In the Spotlight: Bond Activation Reactions with Pyridyl−Triazolylidene Metal Complexes Von der Fakultät 3 Chemie der Universität Stuttgart zur Erlangung der Würde des Doktors der Naturwissenschaften (Dr. rer. nat.) genehmigte Abhandlung. Vorgelegt von Tobias Bens geboren in Bautzen Hauptberichter: Prof. Dr. Biprajit Sarkar Mitberichter: Prof. Dr. Michael R. Buchmeiser Mitberichter: Prof. Dr.-Ing. Elias Klemm Tag der mündlichen Prüfung: 08.09.2023 Institut für Anorganische Chemie Universität Stuttgart 2023 The doctoral studies presented herein were begun in October 2018 at the Institute of Chemistry and Biochemistry of the Freie Universität Berlin under the supervision of Prof. Dr. Biprajit Sarkar. In April 2019, research was conducted under the supervision of Prof. Dr Janez Košmrlj during a research visit to the Univerza v Ljubljaniin. The work was continued at the Institute for Inorganic Chemistry of the University of Stuttgart from October 2019 onward and concluded in May 2023. ”Sorrow is so easy to express and yet so hard to tell.” Joni Mitchell Dedicated to Julia I Declaration of Authorship I hereby confirm that I have prepared this dissertation without the help of any impermissible resources. All citations are marked as such. The thesis has not been accepted in any previous doctorate degree procedure. Tobias Bens Stuttgart, Mai 2024 The results and discussion part of this dissertation have been published as follows: 1) Chromium(0) and Molydenum(0) Complexes with a Pyridyl-Mesoionic Carbene Ligand: Structural, (Spectro)electrochemical, Photochemical, and Theoretical Investigations T. Bens, P. Boden, P. Di Martino-Fumo, J. Beerheus, U. Abold, S. Sobottka, N. I. Neuman, M. Gerhards, B. Sarkar, Inorg. Chem. 2020, 59, 20, 15504-15513. DOI: 10.1021/acs.inorgchem.0c02537. (Chapter 3.3) 2) Impact of Bidentate Pyridyl-Mesoionic Carbene Ligands: Structural, (Spectro)Electrochemical, Photophysical, and Theoretical Investigations on Ruthenium(II) Complexes T. Bens, J. A. Kübler, R. R. M. Walter, J. Beerhues, O. S. Wenger, B. Sarkar, ACS Org. Inorg. Au 2023, 3, 4, 184-198. (Chapter 3.2) DOI: 10.1021/acsorginorgau.3c00005. 3) The Best of Both Worlds: Combining the Power of MICs and WCAs to generate Stable and Crystalline CrI-tetracarbonyl Complexes with 𝜋 −Accepting Ligands T. Bens, R. R. M. Walter, J. Beerhues, M. Schmitt, I. Krossing, B. Sarkar, Chem. Eur. J. 2023, 29, e202301205. (Chapter 3.4) DOI: 10.1002/chem.202301205. II 4) A Different Perspective in Tuning the Photophysical and Photochemical Properties: The Influence of Constitutional Isomers in Group 6 Carbonyl Complexes with Pyridyl-Mesoionic Carbenes T. Bens, D. Marhöfer, P. Boden, S. T. Steiger, L. Suntrup, G. Niedner-Schatteburg, B. Sarkar, Inorg. Chem. 2023, 62, 16182-16195. (Chapter 3.5) DOI: 10.1021/acs.inorgchem.3c02478. 5) Investigations on the Influence of Two Pyridyl-Mesoionic Carbene Constitutional Isomers on the Electrochemical and Spectroelectrochemical Properties of Group 6 Metal Carbonyl Complexes T. Bens, B. Sarkar, Inorganics 2024, 12, 46. (Chapter 3.6) DOI: 10.3390/inorganics12020046. 6) Isolation, Characterization and Reactivity of Key Intermediates Relevant to Reductive (Electro)catalysis with Cp*Rh Complexes containing Pyridyl-MIC (MIC = Mesoionic Carbene) Ligands T. Bens, R. R. M. Walter, J. Beerhues, C. Lücke, J. Gabler, B. Sarkar, Chem. Eur. J. 2024, 30, e202302354. (Chapter 3.7) DOI: 10.1002/chem.202302354. https://www.mdpi.com/2304-6740/12/2/46 III Not included in this dissertation are the following publications: 1) Oxidative Access via Aqua Regia to an Electrophilic, Mesoionic Dicobaltoceniumyltriazolylidene Gold(III) Catalyst S. Vanicek, J. Beerhues, T. Bens, V. Levchenko, K. Wurst, B. Bildstein, M. Tilset, B. Sarkar, Organometallics 2019, 38, 22, 4383-4386. DOI: 10.1021/acs.organomet.9b00616. 2) NIR-Emissive Chromium(0), Molybdenum(0), and Tungsten(0) Complexes in the Solid State at Room Temperature P. Boden, P. Di Martino-Fumo, T. Bens, S. Steiger, U. Albold, Prof. Dr. Gereon Niedner- Schattenburg, Prof. Dr. M. Gerhards, Prof. Dr. B. Sarkar, Chem Eur. J. 2021, 27, 12959- 12964. DOI: 10.1002/chem.202102208. 3) Fluorinated click-derived tripodal ligands drive spin crossover in both iron(II) and cobalt(II) complexes M. Nößler, D. Hunger, N. I. Neuman, M. Reimann, F. Reichert, M. Winkler, J. Klein, T. Bens, L. Suntrup, S. Demeshko, J. Stubbe, M. Kaupp, J. van Slageren, B. Sarkar, Dalton Trans. 2022, 51, 10507–10517. DOI: 10.1039/D2DT01005D. 4) Mechanistic and Kinetic Investigations of ON/OFF (Photo)Switchable Binding of Carbon Monoxide by Chromium(0), Molybdenum(0) and Tungsten(0) Carbonyl Complexes with a Pyridyl-Mesoionic Carbene Ligand P. J. Boden, P. Di Martino-Fumo, T. Bens, S. T. Steiger, D. Marhöfer, G. Niedner- Schattenburg, B. Sarkar, Chem. Eur. J. 2022, e202201038. DOI: 10.1002/chem.202201038. 5) A crystalline cyclic (alkyl)(amino)carbene with a 1,1′-ferrocenylene backbone J. Volk, M. Heinz, M. Leibold, C. Bruhn, T. Bens, B. Sarkar, M. C. Holthausen, U. Siemeling, Chem. Commun. 2022, 58, 10396-10399. DOI: 10.1039/D2CC03871D. IV 6) Liquid crystalline self-assembly of azulene–thiophene hybrids and their applications as OFET materials F. Schulz, S. Takamaru, T. Bens, J. Hanna, B. Sarkar, S. Laschat, H. Iino, Phys. Chem. Chem. Phys. 2022, 24, 23481-23489. DOI: 10.1039/D2CP03527H. 7) Electrochemistry and Spin-Crossover Behaviour of Fluorinated Terpyridine-Based Co(II) and Fe(II) Complexes M. Nößler, R. Jäger, D. Hunger, M. Reimann, T. Bens, N. I. Neuman, A. Singha Hazari, M. Kaupp, J. van Slageren, B. Sarkar, Eur. J. Inorg. Chem. 2023, 26, e202300091. DOI: 10.1002/ejic.202300091. 8) Spin Crossover and Fluorine-Specific Interactions in Metal Complexes of Terpyridines with Polyfluorocarbon Tails M. Nößler, N. I. Neuman, L. Böser, R. Jäger, A. Singha Hazari, D. Hunger, Y. Pan, C. Lücke, T. Bens, J. van Slageren, B. Sarkar, Chem. Eur. J. 2023, 29, e202301246. DOI: 10.1002/chem.202301246. V Acknowledgments I want to express my sincere gratitude to Prof. Dr. Biprajit Sarkar for the opportunity to work on this thesis in his research group and for years of fruitful discussions. Apart from the possibility to learn various new techniques, I want to express some personal thoughts: The ability to maintain a good heart and sympathy is nothing one should take for granted. I am grateful to be able to work with a supervisor who has not lost his faith in humanity during difficult years and was always available for personal discussions whenever needed. Thank you for that Bipro! I am also grateful to Prof. Dr. Michael R. Buchmeiser for agreeing to be the 2nd reviewer for this thesis and I also want to thank Prof. Dr.