the NWChem 6.5 package.16 All single point energy, frequency and solvation energy calculations were performed with the Gaussian09 package.17 Frequency calculations were used to confirm true minima and to determine gas phase free energy values (Ggas). Single point solvation calculations were done using an SMD solvation model with diethyl ether solvent and were used to determine solvated internal energy (Esoln). Free energies of solvation were approximated using the difference in gas phase internal energy 120 (Egas) and solvated internal energy (∆Gsolv ≈ Esoln – Egas) and the free energy of a species in solution was then calculated using the gas phase free energy (Ggas) and the free energy of solvation (Gsoln = Ggas + ∆Gsolv).18 All reduction potentials were calculated referenced to Fc+/0 using the standard Nernst relation G = -nFE0. A2.2. Synthetic Details: A2.2.1 General Procedure for the Synthesis of the Acids: Prior to use the amine was purified (aniline by distillation, diphenylamine and triphenylamine by recrystallization). To a 250 mL round bottom flask in the glovebox was added the amine which was subsequently dissolved in 100 mL of Et2O (no additional drying with NaK). To this was added dropwise (1 equiv) of HOTf with stirring over five minutes. Immediate precipitation of white solid was observed and the reaction mixture was allowed to stir for one hour at RT. The reaction mixture was then filtered and the resulting white powder was washed with Et2O (50 mL), pentane (50 mL) and Et2O again (50 mL). The resulting white microcrystalline material was then dried under vacuum. Yields of greater than 90% of microcrystalline material was obtained in this manner in all cases. A2.3. Ammonia production and quantification studies A2.3.1 Standard NH3 Generation Reaction Procedure: All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, the precatalyst (2.3 μmol) was weighed into a vial.* The 121 precatalyst was then transferred quantitatively into a Schlenk tube using THF. The THF was then evaporated to provide a thin film of precatalyst at the bottom of the Schlenk tube. The tube is then charged with a stir bar and the acid and reductant are added as solids. The tube is then cooled to 77 K in a cold well. To the cold tube is added Et2O to produce a concentration of precatalyst of 2.3 mM. The temperature of the system is allowed to equilibrate for 5 minutes and then the tube is sealed with a Teflon screw-valve. This tube is passed out of the box into a liquid N2 bath and transported to a fume hood. The tube is then transferred to a dry ice/acetone bath where it thaws and is allowed to stir at -78 °C for three hours. At this point the tube is allowed to warm to RT with stirring, and stirred at RT for 5 minutes. To ensure reproducibility, all experiments were conducted in 200 mL Schlenk tubes (51 mm OD) using 25 mm stir bars, and stirring was conducted at ~900 rpm. * In cases where less than 2.3 μmol of precatalyst was used stock solutions were used to avoid having to weigh very small amounts. Table A2.1: UV-vis quantification results for standard NH3 generation experiments with P3BFe+ Entry P3BFe+ Acid (μmol) equiv A 2.3 108a B 2.3 108a C 2.3 108a Avg. D 2.3 322a E 2.3 322a Cp*2Co equiv 54 54 54 162 162 NH4Cl N2H5Cl (μmol) (μmol) 31.4 0.0 28.5 0.0 29.2 0.0 76.4 80.0 2.0 0.7 Equiv NH3/Fe 13.5 12.3 12.6 12.8 ± 0.5 33.0 34.5 % Yield NH3 Based on e75.6 68.6 70.4 72 ± 3 61.4 64.2 122 Avg. 34 ± 1 63 ± 2 a F 2.3 638 322 60.4 0.5 26.0 24.3 G 2.3 638a 322 63.2 0.3 27.3 25.4 Avg. 26.7 ± 0.9 25 ± 1 b H 1.1 108 54 7.8 0.0 6.9 37.6 I 2.3 108b 54 19.2 0.0 8.3 46.3 Avg. 8±1 42 ± 6 c J 2.3 108 54 17.7 N.D. 7.7 43.1 K 2.3 108c 54 13.8 N.D. 6.0 33.6 Avg. 7±1 38 ± 7 c L 2.3 322 162 39.8 N.D. 17.3 32.0 M 2.3 322c 162 31.9 N.D. 13.9 25.7 Avg. 16 ± 3 29 ± 4 N.D. indicates the value was not determined aAcid used is [Ph2NH2][OTf] bAcid used is [Ph2NH2][BArF4] cAcid used is [PhNH3][OTf] Table A2.2: UV-vis quantification results for standard NH3 generation experiments with P3SiFeN2 Entry P3SiFeN2 (μmol) Acid equi v 108a 108a A 2.3 B 2.3 Avg. a Acid used is [Ph2NH2][OTf] Cp*2Co NH4Cl equiv (μmol) 54 54 6.6 2.7 N2H5Cl (μmol) Equiv NH3/Fe % Yield NH3 Based on e- 0.0 0.0 1.7 0.7 1.2 ± 0.2 9.3 3.8 6.5 ± 0.3 Table A2.3: UV-vis quantification results for standard NH3 generation experiments with P3BCoN2Entry P3BCoN2- Acid (μmol) equi v A 2.3 108a Cp*2Co equiv NH4Cl (μmol) N2H5Cl (μmol) Equiv NH3/Fe 54 3.0 0.0 1.3 % Yield NH3 Based on e7.2 123 B 2.3 108a Avg. a Acid used is [Ph2NH2][OTf] 54 1.8 0.0 0.8 1.1 ± 0.4 4.4 6±2 Table A2.4: UV-vis quantification results for standard NH3 generation experiments with P3SiCoN2 Entry P3SiCoN2 (μmol) Acid equiv A 2.3 108a B 2.3 108a Avg. a Acid used is [Ph2NH2][OTf] Cp*2Co equiv NH4Cl (μmol) 54 54 0.0 0.0 N2H5Cl Equiv (μmol) NH3/Fe 0.0 0.0 0.0 0.0 0.0 % Yield NH3 Based on e0.0 0.0 0.0 A2.3.2 Ammonia production studies with [Ph215NH2][OTf]: The procedure was the same as the general procedure presented in section 3.1 with 2.3 μmol of P3BFe+ catalyst, 54 equiv Cp*2Co, and 108 equiv [Ph215NH2][OTf]. Product analyzed by 1H NMR as described in section 1.4 and only the diagnostic triplet of [14NH4][Cl] is observed. 124 Figure A2.1: 1H NMR spectrum (300 MHz, DMSO-d6) of [14NH4][Cl] produced from catalytic N2-to-NH3 conversion conducted with P3BFe+ catalyst, 54 equiv Cp*2Co, and 108 equiv [Ph215NH2][OTf] under an atmosphere of 14N2. A2.4. NH3 Generation Reaction with Periodic Substrate Reloading – Procedure with P3BFe+: All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, the precatalyst (2.3 μmol) was weighed into a vial. The precatalyst was then transferred quantitatively into a Schlenk tube using THF. The THF was then evaporated to provide a thin film of precatalyst at the bottom of the Schlenk tube. The tube is then charged with a stir bar and the acid and reductant are added as solids. The tube is then cooled to 77 K in a cold well. To the cold tube is added 1 mL of Et2O. The temperature of the system is allowed to equilibrate for 5 minutes and then the tube is sealed with a Teflon screw-valve. The cold well cooling bath is switched from a N2(l) bath to a dry ice/acetone bath. In the cold well the mixture in the sealed tube thaws with stirring and is allowed to stir at -78 °C for 3 hours. Then, without allowing the tube to warm above -78 °C, the cold well bath is switched from dry ice/acetone to N2(l). After fifteen minutes the reaction mixture is observed to have frozen, at this time the tube is opened. To the cold tube is added acid (324 equiv) and reductant (162 equiv) as solids. To the tube then 1 additional mL of Na/K-dried Et2O is added. The cold well cooling bath is switched from a N2(l) bath to a dry ice/acetone bath. In the cold well the mixture in the sealed tube thaws with stirring and is allowed to stir at -78 °C for 3 hours. These reloading steps are repeated 125 the desired number of times. Then the tube is allowed to warm to RT with stirring, and stirred at RT for 5 minutes. Table A2.5: UV-vis quantification results for NH3 generation experiments with P3BFe+, with reloading A B Number P3BFe+ Acid of (μmol) equiv Loadings [322] 2 2.3 x2a [322] 2 2.3 x2a Cp*2Co equiv [322] x3a [322] x3a [162]x3 Avg. C 3 2.3 D 3 2.3 Avg. a Acid used is [Ph2NH2][OTf] [162]x2 [162]x2 [162]x3 NH4Cl N2H5Cl Equiv % Yield (μmol) (μmol) NH3/Fe Based on H+ 0.1 115.0 49.6 46.2 0.0 145.6 62.8 58.5 56 ± 9 52 ± 9 0.3 182.4 78.7 48.9 0.1 207.3 89.5 55.5 84 ± 8 52 ± 5 A2.5. Time-resolved H2 quantification of background acid and Cp*2Co reactivity: Inside of a nitrogen filled glovebox, solid acid (0.248 mmol) and Cp*2Co (0.124 mmol) are added to a 260 mL glass tube charged with a stir bar. The vessel is sealed with a septum at RT and subsequently chilled to -196 °C in a cold well in the nitrogen filled glovebox. Et2O (1 mL) is added via syringe into the vessel and completely frozen. The vessel is passed out of the glovebox into a liquid N2 bath, and subsequently thawed in a dry ice/acetone bath with stirring at ~900 rpm. The timer was started as soon as the vessel was 126 transferred to the dry ice/acetone bath. The headspace of the reaction vessel was periodically sampled with a sealable gas sampling syringe (10 mL), which was loaded into a gas chromatograph, and analyzed for the presence of H2(g). From these data, the percent H2 evolved (relative to Cp*2Co) was calculated, correcting for the vapor pressure of Et2O and the removed H2 from previous samplings. Each time course was measured from a single reaction maintained at -78 °C. Table A2.6: Time-resolved H2 quantification for the reaction of Cp*2Co and acid in Et2O at -78 °C in the absence of an Fe precatalyst Acid Time (min) H2(g) (μmol) 10 1.0 ± 0.4 [Ph2NH2][OTf]a 60 2.1 ± 0.6 10 3.7 ± 0.1 [Ph2NH2][BArF4]b 60 12.7 ± 0.8 a Average of two experiments bAverage of three experiments % H2 Based on Cp*2Co 1.6 ± 0.6 3±1 6.0 ± 0.2 21 ± 1 A2.6. Mössbauer Spectra: A2.6.1 General Procedure for Preparation of Rapid-freeze-quench Mössbauer Samples of Catalytic Reaction Mixtures using P3BFe+: All manipulations are carried out inside of a nitrogen filled