-Ing. Elias Klemm for agreeing to be part of the committee. The cooperation partners that contributed to this thesis are kindly acknowledged: Prof. Dr. Markus Gerhards (University Kaiserslautern), Prof. Dr. Gereon Niedner-Schatteburg (University Kaiserslautern), Prof. Dr. Ingo Krossing (University Freiburg) and Prof. Dr. Oliver S. Wenger (University Basel). My sincere gratitude goes to all former and present members of the Sarkar group for filling the last years with joyful moments and unforgettable times inside and outside the lab. Special thanks go to all crystallographers Dr. Lisa Suntrup, Dr. Julia Beerhues, Dr. Uta Albold and Robert R. M. Walter for the single crystal XRD analysis and all the permanent staff at the Freie Universität Stuttgart and University of Stuttgart for performing measurements during the time. Especially, I want to thank Barbara Förtsch, Dr. Wolfgang Frey and Benjamin Rau. I would like to thank Annette Kling and Lale Ötztürk for all their administrative support and for brightening my days whenever we got the chance to chat together. It was always delightful to share some personal experiences and thoughts. I am sincerely grateful to Dr. Christopher Feil and Dr. Arijit Singha-Hazari for proofreading this thesis, the profound discussions we shared about life and the great times we spent during this period. I always appreciated your opinion and support, not only on a scientific but also on a personal basis. Special thanks to Dr. Julia Beerhues and Dr. Shuhadeep Chandra for mastering the movement from Berlin to our Swabian labs and all the fun times we shared together. I thank the whole Gudat group for the warm welcome and the fun wine tastings during that time. VI Furthermore, I want to thank Clemens Lücke and Robert R. M. Walter for starting the adventure in Stuttgart with me and for the endless memories I will keep with me. I really enjoyed the late-night discussions with some beers, our Trash-TV evenings and fun moments. I can ensure that having you as friends enriched my life in many aspects. Thank you for being always there for me! I would like to thank all my colleagues Fridolin Hennhöfer, Dr. Vasileios Filippou, Lasse Dettmann, Marie Leimkühler, Alok Mahata, Maren Neubrand, Maite Nößler, Manuel Pech, Tabea Pfister, Richard Rudolf, Felix Stein. You become more than just colleagues (now it is out!) and I really loved having such a great team with me. With you, everything seemed so much easier. Especially, I want to thank Fred for his emotional support and I hope you will take some time to take a deep breath and realize that everything will turn out fine, as we could prove it many times. Felix, I hope you will keep the sunshine in your heart that makes everyone happy. Special thanks to all my students Cindy Odenwald, Ingo Schneider, Ivan Shestov, Jonas Genz and Julia Gabler and all they have contributed to the success of my thesis. My deepest thanks go to Julia and the permanent support you gave me, all the beautiful moments and laughter we shared and it breaks my heart that we cannot share this moment together. There is not a single day passing by without thinking about you. You had the purest heart I have ever met and I will keep you always in mine. Finally, I want to thank my family for their continuous support all my life. Most importantly, they made me the person I am today to build the strength and devotion for this thesis, knowing that they have my back, whatever happens. It makes me proud to be part of this family. VII Abstract The CuAAC reaction to transform organic azides and terminal alkynes into 1,4-triazoles as precursors for triazolium salts has been known for more than one decade. It comes as no surprise that the synthetic scope of 1,2,3-triazoles has expanded rather fast. To obtain heterocyclic triazolium salts in good yields, selective synthetic strategies are required. They build the fundament of the so-called mesoionic carbenes (MICs), a sub-class of the well-established NHCs. The classification results from the fact that 1,2,3-triazol-5- ylidenes cannot be drawn without a charge separation in their Lewis structure. In the present work, new synthetic routes have been explored to access pyridyl-substituted 1,2,3-triazol-5-ylidenes (pyridyl-MIC) as bidentate ligands for transition metal complexes. The interplay between the strong 𝜎−donating nature of the MIC and the good 𝜋−accepting properties of the ligand plays a crucial role in the photo- and electro-catalytic activity. In chapter 3.2, a series of pyridyl−MIC (py−MIC) and 2,2'-bipyridine (bpy) containing Ru(II) complexes has been synthesized and characterized with techniques, such as 1H and 13C NMR spectroscopy, mass-spectrometry, elemental analysis and X-ray diffraction analysis. Moreover, (spectro)electrochemical measurements (SEC) were performed to explore the nature of different redox states with respect to the number of MIC moieties. Time-dependent density functional theory (TD-DFT) calculations were conducted to get insight into the photophysical properties of the complexes and their potential application as photocatalysts, while exited state lifetimes were investigated in cooperation with the group of Prof. Dr. Oliver Wenger. Chapter 3.3 deals with pyridyl−MIC group 6 (Cr, Mo) carbonyl complexes. CO is an ideal probe to investigate the influence of the ligand at the transition metal complex, not only in the native but also in the oxidized and reduced form. IR-, EPR- and UV/Vis/NIR-SEC combined with TD-DFT were performed to get in-depth understanding of the electronic structure of corresponding redox states, while excitation leads to an unusual reversible binding of the CO ligand after leaving the photoproducts in the dark. Surprisingly, the first stable [M(py−MIC)(CO)4]+ (M = Cr) fragment with 𝜋 −accepting ligands could be isolated and characterized via single-crystal X-ray diffraction analysis (chapter 3.4). EPR-, IR- and UV/VIS-spectroscopy supported by theoretical investigations were performed to shine light on the rare electron-deficient nature of the isolated [Cr(py−MIC)(CO)4]+ complexes. VIII Furthermore, in chapter 3.5 all [M(py−MIC)(CO)4] (M = Cr, Mo, W) complexes were investigated by step scan FTIR-spectroscopy and time-resolved spectroscopy in cooperation with the group of Prof. Dr. Gerhards and Prof. Dr. Niedner-Schatteburg. An unusual (photo)switchable ON/OFF binding of CO in solution was observed. Additionally, all complexes show a NIR-emission in the solid state. In chapter 3.6, the reduced species [M(py−MIC)(CO)4]− (M = Cr, Mo, W) were generated in situ and the reduction was assigned to be predominantly ligand-centered by various (spectro-)electrochemical and theoretical methods. Based on these results, electrochemical CO2 reduction was performed under non-protic conditions. (Spectro)electrochemistry can be a powerful tool to investigate reactive intermediates in electrocatalysis. In the last chapter 3.7, a series of [(py−MIC)Rh(Cp*)X]n+ (X = Cl−, MeCN; n = 1, 2) complexes was synthesized and fully characterized with several techniques. All complexes were tested in electrochemical H+ reduction and a mechanism on the precatalyst formation was proposed. A number of intermediates was chemically isolated, investigated via single-crystal X-ray diffraction analysis and compared to the electrochemically generated species. Theoretical calculations further supported the precatalytic activation pathway. The presented thesis provides an in-depth understanding of the impact of bidentate ligands with 𝜋 −accepting and strong 𝜎 −donating ligands (pyridyl-MIC) in transition metal complexes for potential applications in the field of photochemistry and electrocatalysis. For the first time, detailed (spectro)electrochemical investigations, combined with theoretical calculations, of two constitutional isomers were conducted to explore the influence