glovebox. The precatalyst, [P3B(57Fe)][BArF4], is weighed into a vial (3.5 mg, 2.3 μmol) and transferred 127 using THF into a 150 mL Schlenk tube. The solvent is evaporated to form a thin film of the precatalyst and a stir bar is added. The [Ph2NH2][OTf] (79.4 mg, 0.248 mmol) and Cp*2Co (40.3 mg, 0.124 mmol) are added to the Schlenk tube as solids. The Schlenk tube is then placed in N2(l) and the temperature is allowed to equilibrate. To the tube 1 mL of Et2O is added. The tube is then sealed with a Teflon screw tap and transferred to a prechilled cold well at -78 °C. The timer is set to zero as soon as the stir bar is freed from the thawing solvent. At the desired time, the tube is opened and the well-stirred suspension is transferred to a Delrin cup pre-chilled to -78 °C using a similarly pre-chilled pipette. The sample in the Delrin cup is then rapidly frozen in N2(l). At this point the sample, immersed in N2(l), is taken outside of the glovebox and mounted in the cryostat. A2.6.2 General Procedure for Preparation of Rapid-freeze-quench Mössbauer Samples of the Reaction of P3BFe+ with Reductants: All manipulations are carried out inside of a nitrogen filled glovebox. The precatalyst, [P3B(57Fe)][BArF4], is weighed into a vial (3.5 mg, 2.3 μmol) and .5 mL of THF is added. The solvent is then evaporated to provide a thin film of [P3B(57Fe)][BArF4]. To this is added the desired reductant as a solid (46.0 μmol, 20 equiv). This vial is then placed in N2(l) and the temperature is allowed to equilibrate. To this is added 1 mL of NaK-dried Et2O. The vial is then sealed with a cap and transferred to a pre-chilled cold well at -78 °C. The timer is set to zero as soon as the stir bar is freed from the thawing solvent. After five minutes using a pre-chilled pipette the well-stirred reaction mixture is transferred to a 128 Delrin cup that has been pre-chilled to -78 °C. The sample in the Delrin cup is then rapidly frozen in N2(l). At this point the sample, immersed in N2(l), is taken outside of the glovebox and mounted in the cryostat. A2.6.3 General Procedure for Fitting of Rapid-freeze-quench Mössbauer Samples: Data analysis was performed using version 4 of the program WMOSS (www.wmoss.org) and quadrupole doublets were fit to Lorentzian lineshapes. Simulations were constructed from the minimum number of quadrupole doublets required to attain a quality fit to the data (convergence of χR2). Quadrupole doublets were constrained to be symmetric, unless [P3BFe–N2][Na(12-crown-4)2] was included in the model. With [P3BFe–N2][Na(12-crown4)2] since it is known to have characteristic asymmetry we started with the observed linewidths in the authentic sample and allowed them to then relax. It is known that the exact linewidths are sensitive to the particular sample but the relative line breadth should be fairly constant. Using the non-linear error analysis algorithm provided by WMOSS, the errors in the computed parameters are estimated to be 0.02 mm s-1 for δ and 2% for ΔEq. 129 Figure A2.2: Mössbauer spectrum collected on P3B(57Fe)+ that was used for the Mössbauer experiments conducted in this paper. The parameters used to model this species are extremely similar to those used previously to model this species (δ = 0.75 mm/sec, ΔEq = 2.55 mm/sec, Γr = Γl = 0.52 mm/sec). Table A2.7. Fit parameters for P3BFe+.18 19 Component δ (mm s-1) Fit 0.75 ± 0.02 ΔEQ (mm s-1) 2.50 ± 0.05 Linewidths, ΓL/ ΓR (mm s-1) 0.54/0.58 130 A2.6.4. Details of Individual RFQ Mossbauer spectra: Figure A2.3. Mössbauer spectrum collected from a reaction quenched after 5 minutes between P3BFe+ and excess Cp*2Co (20 equiv). Raw data shown as black points, simulation as a solid red line, with components in green, blue, yellow, and purple (see Table S for parameters). Collected at 80 K with a parallel applied magnetic field of 50 mT as a suspension in Et2O. Fitting details for Figure A2.3: Four quadrupole doublets were found to be necessary to obtain an adequate simulation. The simulation parameters are given in Table S6.2. The two major species in this spectrum are well simulated as P3BFe–N2 and P3BFe–N2-. The residual signal exhibits only two well resolved absorbances but to obtain a good fit with symmetric lineshapes two additional quadrupole doublets were necessary. One of these can be identified as [P3BFe]+ based on the asymmetry in the lineshape of the right feature of P3BFe– N2. The similarity of the other two quadrupole doublets to those identified in the five- 131 minute freeze quench make this a logically consistent fit but one that is not strictly required by the data. Table A2.8. Fit Parameters for Figure A2.3. Component δ (mm s-1) ΔEQ (mm s-1) Linewidths, ΓL/ ΓR (mm s-1) A (green) B (purple) C (yellow) D (blue) 0.57 ± 0.02 0.75 ± 0.02 0.45 ± 0.02 0.40 ± 0.02 3.26 ± 0.06 0.29/0.29 Relative area 0.33 2.55 ± 0.05 0.27/0.27 0.06 1.76 ± 0.04 0.45/0.45 0.23 0.98 ± 0.02 0.48/0.45 0.39 132 Figure A2.4. Mössbauer spectrum collected from a reaction quenched after 5 minutes between P3BFe+ and excess Cp*2Cr (20 equiv). Raw data shown as black points, simulation as a solid red line, with components in green and (see Table S for parameters). Collected at 80 K with a parallel applied magnetic field of 50 mT as a suspension in Et2O. Fitting details for Figure A2.4. The two well-resolved quadrupole doublets can be simulated. The simulation parameters are given in Table S6.3. One of the two major species in this spectrum is well simulated as P3BFe–N2. The other feature has a very similar isomer shift but a significantly narrower quadrupole splitting. Given the labile nature of the N2 ligand this other species may represent a vacant neutral species or a dimeric N2 bridged species. 133 Table A2.9. Fit Parameters for Figure A2.4. Component δ (mm s-1) ΔEQ (mm s-1) A (green) B (brown) 3.22 ± 0.06 1.60 ± 0.05 0.57 ± 0.02 0.58 ± 0.02 Linewidths, ΓL/ ΓR (mm s-1) 0.29/0.29 0.71/0.71 Relative area 0.46 0.54 Figure A2.5. Mossbauer spectrum collected from a catalytic reaction quenched after 5 minutes. Conditions: [P3B(57Fe)][BArF] = 0.23 mM, [Ph2NH2][OTf] = 24.8 mM (108 equiv), and Cp*2Co 12.4 mM (54 equiv). Raw data shown as black points, simulation as a solid red line, with components in green, blue, yellow, and orange (see Table S for parameters). Collected at 80 K with a parallel applied magnetic field of 50 mT. 134 Fitting details for Figure A2.5. Four pairs of quadrupole doublets were found to be necessary to obtain an adequate simulation of these data. The simulation parameters are given in Table S6.1. The outer pair of sharp features clearly belong to P3BFeN2. The inner feature is highly suggestive of P3BFeN2- the presence of which was confirmed by freezequench EPR. The residual then consists of two sharp features which were simulated with the quadrupole doublet in yellow and a broader residual feature that is captured by the quadrupole doublet in orange. The exact isomer shift and quadrupole splitting of orange is not determined by this model but the one here is representative. Table A2.10. Simulation parameters for Mossbauer spectrum in Figure A2.6.4. Component δ (mm s-1) ΔEQ (mm s-1) A (green) B (blue) C (yellow) D (orange) 0.55 ± 0.02 0.40 ± 0.02 0.42 ± 0.02 0.93 ± 0.02 3.24 ± 0.06 Linewidths, ΓL/ ΓR (mm s-1) 0.25/0.25 Relative area 0.32 0.98 ± 0.02 0.49/0.34 0.26 1.82 ± 0.04 0.31/0.31 0.18 2.99 ± 0.06 0.87/0.87 0.24 135 Figure A2.6. Mössbauer spectrum collected from a catalytic reaction quenched after 30 minutes. Conditions: [P3B(57Fe)][BArF] = 0.23 mM, [Ph2NH2][OTf] = 24.8 mM (108 equiv), and Cp*2Co 12.4 mM (54 equiv). Raw data shown as black points, simulation as a solid red line, with components in green, purple, yellow, and orange (see Table A2.6.5 for parameters). Collected at 80 K with a parallel applied magnetic field of 50 mT. Fitting details for Figure A2.6. Four quadrupole doublets were found to be necessary to obtain an adequate simulation. The simulation parameters are given in Table A2.11. The major species in this spectrum is again well simulated as P3BFe–N2. The residual signal exhibits only three well resolved absorbances but to obtain a good fit with symmetric lineshapes three additional quadrupole doublets were necessary. One of these can be identified as [P3BFe]+ based on the asymmetry in the lineshape of the right feature of P3BFe– N2. The similarity of the other two quadrupole doublets to those identified in the five- 136 minute freeze quench make this a logically consistent fit but one that is not strictly required by the data. Table A2.11. Simulation parameters for Mossbauer spectrum in Figure A2.6. Component δ (mm s-1) ΔEQ (mm s-1) A (green) B (purple) C (yellow) D (orange) 0.55 ± 0.02 0.75 ± 0.02 0.44 ± 0.02 1.35 ± 0.02 3.24 ± 0.06 Linewidths, ΓL/ ΓR (mm s-1) 0.29/0.29 2.55 ± 0.05 0.27/0.27 0.08 1.74 ± 0.04 0.48/0.48 0.18 3.00 ± 0.06 0.67/0.67 0.22 Relative area 0.53 A2.7. EPR Spectra: A2.7.1 General Procedure for Preparation of Rapid-freeze-quench EPR Samples of Catalytic Reaction Mixtures using P3BFe+: All manipulations are carried out inside of a nitrogen filled glovebox. The precatalyst, [P3BFe][ BArF4], is