of the electronic structures in bond activation reactions. The investigations of the highly reactive (spectro)electrochemically generated and chemically isolated key intermediates enable a profound understanding of the tailor-made design in new potential transition metal-based photo- and electrocatalysts. IX Kurzzusammenfassung Die CuAAC-Reaktion zur Umwandlung organischer Azide und terminaler Alkine in 1,4- Triazole, als Vorstufen für Triazoliumsalze, ist seit mehr als einem Jahrzehnt bekannt. Es überrascht nicht, dass sich die synthetischen Möglichkeiten für 1,2,3-Triazole rasant weiterentwickelt haben. Um heterocyclische Triazoliumsalze in guter Ausbeute zu erhalten, sind selektive Synthesestrategien erforderlich. Sie bilden die Grundlage der so genannten mesoionischen Carbene (MICs), einer Unterklasse der bekannten NHCs. Die oben genannte Klassifizierung ergibt sich aus der Tatsache, dass 1,2,3-Triazol-5-ylidene nicht ohne Ladungstrennung in ihrer Lewis-Struktur gezeichnet werden können. In der vorliegenden Arbeit wurden neue Synthesewege erforscht, um zu 1,2,3-Triazol-5- ylidenen mit Pyridyl-substituenten (Pyridyl−MIC) als zweizähnige Liganden für Übergangsmetallkomplexe zu gelangen. Das Zusammenspiel zwischen der stark 𝜎 −donierenden Natur des MIC und den guten 𝜋 −akzeptierenden Eigenschaften des Liganden spielt eine entscheidende Rolle für die photo- und elektrokatalytische Aktivität. In Kapitel 3.2 wurde eine Reihe von Pyridyl−MIC (py-MIC) und 2,2'-Bipyridin (bpy) enthaltenden Ru(II)-Komplexen synthetisiert und mit Techniken wie 1H- und 13C-NMR- Spektroskopie, Massenspektrometrie, Elementaranalyse und Röntgenbeugungsanalyse charakterisiert. Darüber hinaus wurden (spektro)elektrochemische Messungen (SEC) durchgeführt, um die Art der verschiedenen Redoxzustände in Bezug auf die Anzahl der MIC-Einheiten zu untersuchen. Zeitabhängige Dichtefunktionaltheorie (TD-DFT) wurden durchgeführt, um einen Einblick in die photophysikalischen Eigenschaften der Komplexe und ihre potenzielle Anwendung als Photokatalysatoren zu erhalten, während die Lebensdauern der Ausgangszustände in Zusammenarbeit mit der Gruppe von Prof. Dr. Oliver Wenger untersucht wurden. Kapitel 3.3 beschäftigt sich mit py-MIC basierten Gruppe 6 (Cr, Mo) Carbonylkomplexen. CO ist eine ideale Sonde, um den Einfluss des Liganden am Übergangsmetallkomplex nicht nur in der nativen, sondern auch in der oxidierten und reduzierten Form zu untersuchen. IR-, EPR- und UV/Vis/NIR-SEC in Kombination mit TD-DFT wurden durchgeführt, um ein tieferes Verständnis der elektronischen Struktur der Redoxzustände zu erhalten, während die Anregung der Komplexe zu einer ungewöhnlichen reversiblen Bindung des CO-Liganden führt, nachdem die Photoprodukte im Dunkeln gelassen wurden. Überraschenderweise konnte das erste stabile [M(py−MIC)(CO)4]+ (M = Cr) Fragment mit 𝜋 −akzeptierenden Liganden isoliert und mittels Röntgeneinkristallbeugungsanalyse X charakterisiert werden (Kapitel 3.4). EPR-, IR- und UV/VIS-Spektroskopie, unterstützt durch theoretische Untersuchungen, wurden durchgeführt um die seltene elektronenarme Natur der isolierten [Cr(py−MIC)(CO)4]+ Komplexe zu erschließen. Darüber hinaus wurden in Kapitel 3.5 alle [M(py−MIC)(CO)4] (M = Cr, Mo, W) Komplexe mittels Step-Scan-FTIR-Spektroskopie und zeitaufgelöste Spektroskopie in Zusammenarbeit mit der Gruppe von Prof. Dr. Gerhards und Prof. Dr. Niedner- Schatteburg untersucht. Es wurde eine ungewöhnliche (photo)schaltbare ON/OFF- Bindung von CO in Lösung beobachtet. Zusätzlich zeigten alle Komplexe eine NIR- Emission im festen Zustand. In Kapitel 3.6 wurden die reduzierte Spezies [M(py−MIC)(CO)4]− in-situ erzeugt und die Reduktion durch verschiedene (spektro-)elektrochemische und theoretische Methoden als überwiegend ligandenzentriert eingestuft. Auf der Grundlage dieser Ergebnisse wurde eine elektrochemische CO2-Reduktion unter nicht-protischen Bedingungen durchgeführt. Die (Spektro)elektrochemie stellt ein leistungsfähiges Instrument zur Untersuchung reaktiver Zwischenstufen in der Elektrokatalyse dar. Im letzten Kapitel 3.7 wurde eine Reihe von [(py−MIC)Rh(Cp*)X]n+ (X = Cl−, MeCN; n = 1, 2) Komplexen synthetisiert und mit verschiedenen Techniken vollständig charakterisiert. Alle Komplexe wurden in der elektrochemischen H+ Reduktion getestet, und es wurde ein Mechanismus für die Bildung des Präkatalysators postuliert. Eine Reihe von Zwischenprodukten wurde chemisch isoliert, mittels Röntgeneinkristallbeugung untersucht und mit den elektrochemisch erzeugten Spezies verglichen. Theoretische Berechnungen untermauerten den präkatalytischen Aktivierungspfad. Die vorliegende Arbeit vermittelt ein tiefgreifendes Verständnis der Auswirkungen von zweizähnigen Liganden mit 𝜋 −akzeptierenden und stark 𝜎 −donierenden Liganden (py−MIC) in Übergangsmetallkomplexen für potenzielle Anwendungen im Bereich der Photochemie und Elektrokatalyse. Zum ersten Mal wurden detaillierte (spektro)elektrochemische Untersuchungen, kombiniert mit theoretischen Berechnungen, von zwei konstitutionellen Isomeren durchgeführt, um den Einfluss der elektronischen Strukturen in Bindungsaktivierungsreaktionen zu erforschen. Die Untersuchungen der hochreaktiven (spektro)elektrochemisch erzeugten und chemisch isolierten Schlüsselintermediate ermöglicht ein tiefgreifendes Verständnis für das maßgeschneiderte Design neuer potenzieller übergangsmetallbasierter Photo- und Elektrokatalysatoren. XI Contents Declaration of Authorship I Acknowledgments V Abstract VII Kurzzusammenfassung IX Contents XI List of Abbreviations XIII 1 Introduction 1 1.1 Conceptual Design of Molecular Photocatalysts 3 1.2 Molecular Electrocatalysts 11 1.3 1,2,3-Triazolylidene-based Ligands 19 1.4 Photochemistry Inspired by Mesoionic Carbenes 25 1.5 Recent Developments in MIC-based Electrocatalysis 30 References 36 2 Scope of this Thesis 46 3 Results & Discussion 48 3.1 Summary and Conclusion 48 3.2 The Impact of Bidentate Pyridyl-Mesoionic Carbene Ligands: Structural, (Spectro)Electrochemical, Photophysical, and Theoretical Investigations on Ruthenium(II) Complexes 68 3.3 Chromium(0) and Molybdenum(0) Complexes with a Pyridyl-Mesoionic Carbene Ligand: Structural, (Spectro)electrochemical, Photochemical and Theoretical Investigations 84 3.4 The Best of Both Worlds: Combining the Power of MICs and WCAs to generate Stable and Crystalline CrI-tetracarbonyl Complexes with 𝝅 −Accepting Ligands 96 XII 3.5 A Different Perspective in Tuning the Photophysical and Photochemical Properties: The Influence of Constitutional Isomers in Group 6 Carbonyl Complexes with Pyridyl- Mesoionic Carbenes 106 3.6 Investigations on the Influence of Two Pyridyl-Mesoionic Carbene Constitutional Isomers on the Electrochemical and Spectroelectrochemical Properties of Group 6 Metal Carbonyl Complexes 122 3.7 Isolation, Characterization and Reactivity of Key Intermediates Relevant to Reductive (Electro)catalysis with Cp*Rh Complexes containing Pyridyl-MIC (MIC = Mesoionic Carbene) Ligands 142 XIII List of Abbreviations η overpotential aNHC abnormal N-heterocyclic carbene Bn benzyl bpy 2,2'-bipyridine btz 3,3-dimethyl-1,1-bis(p-tolyl)-4,4-bis(1,2,3-triazol-5-ylidene) tBu tert-butyl cat. catalyzed C-C pyridyl-4-triazolylidene C−N pyridyl-1-triazolylidene Cp* pentamethylcyclopendatienyl CuAAC copper(I)-catalyzed alkyne-azide cycloaddition CV cyclic voltammetry/cyclic voltammogram Cym p-cymene DCM dichlormethane DFT density functional theory Dipp 2,6-diisopropylphenyl DMF dimethylformamide EC electron-transfer/chemical reaction EECC electron-transfer/electron-transfer/chemical reaction/chemical reaction EPR electron paramagnetic resonance ESI electrospray ionization Et ethyl FcH/FcH+ ferrocene/ferrocenium couple FE Faradaic efficiency GC glassy carbon HEP Huynh's electronic parameter HOMO highest occupied molecular orbital ILCT intra-ligand charge transfer IR infrared ISC intersystem crossing IVCT intervalence charge transfer LC ligand-centered XIV LEP ligand electrochemical parameter LLCT ligand-to-ligand charge transfer LMCT ligand-to-metal charge transfer LUMO lowest occupied molecular orbital MeCN acetonitrile Me methyl Mes mesityl/2,4,6-trimethylphenyl MC metal-centered MIC mesoionic carbene MLCT metal-to-ligand charge transfer MO molecular orbital MS mass spectrometry NHC N-heterocyclic carbene NIR near-infrared nr non-radiative NMR nuclear magnetic resonance OTTLE optically transparent thin-layer electrochemical PCET proton-coupled electron transfer Ph phenyl ppy 2-phenylpyridine iPr isopropyl py pyridine SCE saturated calomel electrode SEC spectroelectrochemistry SET single-electron transfer TEP Tolman's electronic parameter TDDFT time-dependent density functional theory THF tetrahydrofuran TOF turnover frequency Tol tolyl TON turnover number tetr tetrazolate tpy 2,2':6',2''-terpyridine XV triaz 1,2,3-triazole UV ultraviolett vis visible VC vibrational cooling XVI 1 Introduction 1 1 Introduction "When the world is in trouble, chemistry comes to its rescue."[1] With these simple yet so powerful words, Carolyn Bertozzi reacted to the announcement of the prestigious Nobel Prize awarded to her. In the face of rapidly evolving climate change, her words seem more relevant than ever. The explosive growth of the world’s population, progressive industrialization and the scarcity of fossil fuels for societal prosperity have sparked a global debate on resource- efficient access to environmentally friendly and sustainable materials.[2] Despite the tremendous achievements in energy efficiency and the integration of industrial processes into the modern economy, the continued exploitation of fossil fuels results in serious environmental and health concerns.[3] In view of the global energy crisis looming, alternative energy sources such as wind, solar, nuclear and geothermal energy, biomass, dihydrogen and electrocatalytic refinery have been developed in recent decades.[4,5,6] However, according to recent data, 85% of the world’s primary energy depends on fossil fuels, which inevitably release fairly unreactive greenhouse gas, such as CO2 and methane,[5] giving rise to fundamental questions: What does it take to activate chemically inert bonds in small molecules to convert them into economically valuable products and what long-term strategies can be employed to store high energy material?[7,8] The answers appear simple from a chemical perspective – a catalytic transformation. A catalyst is a substance that stabilizes the transition state (𝛥𝐺𝑢𝑛𝑐𝑎𝑡. ‡ ) of a reaction by lowering the energy barrier (𝛥𝐺𝑐𝑎𝑡. ‡ ) between the starting material and the product without affecting the free enthalpy 𝛥𝐺 of the reaction (Figure 1). By definition, a catalyst is not consumed in the reaction and remains unaffected over across multiple transformations.[9] Figure 1. Energy profile of a one-step reaction (grey: uncatalyzed, blue: catalyzed). 1 Introduction 2 Nature undoubtedly contains the most efficient catalysts for the activation of abundant small molecules such as O2, N2, H2, CO2, NOx, and CH4.[10,11,12] A variety of organisms make use of complex enzymes in metabolic processes, such as those found in respiratory processes[13] and photosynthesis,[10,14] overcoming the embedded kinetic and thermodynamic barriers in the multi-step catalytic conversion of chemically inert molecules. In a broader context, the complexity of the enzymes and the simplicity of the target molecules, have parallels to the battle between David and Goliath. Only this time Goliath retains the upper hand. The key role of this outcome, apart from the global structure of the enzymes, can be addressed to the redox-active central metal ion. Indeed, from a (bio-)inorganic perspective, the catalytic activation of small molecules with transition metals forms the foundation of several fields of chemical science, such as heterogeneous catalysis,[15] homogeneous catalysis,[6,16,17] photocatalysis[18,19] and electrocatalysis (Figure 2).[20,21,22,23] Figure 2. Overview of catalytic strategies for small molecule activation. Plenty of industrial manufacturing processes are based on heterogeneous catalyst due to their high stability, easy product separation and excellent recovery. However, heterogeneous catalysts often display poor product selectivity, limited mass transfer, and an undefined catalytic surface precluding from detailed mechanistic studies. On the other hand, from a purely industrial viewpoint, homogeneous catalysts are less attractive considering the difficult product separation, the large amount of waste material, and the corrosion of the reactors. But what makes homogeneous catalysis so unique? 1 Introduction 3 For some, homogeneous catalysis may be a purely academic challenge, but for others, it offers answers to the fundamental mechanistic understanding of small molecule activation in a well-defined molecular environment under relatively mild conditions, as observed in biological systems, such as enzymes. Understanding these complex mechanisms provides access to tailor-made molecular catalysts with extremely high efficiency and selectivity, requiring only low catalyst loadings.[24] The extreme sensitivity of the catalytically active species make them difficult to handle and an appropriate choice of a stable precursor therefore is indispensable. In a classical approach, a homogeneous catalyst requires sacrificial additives to generate the highly reactive intermediate that converts the chemical synthons into valuable products, such as alcohols, formic acid, ammonia, dihydrogen, oxygen or carbon monoxide, to name a few.[5,6,8,12,17] In the last few decades, enormous progress in the field of photo- and electrochemically induced activation of the precatalysts have been achieved to provide an atom economic access to the catalytically active species for the storage of high energy materials and the activation of small molecules.[22,23,25–27] In view of the scope of this thesis and the vastness of the fields, the following discussion will focus on the conceptual design of the homogeneous molecular catalyst and its photo- and electrochemical precatalytic activation concerning small molecules, limited dominantly on bidentate heterocycles that contain N-donors. 1.1 Conceptual Design of Molecular Photocatalysts Photochemistry is often associated in a vernacular sense, with photophysics in an interdisciplinary field that encompasses new synthetic strategies for the design of efficient molecular catalysts. In addition, it is usually accompanied by an in-depth understanding of the physical properties of the applied photocatalyst. The inclusion of rapidly advancing theoretical approaches provides a conceptual understanding of the interplay between the central metal ion and the ligand scaffold, leading to a tremendous development in the field of photocatalysis[18,25,26,28,29,30,31] and dye-sensitized solar cells.