weighed into a vial (3.5 mg, 2.3 μmol) and transferred using THF into a 150 mL Schlenk tube. The solvent is evaporated to form a thin film of the precatalyst and a stir bar is added. The [Ph2NH2][OTf] (79.4 mg, 0.248 mmol) and Cp*2Co (40.3 mg, 0.124 mmol) are added to the Schlenk tube as solids. The Schlenk tube is then placed in N2(l) and the temperature is allowed to equilibrate. To the tube 1 mL of Et2O is 137 added. The tube is then sealed with a Teflon screw tap and transferred to a pre-chilled cold well at -78 °C. The timer is set to zero as soon as the stir bar is freed from the thawing solvent. At the desired time, the tube is opened and the well-stirred suspension is transferred to an EPR tube that is prechilled to -78 °C using a pipette that has similarly been pre-chilled to -78 °C. The EPR sample is then rapidly frozen in N2(l). At this point the sample is quickly transferred out of the glovebox and put into N2(l) before it can warm. A2.7.2 General Procedure for Preparation of Rapid-freeze-quench EPR Samples of the Reaction of P3BFe+ with Reductants: All manipulations are carried out inside of a nitrogen filled glovebox. The precatalyst, [P3BFe][BArF4], is weighed into a vial (3.5 mg, 2.3 μmol) and .5 mL of THF is added. The solvent is then evaporated to provide a thin film of [P3BFe][BArF4]. To this is added (46.0 μmol, 20 equiv) of the desired reductant as a solid. This vial is then placed in N2(l) and the temperature is allowed to equilibrate. To this is added 1 mL of NaK-dried Et2O. The vial is then sealed with a cap and transferred to a pre-chilled cold well at -78 °C. The timer is set to zero as soon as the stir bar is freed from the thawing solvent. At the desired time, the tube is opened and the well-stirred suspension is transferred to an EPR tube that is prechilled to -78 °C using a pipette that has similarly been pre-chilled to -78 °C. The EPR sample is then rapidly frozen in N2(l). At this point the sample is quickly transferred out of the glovebox and put into N2(l) before it can warm. 138 A2.7.3 General Procedure for Preparation of EPR Samples of Cp*2Co, [P3BFe][BArF4], and [P3BFeN2][Na(12-crown-4)2]: The desired species was dissolved in 1 mL of Et2O at RT and transferred to an EPR tube. The EPR tube was then chilled to -78 °C for five minutes. It was then rapidly frozen by transfer to a bath of N2(l). Figure A2.7. The X-band EPR spectrum in a 2-MeTHF glass of 2.3 mM [P3BFe– N2][Na(12-crown-4)2] at 77K. Note that the exceeding insolubility of these species when encapsulated in a crown salt prevented its measurement in ether. We note that this species has significantly different parameters than the species in which the Na is not encapusalated 139 with a crown ether and is therefore interacting with the N2 ligand. We think this species is more representative of what a hypothetical [P3BFe–N2][Cp*2Co] species would look like if isolated. Figure A2.8. The X-band EPR spectrum in Et2O of 2.3 mM [P3BFe][BArF4] at 77K. Note this species is S = 3/2 and as such we would expect no EPR signal at this temperature as is 140 observed here. We attribute the extremely weak signal observed here to background signal from the cavity Figure A2.9. The X-band EPR spectrum in Et2O of 46 mM Cp*2Co at 77K. Decamethylcobaltocene is known to be EPR silent at 77 K but at these high concentrations it becomes apparent that there is a small S = ½ impurity present in this spectrum. This persistent impurity is observable in both freeze quenched reactions of this reductant with [P3BFe][BArF4] and in spectra of the freeze quenched catalytic reaction mixtures. 141 Figure A2.10. The X-band EPR spectrum in Et2O (1 mL) of the reaction between P3BFe+ (3.5 mg, 0.0023 mmol) and Cp*2Co (15.2 mg, 0.046 mmol) stirred for 5 minutes at -78 °C then rapidly frozen to 77 K. 142 Figure A2.11. The X-band EPR spectrum in Et2O (1 mL) of the reaction between P3BFe+ (3.5 mg, 0.0023 mmol) and Cp*2Co (40.3 mg, 0.124 mmol) and [Ph2NH2][OTf] (79.4 mg, 0.248 mmol) stirred for 5 minutes at -78 °C then rapidly frozen to 77 K. Table A2.12. A comparison of the best fitting parameters for the authentic sample of P3BFeN2- (A), the freeze quench of the reaction with the reductant (B), the freeze quench of the catalytic reaction mixture (C). Reaction A B C gx 2.304 2.295 2.298 gy 2.048 2.048 2.048 gz 2.032 2.032 2.032 143 A2.8. Details on DFT Estimates of pKa and BDE A2.8.1 Computational Estimates of pKa in Et2O: The pKa values in diethyl ether were calculated referenced to H(OEt2)2+ and were predicted on the basis of the free-energy change of the exchange reaction with H(OEt2)2+ and application of Hess’ law on the closed chemical cycle. The pKa of H(OEt2)2+ was defined as 0.0. A2.8.2 Computational Estimates of BDEs: Bond dissociation enthalpies (BDE) of X–H bonds were calculated in the gas-phase using a series of known reference compounds containing M–OH, M–H and M–NH bonds.20 The enthalpy difference between the H-atom donor/acceptor pair was calculated based on the thermochemical information provided by frequency calculations after structure optimizations using the procedure described in the general computational section. A linear plot of ΔH vs BDElit was generated to form a calibration curve (Figure A2.8.1). BDE predictions were generated by application of the line of best fit to the calculated ΔH of the unknown species. Error were calculated by application of the trend line to the calculated enthalpies of known species and comparison to their literature BDE value.20-22 Errors are reported as the average of BDEcalc-BDElit (mean signed error, MSE) and the average of the absolute values of BDEcalc-BDElit (mean unsigned error, MUE). The use of 144 the Bordwell equation for bond dissociation enthalpies is well supported by small Scalc = S(X●) – S(XH), as shown in Table A2.14. Table A2.13. Calculated ΔH values and literature BDE values used for BDE calibration Species Cr(H2O)5(OH)2+ Fe(H2O)62+ Cr(H2O)5(OOH)2+ bimFeN22+ P3SiFe-C=NH+ bipFeH22+ TrenFeOH2CpFe(CO)2H [HIPTN3N]Mo-N=NH P3SiFe-N=NMeH+ P3SiFe-C=NH P3SiFe-C=NMeH P3SiFe-C=NMeH+ Hcalc 97.735 77.985 77.175 69.255 65.905 65.475 64.105 57.455 47.715 43.915 38.915 34.375 32.955 BDElit 89 77 79 67 65 62 66 56 49 48 44 45 44 BDEcalc 90 75 75 68 66 65 64 59 51 48 44 40 39 MSE MUE Notes ref 20 ref 20 ref 20 ref 20 ref 21 ref 20 ref 20 ref 20 Truncated; ref 22 ref 21 ref 21 ref 21 ref 21 -0.9 2.1 145 Table A2.14: Calculated entropy (S) for selected XH and X● species. Species P3BFe-N=NH P3BFe=N-NH2+ P3BFe=N-NH2 Cp*Co(η4-C5Me5H)+ Cp*Cr(η4-C5Me5H)+ S(X●) (cal/mol*K) 271.6 266.3 268.9 168.8 159.5 S(XH) (cal/mol*K) 268.9 273.1 281.3 162.0 163.4 S (kcal/mol*K) 2.7x10-3 -6.8x10-3 -1.2x10-2 6.6x10-3 -3.9x10-3 146 A2.9. References 1 Robbins, J. L.; Edelstein, N.; Spencer, B.; Smart, J. C. J. Am. Chem. Soc. 1982, 104, 1882. 2 Anderson, J. S.; Moret, M. E.; Peters, J. C. J. Am. Chem. Soc. 2013, 135, 534. 3 Mankad, N. P.; Whited, M. T.; Peters, J. C. Angew. Chem. Int. Ed. 2007, 46, 5768. 4 Del Castillo, T. J.; Thompson, N. B.; Suess, D. L.; Ung, G.; Peters, J. C. Inorg. Chem. 2015, 54, 9256. 5 Suess, D. L.; Tsay, C.; Peters, J. C. J. Amer. Chem. Soc. 2012, 134, 14158. 6 Moret, M.-E.; Peters, J. C. Angew. Chem. Int. Ed. 2011, 50, 2063. 7 (a) Melzer, M. M.; Mossin, S.; Dai, X.; Bartell, A. M.; Kapoor, P.; Meyer, K.; Warren, T. H. Angew. Chem. Int. Ed. 2010, 49, 904. (b) Vicente, J.; Chicote, M. T.; Guerrero, R.; Jones, P. G. J. Chem. Soc., Dalton Trans. 1995, 8, 1251. 8 Weatherburn, M. W. Anal. Chem. 1967, 39, 971. 9 Watt, G. W.; Chrisp, J. D. Anal. Chem. 1952, 24, 2006. 10 The Spin Count program developed by the Hendrich group at Carnegie Mellon was used to convert them to the g-value and then baseline them. 11 Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178, 42. 12 Zhao, Y; Truhlar, D.G. J. Chem. Phys. 2006, 125, 1. 13 (a) Weigend F.; Ahlrichs,R. Phys. Chem. Chem. Phys., 2005, 7, 3297. (b) Weigend F. Phys. Chem. Chem. Phys., 2006, 8, 1057. 147 14 Andrae D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor. Chim. Acta, 1990, 77, 123. 15 John Towns, Timothy Cockerill, Maytal Dahan, Ian Foster, Kelly Gaither, Andrew Grimshaw, Victor Hazlewood, Scott Lathrop, Dave Lifka, Gregory D. Peterson, Ralph Roskies, J. Ray Scott, Nancy Wilkins-Diehr, "XSEDE: Accelerating Scientific Discovery", Computing in Science & Engineering, vol.16, no. 5, pp. 62-74, Sept.-Oct. 2014, doi:10.1109/MCSE.2014.80 16 Valiev, M.; Bylaska, E.J.; Govind, N; Kowalski, K; Straatsma, T.P.; van Dam, H. J. J.; Wang, D.; Nieplocha, J.; Apra, E.; Windus, T. L.; de Jong, W. A. Comput. Phys. Commun. 2010, 181, 1477. 17 Gaussian 09, Revision B.01, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken,V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian, Inc., Wallingford CT, 2009. 148 18 (a) Ribeiro, R.F.; Marenich, A.V.; Cramer, C.J.; Truhlar, D.G. J. Phys. Chem. B. 2011, 115, 14556. (b) Wang, T.; Brudvig, G; Batista, V.S. J. Chem. Theory Comput. 2010, 6, 755. (c) Marten, B.; Kyungsun, K.; Cortis, C.; Friesner, R.A.; Murphy, R.B.; Ringnalda, M.N.; Sitkoff, D.; Honig, B. J. Phys. Chem. 1996, 100, 11775. 19 Anderson, J. S.; Cutsail 3rd, G. E.; Rittle, J.; Connor, B. A.; Gunderson, W. A.; Zhang, L.; Hoffman, B. M.; Peters, J. C. J. Amer. Chem. Soc. 2015, 137, 7803. 20 Warren, J. J.; Tronic, T. A.; Mayer, J. M. Chem. Rev. 2010, 110, 6961. 21 Rittle, J.; Peters, J. C. Submitted Manuscript. 22 Yandulov, D. V.; Schrock, R.R. Inorg. Chem. 2005, 44, 1103. Appendix 3. Supplementary Information for Chapter 4 149 A3.1.Experimental Details A3.1.1. General Considerations All manipulations were carried out using standard Schlenk or glovebox techniques under an N2 atmosphere. Solvents were deoxygenated and dried by thoroughly sparging with N2 followed by passage through an activated alumina column in a solvent purification system by SG Water, USA LLC. Non-halogenated solvents were tested with sodium benzophenone ketyl in tetrahydrofuran (THF) in order to confirm the absence of oxygen and water. Deuterated solvents were purchased from Cambridge Isotope Laboratories, Inc., degassed, and dried over activated 3-Å molecular sieves prior to use. Cp*2Co,1 [P3BFe][BArF4],2 [P3BFeN2][Na(Et2O)3],3 [P3BFeN2][Na(12-crown-4)2],3 [H(OEt2)][BArF4] (HBArF4; BArF4 = tetrakis- (3,5-bis(trifluoromethyl)phenyl)borate)4, 150 NaBArF44, and [Cp*2Co][BArF4] were prepared according to literature procedures. All other reagents were purchased from commercial vendors and used without further purification unless otherwise stated. Diethyl ether (Et2O) used in the experiments herein was stirred over Na/K (≥ 2 hours) and filtered through celite before use. A3.1.2. Gas Chromatography H2 was quantified on an Agilent 7890A gas chromatograph (HP-PLOT U, 30 m, 0.32 mm ID; 30 °C isothermal; nitrogen carrier gas) using a thermal conductivity detector. A 10 mL manual injection was used and integration area was converted to percent H2 composition by use of a calibration obtained from injection of H2 solutions in N2 of known concentration. A3.1.3. Mössbauer Spectroscopy Mössbauer spectra were recorded on a spectrometer from SEE Co. (Edina, MN) operating in the constant acceleration mode in a transmission geometry. The sample was kept in an SVT400 cryostat form Janis (Wilmington, MA). The quoted isomer shifts are relative to the centroid of the spectrum of a metallic foil of α-Fe at room temperature (RT). Solution samples were transferred to a sample cup and freeze-quenched with liquid nitrogen inside of the glovebox and then immersed in liquid N2 until mounted in the cryostat. Data analysis was performed using version 4 of the program WMOSS (www.wmoss.org) and quadrupole doublets were fit to Lorentzian lineshapes. See discussion below for detailed notes on the 151 fitting procedure. A3.1.4. Ammonia Quantification Reaction mixtures are cooled to 77 K and allowed to freeze. The reaction vessel is then opened to atmosphere and to the frozen solution is slowly added a twofold excess (with respect to acid) solution of a NaOtBu solution in MeOH (0.25 mM) over 1-2 minutes. This solution is allowed to freeze and a Schlenk tube adapter is added and the headspace of the tube is evacuated. After sealing the tube is then allowed to warm to RT and stirred at RT for at least 10 minutes. An additional Schlenk tube is charged with HCl (3 mL of a 2.0 M solution in Et2O, 6 mmol) to serve as a collection flask. The volatiles of the reaction mixture are vacuum transferred at RT into this collection flask. After completion of the vacuum transfer, the collection flask is sealed and warmed to RT and stirred vigorously for 10 minutes. Solvent is removed in vacuo, and the remaining residue is dissolved in DMSO-d6 containing 20 mM 1,3,5-trimethoxybenzene as an internal standard. The ammonium chloride is quantified by integration relative to the 1,3,5-trimethoxybenzene internal standard. A3.1.5. Computational Methods All stationary point geometries were calculated using DFT-D3 (Grimmes D3 dispersion correction5) with an TPSS functional,6 a def2-TZVP7 basis set on transition metals and a def2-SVP7 basis set on all other atoms. Calculations were performed, in part, using Xtreme 152 Science and Engineering Discovery Environment (XSEDE) resources.8 Calculations were performed on the full P3BFe scaffold. Geometries were optimized using the NWChem 6.5 package.9 All single point energy, frequency and solvation energy calculations were performed with the ORCA package.10 Frequency calculations were used to confirm true minima and to determine gas phase free energy values (Ggas). Single point solvation calculations were done using an SMD solvation model11, 12 with diethyl ether solvent and were used to determine solvated internal energy (Esoln). Free energies of solvation were approximated using the difference in gas phase internal energy (Egas) and solvated internal energy (∆Gsolv ≈ Esoln – Egas) and the free energy of a species in solution was then calculated using the gas phase free energy (Ggas) and the free energy of solvation (Gsoln = Ggas + ∆Gsolv).13,14 All reduction potentials were calculated referenced to Fc+/0 and using the standard Nernst relation G = -nFE0. A3.2. Synthetic Details A3.2.1. General Procedure for the Synthesis of the Anilinium Triflates Prior to use the amine was purified (aniline and 2,6-dimethylaniline by distillation and the remaining substituted anilines by sublimation). To a 100 mL round bottom flask in the glovebox was added the desired aniline which was subsequently dissolved in 50 mL of Et2O (no additional drying with NaK). To this was added dropwise (1 equiv) of HOTf with stirring over five minutes. Immediate precipitation of white solid was observed and the reaction mixture was allowed to stir for thirty minutes. The reaction mixture was then filtered and the 153 resulting white powder was washed with Et2O (50 mL) and pentane (50 mL). The resulting white microcrystalline material was then dried under vacuum. Yields of greater than 90% of microcrystalline material was obtained in this manner in all cases. The 1H NMR spectroscopy matched literature reports.15,16 A3.3. Ammonia Generation Details A3.3.1. Standard NH3 Generation Reaction Procedure All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, the precatalyst (2.3 μmol) was weighed into a vial. The precatalyst was then transferred quantitatively into a long tube with a female 24-40 joint at the top using THF. The THF was then evaporated to provide a thin film of precatalyst at the bottom of tube. The tube is then charged with a stir bar, the acid (108 eq), and decamethylcobaltocene (41.2 mg, 54 eq) as solids. The tube is then sealed at room temperature with a septum that is secured with copper wire (this ensures a known volume of N2 in the reaction vessel, which is important for H2 detection). The tube is then chilled to 77 K and allowed to equilibrate for 10 minutes. To the chilled tube is added 1 mL of Et2O. The temperature of the system is allowed to equilibrate for 5 minutes. This tube is passed out of the box into a liquid N2 bath and transported to a fume hood. The tube is then transferred to a dry ice/acetone bath where it thaws and is allowed to stir at −78 °C for four hours. At this point the headspace of the tube is sampled with a 10 mL gas, locking syringe which is used to analyze for H2. The tube is then allowed to warm to RT with stirring and then stirred at room temperature for a further ten minutes. At this point the previously described procedure for quantifying ammonia was 154 employed. To ensure reproducibility, all experiments were conducted in 395 mL tubes (51 mm OD) using 25 mm stir bars, and stirring was conducted at ~650rpm. Table A3.1. NMR quantification results for standard NH3 generation experiments with P3BFe+ Acid Integration Relative % Yield NH3 % to (error) Yield (error) Internal Standard [4-methoxyanilinium][OTf] 0.01, 0.02 0.2(0.1) 89.1(0.2) [anilinium][OTf] 3.42, 3.33 40.4(0.5) 48.6(0.7) [2,6-dimethylanilinium][OTf] 4.30, 3.63 47.5(4.0) 37.8(0.2) [2-chloroanilinium][OTf] 4.98, 4.92 59.3(0.4) 26.1(1.9) [2,5-dichloroanilinium][OTf] 6.78, 6.15 77.5(3.8) 10.5(1.1) [2,6-dichloroanilinium][OTf] 6.81, 6.00 76.7(4.9) 12.6(2.5) [2,6- 6.60, 5.81 74.4(4.7) 14.2(3.4) 4.12, 3.0 42.7(6.7) 18.8(0.8) 5.73, 6.10 70.9(2.2) 12.0(0.8) dichloroanilinium][OTf]* [2,6dichloroanilinium][BArF4] [2,4,6trichloroanilinium][OTf] H2 155 [pentachloroanilinium][OTf] 1.62, 1.70 19.9(0.5) 63.5(1.1) *Run performed with [P3BFeN2][Na(Et2O)3] as the precatalyst. A3.4. H2 Monitoring Details A3.4.1. Standard Background Generation Reaction Procedure All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, a long tube with a female 24-40 joint is charged with a stir bar, the acid (108 eq) and decamethylcobaltocene (41.2 mg, 54 eq). The tube is then sealed at room temperature with a septum that is secured with copper wire. The tube is then chilled to 77 K and allowed to equilibrate for 10 minutes. To the chilled tube is added 1 mL of Et2O. The temperature of the system is allowed to equilibrate for 5 minutes. This tube is passed out of the box into a liquid N2 bath and transported to a fume hood. The tube is then transferred to a dry ice/acetone bath where it thaws and is allowed to stir at −78 °C for four hours. At this point the headspace of the tube is sampled with a 10 mL gas, locking syringe which is used to analyze for H2. Table A3.2: Data for Background H2 Quantification Experiments Acid GC Integration for H2 % Yield H2 4-methoxyanilinium triflate 49.8 31.5 anilinium triflate 24.0 15.2 156 2,6-dimethylanilinium triflate 8.2 5.2 2-chloroanilinium triflate 47.2 29.9 2,5-dichloroanilinium triflate 37.1 23.5 2,6-dichloroanilinium triflate 77.8 49.2 2,4,6-trichloroanilinium 34.8 22.0 98.3 62.3 triflate pentachloroanilinium triflate A3.4.2. H2 Evolution Kinetics All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. For the catalyzed run, the precatalyst was then transferred quantitatively into a Schlenk tube using THF. The THF was then evaporated to provide a thin film of precatalyst at the bottom of the long tube with a female 24-40 joint. The tube is then charged with a stir bar and the 2,6- dichloroanilinium triflate (77.9 mg, 108 eq) and decamethylcobaltocene (41.2 mg, 54 eq) are added as solids. The tube is then sealed at room temperature with a septum that is secured with copper wire. The tube is then chilled to 77 K and allowed to equilibrate for 10 minutes. To the chilled tube is added 1 mL of Et2O. The temperature of the system is allowed to equilibrate for 5 minutes. This tube is passed out of the box into a liquid N2 bath and transported to a fume hood. The tube is then transferred to a dry ice/acetone bath where it thaws and is allowed to stir at −78 °C. As soon as the stir bar is freed from the frozen solution 157 and stirring begins the timing is started. At the time points noted below the headspace was sampled for H2 with a 10 mL gas tight syringe. Table A3.3. Time points for catalyzed H2 evolution from 2,6-dichloroanlinium triflate and decamethylcobaltocene Time GC Integration for H2 % Yield H2 (error) 5 3.8, 6.4 3.3(0.9) 15 11.6, 16.9 9.3(1.8) 25 14.7, 26.2 13.4(3.8) 35 22.5, 20.8 13.9(0.5) Table A3.4. Time points for uncatalyzed H2 evolution from 2,6-dichloroanlinium triflate and decamethylcobaltocene Time GC Integration for H2 % Yield H2 (error) 158 5 3.3, 2.9 2.0(0.1) 15 7.0, 6.2 4.3(0.3) 25 8.8, 11.1 6.3(0.8) 65 20.7, 27.0 14.5(1.7) 18 16 14 %H2 Evolution 12 10 8 6 4 2 0 0 10 20 30 40 Time (min) Catalyzed H2 Formation 50 60 Background H2 Evolution 70 159 Figure A3.1: Comparison of catalyzed and uncatalyzed H2 evolution from 2,6dichloroanlinium triflate and decamethylcobaltocene at early time points. A3.5. Mössbauer Spectroscopy A3.5.1. General Procedure for Freeze-Quench Mössbauer Spectroscopy All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, the desired 57Fe species (0.0023 mmol) is quantitatively transferred using THF to a vial and then evaporated to yield a thin film. That vial is charged with a small stir bar and the other reagents as solids. The vial is then chilled to 77 K in a liquid nitrogen bath and allowed to equilibrate for five minutes. To the chilled tube is added 1 mL of Et2O and this allowed to equilibrate for another five minutes. The vial is then transferred to a cold well that has been pre-cooled for at least fifteen minutes to −78 °C with a dry ice/acetone bath. When the stir bar is freed from the frozen solvent and begins to stir the time is started. At the time noted the stirring is stopped and using a prechilled pipette the reaction mixture is transferred in one portion to a pre-chilled Mössbauer cup sitting in a vial. The vial is then placed in a liquid nitrogen bath causing the reaction mixture to freeze in approximately twenty seconds. The Mössbauer cup is then submerged in the liquid nitrogen and then removed from the glovebox and standard procedure is used to mount the sample on the Mössbauer spectrometer. 160 Figure A3.2. Mössbauer spectrum collected from a reaction freeze quenched after stirring for 5 minutes at −78 °C in 1 mL of Et2O between [P3B(57Fe)N2][Na(Et2O)3] and excess 2,6-dichloroanilinium triflate (50 equiv). Raw data shown as black points, simulation as a solid red line, with components in (see Table A3.3 for parameters). The spectrum was collected at 80 K with a parallel applied magnetic field of 50 mT as a suspension in Et2O. Fitting details for Figure A3.2: Three quadrupole doublets were found to be necessary to obtain an adequate simulation. Although a variety of parameters could potentially simulate the relatively broad absorptions observed here, previous reactivity of [P3BFeN2]- with acid suggested that P3BFeN2 and [P3BFe]+ were likely products. Satisfyingly if one of those components was provided to the fitting program the other major component was found to be the other species after refining freely. The third species was unchanged in these simulations and represents an unknown species but is demanded by the inflection point on the more negative side of the right-hand absorbance. Modeling this feature also helps to capture the asymmetry of the left-hand absorbance while using the symmetric line-shapes we expect for 161 P3BFeN2 (green) and P3BFe+ (purple). The broad linewidths for P3BFe+ have been observed previously and may be explained by the existence of unbound and bound varieties of the species with the reaction mixture providing potential ligands such as OTf-, 2,6-ClPhNH, and N2. Table A3.5: Simulation parameters for Mossbauer spectrum in Figure A3.2. Component δ (mm s-1) ΔEQ (mm s-1) Linewidths, Relative area ΓL/ ΓR (mm s-1) A (green) 0.58 ± 0.02 3.28 ± 0.07 0.52/0.52 0.26 B (purple) 0.76 ± 0.02 2.57 ± 0.05 1.10/1.10 0.63 C (yellow) 0.13 ± 0.02 2.24 ± 0.04 0.50/0.50 0.11 A3.6. EPR Spectroscopy A3.6.1 General Procedure for EPR Spectroscopy All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, the desired Fe species (0.0023 mmol) is quantitatively transferred using THF to a vial and then evaporated to yield a thin film. That vial is charged with a small stir bar and the 162 acid (0.116 mmol, 50 eq) as solids ([2,6-dichloroanilinium][OTf] or [2,6- dichloroanilinium][BarF4]). The vial is then chilled to 77 K in a liquid nitrogen bath and allowed to equilibrate for five minutes. To the chilled tube is added 1 mL of Et2O (for HOTf 50 eq have been dissolved in this 1 mL of Et2O at room temperature) and this allowed to equilibrate for another five minutes. The vial is then transferred to a cold well that has been pre-cooled for at least fifteen minutes to −78 °C with a dry ice/acetone bath. When the stir bar is freed from the frozen solvent and begins to stir the time is started. The reaction mixture is stirred for five minutes and then stirring is stopped. Using a pre-chilled pipette approximately 0.5 mL of the reaction mixture is rapidly transferred to a pre-chilled X-band EPR tube. The X-band EPR tube is then placed in a liquid nitrogen bath causing the reaction mixture to freeze in approximately twenty seconds. The EPR tube is then sealed and removed from the glovebox in liquid nitrogen. A3.6.2. Comment on Stoichiometric Reactivity In our attempt to model the catalytic reaction mixture we were interested in the reactivity of [P3BFeN2]- (observed previously both from mixing [P3BFe][BarF4] with excess Cp*2Co and under the catalytic reaction conditions) with acid. In order to achieve this we wanted to prepare independently known [P3BFeN2]- species to model the proposed catalytic intermediate [P3BFeN2][Cp*2Co]. We chose [P3BFeN2][Na(Et2O)3] because its solubility in Et2O modeled that of [P3BFeN2][Cp*2Co]. 163 Figure A3.3. The continuous wave, X-band EPR at 77K in Et2O of reaction mixtures freeze-quenched after five minutes. In red is the reaction of [P3BFeN2][Na(Et2O)3] with 50 eq of [2,6-dichloroanilinium][BArF4] clearly demonstrating the formation of [P3BFeNNH2][BArF4]. In green is reaction of [P3BFeN2][Na(Et2O)3] with 50 eq of [2,6dichloroanilinium][OTf] in which the small residual species is neither the starting material ([P3BFeN2][Na(Et2O)3]) or the desired product ([P3BFeNNH2][OTf]). Although we do not know the chemical identity of this species we note that it is very similar to the EPR 164 observed in the reaction of [P3BFeN2][Na(12-crown-4)2] with 1 eq. of HBArF4.17 We hypothesize therefore that it may represent a Fe–H side product. Figure A3.4. In blue is the continuous wave, X-band EPR spectrum at 77K of a reaction mixture of 50 eq. [2,6-dichloroanilinium][BarF4] with [P3BFeN2][Na(12-crown-4)2] quenched with liquid nitrogen after 5 minutes. In orange is the simulation of this spectrum (fitting details below) Fitting details for Figure A3.4: The parameters used to fit the spectrum were obtained using the esfit application in the easyspin program.18 The fitting program obtains the best fit by minimizing the root mean square deviation from the data. 165 The data was fit with the following parameters: g1 = 2.23899, g2 = 2.09189, g3 = 2.00664, and a line broadening of 323.8530, 71.2309, and 38.7902 MHz respectively. These parameters represent only a very small perturbation from those used previously to model [P3BFeNNH2][BarF4]: g1 = 2.222, g2 = 2.091, g3 = 2.006 and a line broadening of 256, 113, and 41 MHz respectively.17 The slightly broader spectrum observed here precludes