[32] In photochemistry, ground-breaking discoveries have been made in the field of supramolecular chemistry,[33] early[34–36] and late[28,29,36] transition metal complexes, due to the endless number of ligand frameworks, such as tripodal ligands[34–40] or cyclic ligands,[41] to name a few. 1 Introduction 4 For the scope of this thesis, the following discussions will briefly focus on the octahedral group 8 transition metal complexes ([M(bpy)3]2+ with M = Ru, Fe) containing bidentate polypyridine ligands. Photocatalysts are multi-electron systems and the resulting electronic wave functions (molecular orbital = MO) can be visualized in a Jablonski diagram according to their predominant atomic orbital contributions (Figure 3). Figure 3. Molecular orbital diagram for an octahedral transition metal complex (arrows indicating electronic transitions). Strongly bonding, predominantly ligand-centered orbitals are represented as 𝜎𝐿 for 𝜎 −bonding and 𝜋𝐿 for 𝜋 −bonding orbitals, while essentially non-bonding metal- centered orbitals of 𝑡2𝑔 symmetry are depicted as 𝜋𝑀 orbitals. The anti-bonding, predominantly metal-centered orbitals of 𝑒𝑔 ∗ symmetry are classified as 𝜎𝑀 ∗ orbitals and the anti-bonding ligand-based orbitals as 𝜋𝐿 ∗ orbitals.[42] At relatively low ground state energies, electronic transitions of the type metal-centered (MC), ligand-to-metal charge transfer (LMCT), metal-to-ligand charge transfer (MLCT) and ligand-centered (LC) are expected. Although MC transitions are forbidden in octahedral complexes with inversion symmetry (Laporte’s rule),[43] lowering the dynamic symmetry induces partially allowed transitions. Octahedral complexes with a d6 electron configuration typically have fully occupied 𝜎𝐿 and 𝜋𝐿 orbitals, resulting in a closed-shell ground state configuration 𝐴1𝑔 1 . Since the electronic transitions from the ground state to the excited state occur instantaneously compared to their nuclear motions, the geometry of a metal complex does not change within the time scale of the electronic transition, according to the Franck- Condon principle.[42] 1 Introduction 5 However, the spatial change in electron density induces a nuclear motion leading to a new minimum geometry of the excited complex. The changes in the bond distances, angles and torsion angles are best described as a combination of normal modes (vibrational cooling, VC). Their intensity is proportional to the square of the overlap integral between the vibrational wave functions of the transitions involved (absorption and emission).[43] To a first approximation, the electronic transitions and their vibronic modes represented by the potential energy surfaces (anharmonic oscillator) depends to a large extent on the nature of the metal center, the coordinated ligands and the overall symmetry of the molecule, defining the nature of the lowest excited state and multiplicity.[34,40] These criteria form the basis for an efficient photocatalyst facilitating sufficient lifetime of the excited state. From a purely statistical perspective, an increased lifetime leads to a higher probability of efficient energy transfer or chemical reaction with the targeted molecule – but what factors limit the excited-state lifetime? One of the well-established, yet so simple photocatalysts are the polypyridine complexes of group 8 [M(bpy)3]2+ (M = Ru, Fe) metals (Figure 4).[34,40,42] The higher homologue [Ru(bpy)3]2+ shows an excited-state lifetime of 890 ns at room temperature, while [Fe(bpy)3]2+ has only an excited-state lifetime of only 50 fs.[34,40] The drastic decrease in excited-state lifetime can be attributed to the central metal atom. After excitation from the 𝐴1𝑔 1 ground state to the excited 1MLCT state and subsequent vibrational cooling, a rapid intersystem crossing (kISC) in the 3MLCT state (nesting state) occurs in both complexes. However, the weaker ligand-field splitting of [Fe(bpy)3]2+, as a result of the smaller radial distribution of 3d electrons on the metal core,[44] allows access to energetically low-lying MC states, decreasing the energetic barrier (E) for the internal conversion and initiating a non-radiative decay (knr) to the ground state. The shift of the MC states can be attributed to the population of the anti-bonding 𝑒𝑔 ∗ orbitals, resulting in the elongation of the metal-ligand bond. Therefore, to obtain long- lived emissive MLCT states, the non-emissive MC states must be shifted to higher energies. In this context, different conceptual strategies for ligand design have been explored to increase the thermal energy barrier between the MLCT and MC states. 1 Introduction 6 Figure 4. Schematic potential energy surface diagrams of [M(bpy)3]2+ (M = Ru, Fe) and their electronic states (excitation (h): black, phosphorescence (kp): red, intersystem crossing (kISC): blue, internal conversion (kIC): dark grey, non-radiative decay (knr): light grey).[34] [Fe(bpy)3]2+ [Ru(bpy)3]2+ 1 Introduction 7 Increasing the activation barrier between MLCT states and the MC states can be achieved by the inclusion of highly symmetric ligands, push-pull systems, highly strained complexes, strongly donating ligands or combined 𝜎 −donor and 𝜋 −acceptor ligands (for detailed discussion see Section 1.4).[40] In a more general perspective, the effect on ligand-field splitting is determined by the metal-ligand interaction according to ligand-field theory: • 𝜎 −donor: overlap of the filled 𝜎 orbitals of the ligand with the metal-centered d−orbitals of 𝜎 −symmetry (𝑑𝑧2 , 𝑑𝑥2−𝑦2) • 𝜋 −donor: overlap of the filled 𝜋∗ orbitals of the ligand with the metal-centered d−orbitals of 𝜋 −symmetry (𝑑𝑥𝑦, 𝑑𝑥𝑧, 𝑑𝑦𝑧) • 𝜋 −acceptor: overlap of an unoccupied 𝜋∗ orbitals of the ligand with the metal- centered d−orbitals of 𝜋 −symmetry (𝑑𝑥𝑦, 𝑑𝑥𝑧, 𝑑𝑦𝑧), so-called 𝜋 −backbonding A weak 𝜎 −donor ligand leads to a weak ligand-field splitting between the non-bonding 𝑡2𝑔 orbitals and the anti-bonding 𝑒𝑔 ∗ orbitals, while a strong 𝜎 −donor ligand destabilizes the anti-bonding 𝑒𝑔 ∗ orbitals or in other words – the 3/5MC states (Figure 5). On the other hand, strong 𝜋 −acceptor ligands decrease the energy of the 𝑡2𝑔 orbitals and consequently the thermal barrier to internal conversion. A different situation is observed in the presence of a strong 𝜋 −donor ligand. With a right choice of ligand, destabilization of the 𝜋 orbitals can lead to a so-called HOMO inversion between the metal-centered and the ligand-centered orbitals and is currently state of the art.[45] However, to achieve optimal ligand-field splitting, an ideal octahedral geometry of the N−M−N trans angles (= 180°) is required to maximize the overlap of the metal-ligand orbitals. Otherwise, a lowering of the symmetry in the metal center results in a degeneration of the participating orbitals and, consequently, to a decrease of the exited- state lifetime.[40] Formally, the excitation in [M(bpy)3]2+ (M = Ru, Fe) complexes can be described as an oxidized metal center and a reduced ligand radical [MIII(bpy)2(bpy∙−)]* in the 3MLCT state.