resolution of the small phosphorus coupling on g3. We believe that this broadening arises from either the use of a non-glassing solvent (Et2O vs 2-MeTHF) or via small differences in hydrogen-bonding that arise from the presence of 2,6-dichloroaniline. A3.7. Acid Quench of P3BFeN2A3.7.1 Standard Acid Quench Procedure All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, the desired Fe species (2.3 μmol) was weighed into a vial. The precatalyst was then transferred quantitatively into a Schlenk tube using THF. The THF was then evaporated to provide a thin film of Fe species at the bottom of the Schlenk tube. The tube is then charged with a stir bar and [2,6-dichloroanilinium][OTf (36.2 mg, 115 μmol, 50 eq) is added as a solid. The tube is then sealed at room temperature with a septum and a Konte’s valve that is left partially open. The tube is then chilled to 77 K and allowed to equilibrate for 10 minutes. To the chilled tube is added 1 mL of Et2O through the septum. The temperature of the system is allowed to equilibrate for 5 minutes and then the Konte’s valve is sealed. This tube is passed out of the box into a liquid N2 bath and transported to a fume hood. The tube is then 166 transferred to a dry ice/acetone bath where it thaws and is allowed to stir at −78 °C for three hours. At the end of the reaction the Konte’s valve is opened and the reaction headspace is allowed to equilibrate. At this point the headspace of the tube is sampled with a 10 mL gas, locking syringe which is used to analyze for H2. The tube is then allowed to warm to RT with stirring and then stirred at room temperature for a further ten minutes. At this point the previously described procedure for quantifying ammonia was employed. Table A3.6. Comparative NH3 and H2 Yields for [2,6-ClPhNH3][OTf] and [2,6ClPhNH ][BArF ] 3 4 Acid Integration Relative % Yield % Internal NH3 Yield % Yield H2 H2 (error) Standard (error) [2,6-dichloroanilinium][OTf] 0.00, 0.00 0.0(0.0) 39.0, 48.2 43.7(4.6) [2,6-dichloroanilinium][BArF4] 0.36, 0.25 0.20(0.03) 30.3, 45.6 37.8(7.6) A3.8. Solubility Measurement A3.8.1. Procedure for Measuring Solubility of Cp*2Co All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, a Schlenk tube is charged with a stir bar and the Cp*2Co (41.2 mg, 0.125 mmol) 167 is added to the tube. The tube is then chilled to 77 K in a liquid nitrogen bath and allowed to equilibrate for 5 minutes. To the chilled tube is added 1 mL of Et2O. The temperature of the system is allowed to equilibrate for 5 minutes and then the Schlenk tube is transferred to the cold well which has been prechilled to −78 °C for fifteen minutes. After five minutes of stirring at ~620 rpm, the stirring is stopped. With a prechilled pipette the entirety of the reaction mixture is transferred to a similarly prechilled celite pad for filtration. Filtration yielded a pale green solution that was then warmed to room temperature and the solvent was removed under reduced pressure. The vial was then extracted with a 20 mM solution of 1,3,5trimethoxybenzene in C6D6. The NMR was then measured and the Cp*2Co signal was integrated relative to the 1,3,5-trimethoxybenzene standard. The accuracy of this integration procedure was confirmed by performing this procedure on a sample of Cp*2Co that had simply been weighed into a vial. Repetition of this experiment resulted in Cp*2Co concentrations between 5-6 mM. A3.8.2. Procedure for Measuring Solubility of [2,6-dichloroanilinium][OTf]: All solvents are stirred with Na/K for ≥2 hours and filtered prior to use. In a nitrogen-filled glovebox, a Schlenk tube is charged with a stir bar and the [2,6-dichloroanilinium][OTf] (77.9 mg, 0.250 mmol) is added to the tube. The tube is then chilled to 77 K in a liquid nitrogen bath and allowed to equilibrate for 5 minutes. To the chilled tube is added 1 mL of Et2O. The temperature of the system is allowed to equilibrate for 5 minutes and then the Schlenk tube is transferred to the cold well which has been prechilled to −78 °C for fifteen minutes. After five minutes of stirring at ~620 rpm, the stirring is stopped. With a prechilled 168 pipette the entirety of the reaction mixture is transferred to a similarly prechilled celite pad for filtration. Filtration yielded a colorless solution that was then warmed to room temperature and the solvent was removed under reduced pressure. The vial was then extracted with a 20 mM solution of 1,3,5-trimethoxybenzene in THF-d8. The NMR was then measured and the two signals for [2,6-dichloroanilinium][OTf] were integrated relative to the 1,3,5-trimethoxybenzene standard. The result was a [2,6-dichloroanilinium][OTf] concentration of 0.4 mM. A3.9. Controlled Potential Electrolysis (CPE) and Cyclic Voltammetry (CV) Details General considerations. All manipulations are carried out in an N2 filled glove box. For CPE experiments a sealable H-cell consisting of two compartments separated by a fine porosity sintered glass frit is cooled to −35 °C in a cold well and charged with 4 mL (working chamber) and 4 mL (auxiliary chamber) of 0.1 M NaBArF4 solution in Et2O, the solutions are also cooled to −35 °C and the solution for the working chamber may contain additional chemical components as described below. The working chamber is outfitted with a glassy carbon working electrode, rectangular prismatic in shape with dimensions of 10 mm x 2 mm and submerged in the working chamber solution to a depth of ~10 mm. The working chamber is also equipped with a Ag/AgPF6 in 0.1 M NaBArF4 Et2O reference electrode isolated by a CoralPor™ frit (obtained from BASi) and referenced externally to Fc/Fc+. The auxiliary chamber is outfitted with a solid sodium auxiliary electrode (~5 mm by ~1 mm rectangular prism, submerged to ~5 mm). The cell is sealed before electrolysis. The cell is connected to 169 a CH Instruments 600B electrochemical analyzer and controlled potential bulk electrolysis experiments were performed at −35 °C with stirring, cold well external bath temperature maintained by a SP Scientific FTS Systems FC100 immersion cooler. CV experiments are conducted in a single compartment cell cooled to −35 °C in a cold well in 0.1 M NaBArF4 Et2O solution, again cold well external bath temperature maintained by a SP Scientific FTS Systems FC100 immersion cooler. The working electrode is a glassy carbon disk, the reference electrode is a Ag/AgPF6 in 0.1 M NaBArF4 Et2O reference electrode isolated by a CoralPor™ frit (obtained from BASi) and referenced externally to Fc/Fc+, the auxiliary electrode is a platinum wire. Measurements conducted with a CH Instruments 600B electrochemical analyzer General methodology for controlled potential electrolysis experiments: To the working chamber is added 3 mg of [P3BFe][BArF4] (2 μmol), 100 μmol of acid (e.g. [Ph2NH2][OTf]), 0-23.8 mg of [CoCp*2][BArF4] (0-20 μmol), and a magnetic stir bar. The cell is held at a working potential of −2.1 V vs Fc/Fc+ until the current passed in the cell falls to 1% of the initial current pass or until 21.5 hours have passed. After that time the potential bias is removed, the headspace of the cell is sampled with a sealable gas syringe (10 mL), which is immediately analyzed by GC for the presence of H2. Then an additional 100 μmol of acid in 2 mL 0.1 M NaBArF4 solution in Et2O is injected through rubber septa into both chambers to sequester NH3 as [NH4][OTf]. The cell is allowed to stir at -35 °C for 10 minutes and then warmed to room temperature. The contents of both chambers are then transferred to a 170 Schlenk tube (cell washed with additional Et2O) and this material is analyzed for NH3 by base digestion, vacuum transfer of volatiles, and NMR integration as described in section A1.4 Methodology for controlled potential electrolysis experiments with reloading of substrate: To the working chamber is added 3 mg of [P3BFe][BArF4] (2 μmol), 100 μmol of acid (e.g. [Ph2NH2][OTf]), 0-23.8 mg of [CoCp*2][BArF4] (0-20 μmol), and a magnetic stir bar. The cell is held at a working potential of −2.1 V vs Fc/Fc+ until the current passed in the cell falls to 1% of the initial current pass or until 21.5 hours have passed. After that time the potential bias is removed. An additional 100 μmol of acid in 2 mL 0.1 M NaBArF4 solution in Et2O is then added to the working chamber of the cell via injection through a rubber septum. The cell is then held at a working potential of −2.1 V vs Fc/Fc+ until the current passed in the cell falls to 1% of the initial current pass or until 21.5 hours have passed. After that time the potential bias is removed, the headspace of the cell is sampled with a sealable gas syringe (10 mL), which is immediately analyzed by GC for the presence of H2. Then an additional 100 μmol of acid in 2 mL 0.1 M NaBArF4 solution in Et2O is injected through rubber septa into both chambers of the cell to sequester NH3 as [NH4][OTf]. The cell is allowed to stir at -35 °C for 10 minutes and then warmed to room temperature. The contents of both chambers are then transferred to a Schlenk tube (cell washed with additional Et2O) and this material is analyzed for NH3 by base digestion, vacuum