[44,46] The light-induced charge separation and the extended lifetime of the excited- state are essential for the activity of the photocatalyst in electron and energy transfer processes. 1 Introduction 8 Figure 5. Ligand field effects between metal d−orbitals and ligand orbitals.[40] The energy level of the generated electron-hole at the metal center and the excited ligand- centered electron can be tuned by the synergy between the metal center and the electronic properties of the incorporated ligand to provide an efficient photocatalyst for tailor-made applications. To determine, whether the applied photocatalyst is a suitable candidate for chemical and energy transfer processes, Latimer diagrams have proven to be particularly useful.[31,47] The combination of the oxidative and reductive potentials of the ground state with the excited state energy (E00) is frequently used to estimate the oxidative and reductive potentials of the excited state of the photocatalyst (Scheme 1). Scheme 1. Simplified Latimer-diagram for electron/energy transfer processes. 1 Introduction 9 The ambivalence of the excited state, as a strong reductant and at the same time as a strong oxidant, leads to two different quenching pathways. One is the so-called reductive quenching cycle. After excitation, the photocatalyst [M]* can undergo an one-electron reduction in a bimolecular reaction with the substrate (D) to generate the singly-reduced species [M]− and D∙+, while in the oxidative quenching cycle, the excited state transfers an electron to the substrate (A) of interest, forming the singly-oxidized species [M]+ and A∙−. It is important to mention, that the electron transfer rate competes with the radiative emission and vibrational relaxation (dynamic quenching) of the excited state and the interaction of the substrate with the photocatalyst (static quenching).[31] A sufficiently long-lived resting state is therefore crucial for an excellent catalytic performance. So far, only non-catalytic conditions have been presented. In catalytic conditions, the native photocatalyst must be recovered from its oxidized or reduced form. In photocatalysis, sacrificial donors (or acceptors), such as amines,[31] are usually added to regenerate the catalyst. Once the catalytic conditions are fulfilled, the excited photocatalyst can either directly undergo a single-electron transfer (SET) to the substrate, to initiate a chemical reaction or act as a photosensitizer for a co-catalyst, to perform the redox-reaction (Scheme 2). Scheme 2. Possible photocatalytic SET reactions in an oxidative quenching cycle (D = sacrificial donor). The thermodynamic ability of the excited state to intervene in a bimolecular energy- transfer process follows the Marcus-theory[48] and the Franck-Condon factor.[49] 1 Introduction 10 The non-radiative energy transfer can be described classically as combined effects of energy gradient and nuclear reorganization or quantum mechanically, as the thermally averaged sum of vibrational overlap integrals between the donor and acceptor molecule. They are divided into two mechanisms - the Coulomb mechanism, and the exchange mechanism (Scheme 3).[42] However, from a general perspective, the mechanism depends on the spin of the ground state, the excited state and the donor-acceptor distance. Scheme 3. Energy transfer mechanism (top: coulomb mechanism, bottom: exchange mechanism).[42] The Coulomb (trough space) mechanism requires physical contact between the donor and acceptor molecules and a large dipole-dipole interaction. This mechanism is usually observed in singlet-singlet energy transfers with large aromatic systems. On the other hand, the exchange mechanism strongly depends on the orbital overlap between the donor and acceptor molecules, typically mediated by a bridging fragment (trough bond). The energy transfer rate therefore increases with decreasing distance. Notably, the exchange mechanism allows spin-forbidden transitions to obey spin conservation, making a photocatalytic application particularly attractive. Ideal photocatalysts show an absorption maximum in the visible spectrum (𝜆𝑚𝑎𝑥 = 390 − 700 nm) to perform the catalytic transformation under mild conditions while avoiding substrate and product decomposition.[50] 1 Introduction 11 They are of particular interest for the activation of small molecules, such as organohalides,[51] non-halogenated substrates (C−C bond cleavage,[52] C−H functionalization,[53] oxidative cyclization,[54,55] alkene reduction,[56] cyclcoaddition,[57] trifluormethylation[55,58]), in dehydrogenation reactions,[59,60] CO2 reduction,[61,62,63] polymerization[64] and lignin degradation,[65] providing a wide range of products for synthetic transformations and sustainable energy conversion (Figure 6). Figure 6. Photocatalytic small molecule activation.[31] 1.2 Molecular Electrocatalysts The electrochemical conversion of low energy feedstocks, such as H2O, H+, and CO2, has led to the discovery of an ever-growing number of homogeneous transition metal complexes to lower the kinetic barriers and to improve the selectivity for the generation of renewable energy sources.[66] However, non-trivial multi-electron conversion with high-energy intermediates allows for multiple reaction pathways and electrocatalyst degradation.[23,67,68] Therefore, the identification of the catalytically active species is of fundamental interest to optimize the stability and activity of the electrocatalyst. In general, the homogeneous precatalyst (P) functions as electron shuttle between the electrode surface and the substrate (A) to induce the chemical reaction generating product B. The catalytically active species (Q) performs the substrate conversion in the diffusion layer of the electrochemical cell and requires lower energies (𝐸𝑐𝑎𝑡/2 0 ) compared to direct product formation at the electrode surface (𝐸𝑜𝑛𝑠𝑒𝑡) (Scheme 4). 1 Introduction 12 The catalytically active species is therefore of fundamental interest to optimize the stability and activity of the electrocatalyst. However, the performance of the activated electrocatalyst is determined by the thermodynamic potential (𝐸𝐴/𝐵 0 ) of the product formation (A/B). In the case of 𝐸𝑐𝑎𝑡/2 0 < 𝐸𝐴/𝐵 0 , the chemical transformation from A→B is usually limited by the insufficient driving force of the catalyst. Hence, most of the reported electrocatalysts require additional energy, better known as overpotential (𝜂).[23] Scheme 4. Suitable range for a homogeneous electrocatalyst (top) and simplified electron transfer at the diffusion layer (bottom left) with a schematic inner- and outer- sphere mechanism (bottom right). In an outer-sphere mechanism, the electrocatalyst acts as an electron-transfer reagent, that induces product formation, whereas in an inner-sphere mechanism, the electrocatalyst binds the substrate to produce an intermediate, which in turn undergoes a redox process to generate the desired product. But how can one assure that the precatalyst used is a suitable candidate under the electrocatalytic conditions? One of the most popular techniques for evaluating the precatalyst is cyclic voltammetry. It not only provides information on the redox potentials to determine whether a catalyst can be used in an appropriate regime, but also in-depth kinetic and mechanistic details of the molecular (pre-)catalyst and its activity.[23,68,69] 1 Introduction 13 However, the identification of the active species at the electrode surface is still relatively unexplored but is crucial for understanding the catalytic performance. The extreme potentials required for electrochemical transformation can lead to the decomposition or demetallation of the complex, resulting in catalytically active nanoparticles or heterogeneous species, that are adsorbed on the electrode surface.[68] Given scope of this thesis, the following brief discussion will focus mainly on the precatalytic activation of the electrocatalyst and how the performance of the active species can be influenced by the metal center and the ligand framework. In the absence of a substrate, molecular electrocatalysts exhibit individual redox processes, which, depending on the nature of the electrocatalyst, are either electrochemically reversible, quasi-reversible or irreversible. For the sake of simplicity, only electrochemically reversible processes in the presence of the substrate will be discussed here. The first mathematical description of a one-electron, one-substrate reaction was described by Savéant and Su, who figuratively classified the limiting wave forms into eight kinetic zones.[70] According to their classification, catalysts under an ideal condition show a S-shaped (K or KS zone) response in the presence of the substrate (Figure 7, I). However, the shape strongly depends on the scan rate, electron transfer rate and individual substrate and catalyst concentrations in addition to competing with certain side phenomena such as substrate consumption, product inhibition and catalyst deactivation.[71] Figure 7. Qualitative cyclic voltammograms of catalyst transformation or decomposition adapted from Dempsey and co-workers.[68] 1 Introduction 14 The complexity of catalytic systems has shown that direct translatability of kinetic zones is difficult to achieve due to multiple elementary steps and intermediates, that can undergo several reaction pathways. Some of the most common deviations from ideal S-shaped catalysis wave forms are shown in Figure 7. In the one-electron, one-substrate catalytic reaction (Figure 7, II), a diffusion peak is observed after reversing the scan direction of the catalytic process. In this case, the redox event originates from a new homogeneous species, formed during the catalysis, as a consequence of a slow reaction rate constant or substrate consumption.[72] Additionally, the diffusion peak sometimes also indicate oxidation or reduction of the catalytic product, a decomposition product or even a feature of the active catalyst. Identifying the diffusional process can be challenging but can provide useful information about possible decomposition pathways, product formation or even important intermediates in the catalytic cycle. The in-situ formation of the active catalyst is often accompanied by an induction period, manifested by the catalytic activity as a function of time. As the cyclic voltammogram (Figure 7, III) progresses, a higher concentration of the catalytically active species is produced. The accumulation of the active catalyst results in a continuous increase of the current, during reverse scan, causing a so-called "crossing curve". However, this phenomenon should not always be interpreted as the slow formation of the active species or as electrochemically induced autocatalysis,[73] but rather as possible electrodeposition at the electrode surface.[74] Perhaps the most clear indication of precatalyst transformation is the appearance of an irreversible prewave (Figure 7, IV).[75–77] An electron transfer reaction can induce a chemical reaction leading to a catalytic intermediate that does not carry out the catalytic reaction immediately. In this case, the potential shift of the prewave 𝐸1/2 strongly depends on the kinetics of the precatalyst formation. An anodic shift of the potential indicates a rapid formation of the precatalyst, while a cathodic shift is associated with a slow formation of the catalytic intermediate (Figure 8). 1 Introduction 15 Figure 8. Kinetic influence of precatalytic formation in presence of the substrate.[68] The investigation of the precatalytic wave plays a crucial role in understanding of the catalytic performance, considering the presence of competing chemical decomposition leading to the formation of heterogeneous intermediates. During the past decades, various techniques have been developed to confirm the presence of a homogeneous (pre-)catalyst, such as mercury poisoning,[78,79] electrode surface analysis,[77,79–81] rotating ring-disc electrochemistry,[76,82] rinse tests[75,76,78,81,83] and spectroelectrochemistry.[84] Each of the above-mentioned methods provide detailed insights into the nature of the electrochemically generated species. Particularly both rinse tests and spectroelectrochemistry are fruitful due to their simplicity and spectroelectrochemistry due to the scope to combine electrochemistry with classical spectroscopic techniques, such as UV/vis/NIR-, EPR- and IR-spectroscopy.[85,86] The rinse test represents one of the most common methods for detecting a heterogeneous or heterogenized active catalyst. In this process, the electrode is rinsed and transferred into a freshly prepared substrate-only solution after stopping the cyclic voltammogram passing the precatalytic wave or the catalytic potential. The absence of any significant changes in the current beyond the background clearly exclude the formation of any heterogeneous particles or adsorption on the electrode surface. However, correct handling of the rinse test determines whether a heterogeneous species can be detected or not. The observation of an additional current depends strongly on the metastable film, that forms on the electrode surface. Without the application of an appropriate potential during the rinse test to prevent the diffusion of decomposition of the molecular species at the electrode surface, the assignment of an electrodeposited catalyst could lead to a misinterpretation of the results.[81] 1 Introduction 16 An excellent method to follow in-situ formation of the precatalyst, without transferring the electrode is spectroelectrochemistry (SEC). Real-time detection in the absence or presence of the substrate provides detailed information about the stability and electronic structure of the electrochemically generated precatalyst.[86] However, from a purely electrochemical point of view the short-lived intermediates can undergo rapid follow-up reactions, even though they follow the criteria of a reversible electron transfer process in cyclic voltammetry (peak/peak current ratio = 1, peak/peak voltage difference = 59 mV for one-electron transfer at 298 K),[87] In terms of spectroelectrochemistry, the time scale of bulk electrolysis competes with the rapid chemical transformation of a highly reactive intermediate generated during the fast sweep rate in cyclic voltammetry. To shorten the bulk electrolysis time, narrow glass tubes (EPR-SEC) or optically transparent thin-layer electrodes (OTTLE) cells (UV/vis/NIR- or IR-SEC) are commonly used to increase the concentration of the electrochemically generated species at the electrode surface and consequently make detection of the electrochemically reduced or oxidized species more accessible.