transfer of volatiles, and NMR integration as described in section A1.4 171 Table A3.7. Controlled Potential Electrolysis Data. Entry Acid Equiv Time Charge Yield of FE [CoCp*2] (h) Passed NH3 (C) (equiv [BArF4] NH3 FE (%) (%) per Fe) 1 [Ph2NH2][OTf] 0 42 7.5 2.3 18 80 2 [Ph2NH2][OTf] 0 63 6.2 2.8 26 25 2.6 ± 0.4 22 ± 6 Avg 3b [Ph2NH2][OTf] 0 43 7.5 2.2 17 67 4 [Ph2NH2][OTf] 1 17 8.1 4.4 31 56 5 [Ph2NH2][OTf] 1 22 8.3 3.5 24 47 4.0 ± 0.6 28 ± 5 Avg 6 [Ph2NH2][OTf] 5 17 8.5 3.9 26 61 7 [Ph2NH2][OTf] 5 21 9.1 3.5 22 57 8 [Ph2NH2][OTf] 5 22 9.5 4.6 28 27 4.0 ± 0.6 25 ± 3 Avg 9 [Ph2NH2][OTf] 10 21 9.4 3.0 19 64 10 [Ph2NH2][OTf] 10 10 10.2 5.1 29 47 H2a 172 Avg 4±1 24 ± 7 11 [PhNH3][OTf] 5 15 9.0 1.2 8 48 12 [PhNH3][OTf] 5 22 7.8 0.6 4 35 0.9 ± 0.4 6±3 Avg 13 [2,6-Cl2 5 17 10.6 2.0 11 44 5 17 10.7 1.7 9 41 1.9 ± 0.2 10 ± 1 PhNH3][OTf] 14 [2,6-Cl2 PhNH3][OTf] Avg 15b [Ph2NH2][OTf] 5 32 17.3 6.1 20 43 16b [Ph2NH2][OTf] 5 22 18.7 6.7 21 32 17b [Ph2NH2][OTf] 5 37 13.7 4.7 20 38 18b [Ph2NH2][OTf] 5 41 15.3 4.8 18 52 19b [Ph2NH2][OTf] 5 43 17.8 5.4 18 31 5.5 ± 0.9 19 ± 1 Avg 20Ac [Ph2NH2][OTf] 5 21.5 9.5 4.6 28 27 20Bc [Ph2NH2][OTf] 5 11.5 9.2 0.0 0 88 173 21d [Ph2NH2][OTf] 5 16 9.2 0.0 0 75 22e [Ph2NH2][OTf] 5 43 0.0 0.3 N/A N/A 23f [Ph2NH2][OTf] Chemical 21.5 N/A 1.3 7.8 e- 50 e- 21.5 N/A 2.3 13.8 e- 31 e- 1.8 ± 0.7 11 ± 4 runs 24f [Ph2NH2][OTf] Chemical runs Avg a Some ports of the cell are sealed with septa and one of these is pierced before the electrolysis begins to pressure equilibrate the cell as it cools to -35 °C, we note therefore that H2 gas may escape from the cell particular during long experiments, indeed a test of H2 retention in the cell under equivalent conditions revealed leakage of H2 thus the detected % yield of H2 reported here should be considered a lower limit. bThese experiments were conducted using the reloading protocol as described above. cElectrode rinse test as described in main text. d Control experiment with no [P3BFe][BArF4] included but including a typical loading of 11.9 mg (10 μmol) of [CoCp*2][BArF4]. eControl experiment where the cell, with all components including the sodium auxiliary electrode, was assembled and stirred at -35 °C for 43 hours but no potential bias was applied. fChemical catalysis runs at -35 °C in 0.1 M NaBArF4 Et2O solution with 50 equiv (100 μmol) of Cp*2Co included as a chemical reductant as well as [P3BFe][BArF4] (2 μmol) and 100 μmol of acid ([Ph2NH2][OTf]). 174 A3.10. Computational Details A3.10.1. Calculation of Acid Dissociation Constants Acid dissociation constants (pKa and pKd) were performed were optimized and solvated as discussed in the general methods section. For pKa values, the G for the exchange of a proton (H+) between the acid of interest and 2,6-ClPhNH2/2,6-ClPhNH3+. For pKd values, the same approach was used except that the net exchange of a HOTf unit was calculated. In all cases the dissociation constant was reference to the literature value for the pKa of 2,6-ClPhNH3+ in THF. A4.10.2. Determination of PT, ET and PCET Kinetics Kinetic barriers for reported for PT, ET and PCET were performed in one of two ways. Internal consistency between the methods was determined where possible. Values are summarized in Table A3.8. Method A. Marcus Theory. Standard Marcus theory expressions19 were used in method A. Inner sphere reorganization energies for PT or PCET were calculated using the method developed by the group of Hammes-Schiffer (Eq. A3.1) utilizing the force constants for the reactant (𝑓𝑗𝑟 ) and product (𝑓𝑗𝑝 ) species and the change in equilibrium bond length (Dqj).20 𝜆𝑖𝑠,𝑃𝑇/𝑃𝐶𝐸𝑇 = 𝑓𝑗𝑟 𝑓 𝑝 ∑𝑗 𝑟 𝑗 𝑝 ∆𝑞𝑗 2 𝑓𝑗 +𝑓𝑗 175 (Eq. A3.1) Outer sphere reorganization energies were calculated using a continuum solvation model for the solvation of a point charge (λos,ET)19 or a dipole (λos,PT).20-22 The λos,PCET was approximated using equation A3.2, where θ is the angle between the ET and PT vectors.20 It was determined via analysis of the structure of a constrained optimization (in which the Fe−H–Co distance was kept constant) that θ is between 0 and 45o, a range which corresponds to an insignificant variation (less than 0.2 kcal/mol) in λos,PCET. λos,PCET = λos,PT + λos,ET – (λos,PT* λos,ET)cos(θ) (Eq. A3.2) Relative rates for a bimolecular PT/ET vs PCET (kbi) pathway for reaction shown in Table A3.8, Equation 6 were determined via the method outline by the group of Hammes-Schiffer in which the bimolecular rate constant for PT, ET or PCET is approximated by equation A3.3. kbi = KA*kuni (Eq. A3.3) KA represents the pre-arrangement equilibrium constant and kuni represents the unimolecular rate constant for PCET or ET.23 Along an PT/ET pathway, the barriers calculated suggest that kPT > kET. In approximating kuni for PCET and ET, we made extensive use of the webPCET portal.22 The electronic coupling for PCET and ET was assumed to be equal. In order to approximate a lower bound for kPCET/kET, the pre-arrangement equilibrium (KA) was also assumed to be equal for PCET and ET. We believe this represents a lower bound as the approximation for KA does not include any hydrogen bonding interactions for a PCET 176 pathway. Method B. Optimization of a 1st Order Saddle Point. PT barriers for the protonation of Cp*2Co was also performed using optimization of a 1st order saddle point. Table A3.8. Overview of Parameters Used to Calculate Kinetic Barriers 1. [2,6-ClPhNH3][OTf] + Cp*2Co  Cp*Co(η4-C5Me5H)-OTf + 2,6-ClPhNH2 2. [2,6-MePhNH3][OTf] + Cp*2Co  Cp*Co(η4-C5Me5H)-OTf + 2,6-MePhNH2 3. [4-OMePhNH3][OTf] + Cp*2Co  Cp*Co(η4-C5Me5H)-OTf + 4-OMePhNH2 4. P3BFe(NNH) + [Cp*Co(η4-C5Me5H)][OTf]  [P3BFe(NNH)2][OTf] + Cp*2Co 5. [P3BFe(NNH)2][OTf] + Cp*2Co  P3BFe(NNH2) + [Cp2*Co][OTf] 6. P3BFe(NNH) + [Cp*Co(η4-C5Me5H)][OTf]  P3BFe(NNH)2+ [Cp*2Co][OTf] Reaction λis λos Barrier Method {krel} 1 N/A N/A 1.3 kcal/mol A 1 7.5 kcal/mol 6.3 kcal/mol 1.3 kcal/mol B 2 N/A N/A 3.8 kcal/mol A 2 7.5 kcal/mol 6.3 kcal/mol 3.6 kcal/mol B 177 3 N/A N/A 4.5 kcal/mol A 3 7.5 kcal/mol 6.3 kcal/mol 4.8 kcal/mol B 4 8.9 kcal/mol 6.3 kcal/mol 1.5 kcal/mol A 5 8.9 kcal/mol 25.0 kcal/mol 4.1 kcal/mol Aa {krel ≡ 1 M-1s-1} 6 13.7 kcal/mol 0-10 kcal/mol 0.2 – 0.6 kcal/mol A {2000 – 4500 M-1s-1} a The barrier for [P3BFe(NNH2)][OTf] reduction was calculated assuming that rate- determining reduction to [P3BFe(NNH2)][OTf]− precedes OTf – release. A3.10.3. BDFE Calculations Bond dissociation free energies (BDFE) of X–H bonds were calculated in the gas-phase using a series of known reference compounds.25 The free-energy difference between the Hatom donor/acceptor pair was calculated based on the thermochemical information provided by frequency calculations after structure optimizations using the procedure described in the general computational section. A linear plot of ΔG vs BDFElit was generated to form a calibration curve (Figure A3.5.). BDFE predictions were generated by application of the line of best fit to the calculated ΔG of the unknown species. 178 Figure A3.5. BDFEcalc and BDFElit plotted for species of known BDFEE-H. Line of best fit is shown. Table A3.9. Data used to generate the plot and line of best fit shown in Figure A3.5. HOOH DG (E-H) DG (E●) DGcalc BDFEE-H -151.4 -150.8 79.7 69.8 179 MeOH -115.6 -115.0 88.3 96.4 EtOOH -230.0 -229.4 68.7 76.6 H2O -76.4 -75.7 104.2 111.0 NH3 -56.5 -55.8 94.0 99.4 Me3CH -158.3 -157.6 82.7 88.3 PhOH -307.2 -306.6 74.0 79.8 Et2NH -213.6 -212.9 81.0 86.4 NH2NH2 -111.8 -111.1 67.3 72.6 OH− -75.7 -75.0 98.6 103.1 PhSH -630.2 -629.5 70.3 75.3 NH4+ -56.8 -56.1 113.0 116.9 Me2CH2 -119.0 -118.4 85.9 90.4 HC(O)OOH -264.7 -264.1 82.2 86.8 OOH -150.8 -150.2 37.5 42.7 C6H6 -232.1 -231.4 101.6 104.7 C2H4 -78.5 -77.8 99.7 102.5 180 C2H6 -79.7 -79.1 90.0 92.9 PhCH3 -271.3 -270.7 79.0 81.6 CH4 -40.5 -39.8 95.1 96.8 CpH -193.9 -193.3 71.0 73.2 EtSH -477.8 -477.2 77.2 79.1 MeSH -438.6 -437.9 77.3 79.2 PhNH2 -287.4 -286.7 79.8 81.5 NHNH -110.6 -110.0 51.0 52.6 H2S -399.3 -398.7 83.1 83.0 H2 -1.2 -0.5 98.8 97.2 A3.11. X-ray Photoelectron Spectroscopy (XPS) details The surface composition of the carbon electrode surface after a 15 hour bulk electrolysis in the presence of P3BFe+, Cp*2Co+, [Ph2NH2][OTf] and N2 was determined via XPS on a Kratos Axis Nova spectrometer with DLD (Kratos Analytical; Manchester, UK). The excitation source for all analysis was monochromatic Al Kα1,2 (hv = 1486.6 eV) operating at 10 mA and 15 kV. The X-ray source was directed 54° with respect to the sample normal. A base pressure of 1 × 10−9 Torr is maintained in the analytical chamber, which rises to 5 × 181 10−9 Torr during spectral acquisition. All spectra were acquired using the hybrid lens magnification mode and slot aperture, resulting in an analyzed area of 700 μm × 400 μm. Survey scans were collected using 160 eV pass energy, while narrow region scans used 10 eV; charge compensation via the attached e−-flood source was not necessary in this study. Subsequent peak fitting and composition analysis was performed using CasaXPS version 2.3.16 (Casa Software Ltd.; Teignmouth, UK). Energy scale correction for the survey and narrow energy regions was accomplished by setting the large component in the C 1s spectrum, corresponding to a C 1s C(=C) transition, to 284.8 eV. All components were fitted using a Gaussian 30% Lorentzian convolution function. For quantification, Shirley baselines were employed where there was a noticeable change in CPS before and after the peak in the survey spectrum; otherwise, linear was chosen. Atomic percentages were calculated using the CasaXPS packages for regions and/or components and are reported herein. Calculations were performed using region or component areas normalized to relative sensitivity factors specific to the instrument conditions with deconvolution from the spectrometer transmission function. 182 Figure A3.6. XPS survey scan of a section of a glassy carbon plate which was not exposed to the working chamber solution during a 15 hour bulk electrolysis in the presence of P3BFe+, Cp*2Co+, [Ph2NH2][OTf] and N2 at -2.1 V (vs Fc/Fc+). XPS and Auger peaks are assigned as labeled in the legend, which also includes atomic percentages calculated from 183 component fits from scans of individual XPS regions. This material represents a baseline of the electrode surface composition resulting from cleaning, polishing, and handling prior to CPE experiments and is provided for comparison to a XPS survey scan of a section of the same glassy carbon plate which was exposed to the working chamber solution during a 15 hour bulk electrolysis in the presence of P3BFe+, Cp*2Co+, [Ph2NH2][OTf] and N2 at 2.1 V (vs Fc/Fc+) presented in Figure A3.7. 184 185 Figure A3.7. XPS survey scan of a section of a glassy carbon plate which was exposed to the working chamber solution during a 15 hour bulk electrolysis in the presence of P3BFe+, Cp*2Co+, [Ph2NH2][OTf] and N2 at -2.1 V (vs Fc/Fc+). XPS and Auger peaks are assigned as labeled in the legend, which also includes atomic percentages calculated from component fits from scans of individual XPS regions. This material represents a postelectrolysis state of the electrode surface composition for comparison to a XPS survey scan of a section of the same glassy carbon plate which was not exposed to the working chamber solution during a 15 hour bulk electrolysis in the presence of P3BFe+, Cp*2Co+, [Ph2NH2][OTf] and N2 at -2.1 V (vs Fc/Fc+) presented in figure A3.3.1.Notably this active surface scan reveals a small Fe signal attributed to some degree of decomposition of the P3BFe+ catalyst over the course of the 15 hour electrolysis, also notable is that no new Co signal is observed in the post-electrolysis scan suggesting that Cp*2Co+ does not decompose to a surface bound Co species in appreciable amounts during the electrolysis. A3.12. pKa Determination Strategy Bosch et al. published a procedure for converting a pKa in THF into the equivalent pKa in different solvents.26 Although not all of the pKa values have been experimentally determined in THF the values obtained from converting from MeCN or H2O into a THF value is quite accurate. So we have used these converted values in the text. Where available a number measured in THF has been used, if not the average of the two values has been used if available. These choices do not materially affect the analysis as the interpretation of pKa effect is qualitative in nature. Solvent conversion equations: 186 pKa(THF) = 0.78×pKa(MeCN) − 0.52 pKa(THF) = 1.19×pKa(H2O) + 2.13 pKa in pKa in Converted Experimental pKa a pKa in THF 8.8 (8.4) 8.826 7.8 (7.6) 8.026 Acid MeCN H2O 11.862 5.2929 [4-methoxyanilinium][OTf] 7 10.622 4.5829 [anilinium][OTf] 7 [2,6-dimethylanilinium][OTf] -- 3.8929 -- (6.8) [2-chloroanilinium][OTf] 7.8627 2.6429 5.6 (5.3) 6.026 [2,5-dichloroanilinium][OTf] 6.2128 1.5327 4.3 (4.0) 4.526 [2,6-dichloroanilinium][OTf] 5.0627 0.4230 3.4 (2.6) [2,4,6-trichloroanilinium][OTf] -- −0.0330 -- (2.1) [pentachloroanilinium][OTf] 2.3528 -- 1.3 (--) a First is listed the value converted from THF and then in parentheses is the value converted from H2O. A3.13. References 1 Robbins, J. L.; Edelstein, N.; Spencer, B.; Smart, J. C. Syntheses and Electronic Structures of Decamethylmetallocenes. J. Am. Chem. Soc. 1982, 104, 1882. 187 2 Anderson, J. S.; Moret, M.-E.; Peters, J. C. Conversion of Fe–NH2 to Fe–N2 with Release of NH3. J. Am. Chem. Soc. 2013, 135, 534. 3 Moret, M.-E.; Peters, J. C. N2 Functionalization at Iron Metallaboratranes. J. Am. Chem. Soc. 2011, 133, 18118. 4 Del Castillo, T. J.; Thompson, N. B.; Peters, J. C. A Synthetic Single-Site Fe Nitrogenase: High Turnover, Freeze-Quench 57Fe Mössbauer Data, and a Hydride Resting State. J. Am. Chem. Soc. 2016, 138, 5341. 5 Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104. 6 Tao, J.; Perdew, J. P.; Staroverov, V. N.; Scuseria, G. E. Climbing the Density Functional Ladder: Nonempirical Meta-Generalized Gradient Approximation Designed for Molecules and Solids. Phys. Rev. Lett. 2003, 91, 146401. 7 Weigend, F.; Ahlrichs, R. Balanced Basis Sets of Split Valence, Triple Zeta Valence and Quadruple Zeta Valence Quality for H to Rn: Design and Assessment of Accuracy. Phys. Chem. Chem. Phys. 2005, 7, 3297. 8 John Towns, Timothy Cockerill, Maytal Dahan, Ian Foster, Kelly Gaither, Andrew Grimshaw, Victor Hazlewood, Scott Lathrop, Dave Lifka, Gregory D. Peterson, Ralph Roskies, J. Ray Scott, Nancy Wilkins-Diehr, "XSEDE: Accelerating Scientific Discovery", Computing in Science & Engineering, vol.16, no. 5, pp. 62-74, Sept.-Oct. 2014, doi:10.1109/MCSE.2014.80 9 Valiev, M.; Bylaska, E. J.; Govind, N.; Kowalski, K.; Straatsma, T. P.; Van Dam, H. J. J.; 188 Wang, D.; Nieplocha, J.; Apra, E.; Windus, T. L.; de Jong, W. A. NWChem: A Comprehensive and Scalable Open-Source Solution for Large Scale Molecular Simulations. Comput. Phys. Commun. 2010, 181, 1477. 10 Neese, F. The ORCA program system. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2012, 2, 73. 11 Klamt, A.; Schüürmann, G. COSMO: A New Approach to Dielectric Screening in Solvents with Explicit Expressions for the Screening Energy and its Gradient. J. Chem. Soc. Perkin Trans. 2. 1993, 2, 799. 12 Marten, B.; Kim, K.; Cortis, C.; Friesner, R. A.; Murphy, R. B.; Ringnalda, M. N.; Sitkoff, D.; Honig, B. New Model for Calculation of Solvation Energies: Correction of SelfConsistent Reaction Field Continuum Dielectric Theory for Short-Range Hydrogen Bonding Effects. J. Phys. Chem. 1996, 100, 11775. 13 Ribeiro, R. F.; Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. Use of Solution-Phase Vibrational Frequencies in Continuum Models for the Free Energy of Solvation. J. Phys. Chem. B 2011, 115, 14556. 14 Wang, T.; Brudvig, G.; Batista, V. S. Characterization of Proton Coupled Electron Transfer in a Biomimetic Oxomanganese Complex: Evaluation of the DFT B3LYP Level of Theory. J. Chem. Theory Comput. 2010, 6, 755. 15 Hamashima, Y.; Somei, H.; Shimura, Y.; Tamura, T.; Sodeoka, M. Amine-Salt- Controlled, Catalytic Asymmetric Conjugate Addition of Various Amines and Asymmetric 189 Protonation. Org. Lett. 2004, 6, 1861. 16 Anderson, L. L.; Arnold, J.; Bergman, R. G. Proton-Catalyzed Hydroamination and Hydroarylation Reactions of Anilines and Alkenes: A Dramatic Effect of Counteranions on Reaction Efficiency. J. Amer. Chem. Soc. 2005, 127, 14542. 17 Anderson, J. S.; Cutsail, G. E., III; Rittle, J.; Connor, B. A.; Gunderson, W. A.; Zhang, L.; Hoffman, B. M.; Peters, J. C. Characterization of an Fe=N-NH2 Intermediate Relevant to Catalytic N2 Reduction to NH3. J. Am. Chem. Soc. 2015, 137, 7803. 18 Stoll, S.; Schweiger, A. EasySpin, a Comprehensive Software Package for Spectral Simulation and Analysis in EPR. J. Magn. Reson. 2006, 178, 42. 19 Marcus, R. A. On the Theory of Oxidation-Reduction Reactions Involving Electron Transfer. I. J. Chem. Phys. 1956, 24, 966. 20 Iordanova, N.; Decornez, H.; Hammes-Schiffer, S. Theoretical Study of Electron, Proton, and Proton-Coupled Electron Transfer in Iron Bi-Imidazoline Complexes. J. Am. Chem. Soc. 2001, 123, 3723. 21 Koper, M. T. M. Theory of Multiple Proton–Electron Transfer Reactions and its Implications for Electrocatalysis. Chem. Sci. 2013, 4, 2710. 22 Onsager, L. Electric Moments of Molecules in Liquids. J. Am. Chem. Soc. 1936, 58, 1486. 23 Thoss, M.; Evers, F. Perspective: Theory of Quantum Transport in Molecular Junctions. J. Chem. Phys. 2018, 148, 030901. 24 (a) webPCET Application Server, University of Illinois at Urbana-Champaign, 190 http://webpcet.scs.uiuc.edu (2009) (b) Hammes-Schiffer, S.; Soudackov, A. V. J. Phys. Chem. B 2008, 112, 14108-14123. 25 Warren, J. J.; Tronic, T. A.; Mayer, J. M. Thermochemistry of Proton-Coupled Electron Transfer Reagents and Its Implications. Chem. Rev. 2010, 110, 6961. 26 Garrido, M. Rosés, C. Ràfols, E. Bosch Acidity of Several Anilinium Derivatives in Pure Tetrahydrofuran. J. Soln. Chem. 2008, 37, 689. 27 Kaljurand, I.; Kütt, A.; Sooväli, L.; Rodima, T.; Mäemets, V.; Leito, I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019. 28 Haav, K.; Saame, J.; Kütt, A.; Leito, I. Eur. J. Org. Chem. 2012, 2167. 29 Pankratov, A. N.; Uchaeva, I. M.; Doronin, S. Y.; Chernova, R. K. J. Struct. Chem. 2001, 42, 739. 30 Liao, K.; Pack, B. W.; Toltl, N. P. J. Pharm. Biomed. Anal. 2007, 44, 118.