[86] The stability of the electrochemically generated species is expressed by isosbestic points during controlled potential electrolysis and the complete recovery of the initial spectra after spectroelectrochemical measurements.[86] Any deviation from these criteria indicates electrochemically induced chemical transformation of the precatalyst. The information that can be obtained from precatalytic wave formation paves the way for the optimal design of a highly active and stable catalyst. An efficient electrocatalyst should be capable of performing multi-electron substrate conversion at low overpotential to access high energy materials, such as H2, CO or HCOOH, to name a few.[88] From a mechanistic point of view, the molecular catalyst can perform electron transfer steps directly at the metal center, in a combined metal-ligand pathway or exclusively at the ligand framework. This structure-reactivity relationship between the metal center and the ligand defines the catalytic performance of the electrocatalyst and is of great importance for modifying the catalytic activity – but nothing comes without a price. The incorporation of an electron-withdrawing ligand is expected to lower the overpotential, by reducing the electron density at the active site of the metal center. This makes the electron uptake more accessible at mild potentials. 1 Introduction 17 In contrast, electron-donating ligands increase the electron density at the electrocatalyst and thereby shift the electron uptake to higher potentials. Activation of the precatalyst often follows subsequent substrate binding according to an inner-sphere mechanism. However, the initial step is strongly influenced by the electronic structure of the catalyst. Electron-withdrawing ligands, for instance, reduce the substrate affinity and thus the overall activity of the catalyst, while electron-donating ligands favor substrate binding. This contradiction between the requirement of a low overpotential and strong substrate activation is one of the major challenges in electrocatalysis. Savéant et al. described this paradox as the "iron law" of electrocatalysis.[89] However, the modular synthesis of homogeneous electrocatalysts allows a tailor-made design for selective and highly efficient conversion of the substrate. The nature of the active metal center plays probably the most obvious influence on the stability and activity of the electrocatalyst. Comparison of 3d metal centers with 4d and 5d metal centers reveals some interesting trends. First, 3d metals possess smaller atomic radii and consequently stronger Coulomb repulsion between the metal centers and the ligand frameworks, while 4d and 5d metals have more diffuse orbitals leading to better overlap (primogenic repulsion).[44] As a result, ligand dissociation is more likely in 3d metal complexes and usually (but not always)[90] higher overpotentials are observed due to the higher reorganization energies during electron transfer. The high geometric changes in 3 d metal complexes also reduce the electron transfer rate during the redox event.[88] However, this should be taken with great caution, as the relative energies of the d orbitals involved (3d < 4/5d) with respect to the incorporated ligand have a drastic effect on the activation of the substrate, reaction rate and activation of the precatalyst.[90] Second, changing the oxidation state can affect product selectivity and overpotential. Higher oxidation states usually decrease the overpotential of the electrocatalyst as the overall charge of the complex decreases. In 2015, Robert et al. showed that in a similar ligand framework the active metal center has a drastic impact on the product selectivity in electrochemical CO2 reduction. The Fe(II) complexes, with an 𝜂1 CO2 binding mode selectively produced formic acid as a product, while in the corresponding Co(III) complex CO2 was converted to CO. The product selectivity can be explained by the weaker 𝜋 −backbonding in the Fe(II) complex leading to a weaker metal-carbon bond, which facilitates the isomerization from Fe−COOH to Fe−OCHO.[61] 1 Introduction 18 However, the catalytic activity of the central metal atom is strongly influenced by the electronic and steric properties of the ligand framework. The influence of the ligand system can be divided into two categories: the inner coordination sphere and the secondary coordination sphere (Figure 9). The inner-sphere primarily determines the electronic properties of the complex, as the ligands are directly bound to the active site, whereas in the outer-sphere the steric repulsions, weak bonding interactions and cooperative effects with the metal-bound substrate are decisive. An extended 𝜋 −system lowers the HOMO−LUMO energy gap and facilitates electron uptake, leading to lower overpotentials, while strongly electron-donating ligands increase the activity of the metal center, as mentioned earlier. Figure 9. Schematic representation of a modular electrocatalyst. The incorporation of a labile co-ligand, such as halides or solvent molecules, provides access to a vacant coordination site at the metal center for substrate binding. Contrary to halides, neutral ligands are easier to dissociate from the metal center under reductive conditions. As a result, electrocatalysts with neutral ligands have lower overpotentials compared to their halide-containing counterparts.[90] A vacant coordination site can lead to the dimerization of the catalytic intermediate and consequently to a decrease in the catalytic rate. To prevent catalyst deactivation, sterically demanding substituents can be incorporated into the secondary-sphere of the ligand.[91] In addition, the repulsive interaction of the ligand at the metal center can favor ligand dissociation of the labile co-ligand, lowering the overpotential for substrate binding and even altering the mechanism of electrocatalysis.[92] 1 Introduction 19 Secondary-sphere effects have been shown to be effective in enhancing catalytic activity. Substitution of ligands with charged substituents or perfluorination of ligands lowers the overpotential of the catalyst, but the electronic changes induced by ligand substitution follows the "iron law".[89] Another attempt is the inclusion of hydrogen bonding donor substituents. The hydrogen substrate interaction not only affects the stability and solubility of the complex, but also lowers the kinetic barrier for the substrate conversion and product selectivity.[89,90,93] However, a strong ligand-substrate interaction in the secondary-sphere can lead to a decrease in the reaction rate, again highlighting the importance of a well-balanced interplay between the electrocatalyst and the substrate.[89] 1.3 1,2,3-Triazolylidene-based Ligands A common motif for complexes based on transition metals are N−donor ligands, such as pyridine and bipyridine (I, Figure 10). The ligands exhibit both good 𝜎 −donor and 𝜋 −acceptor properties, making them one of the most commonly used moieties for photocatalytic[34,39,40,44,59] and electrocatalytic applications.[20,23,91,94] Figure 10. Relative donor/acceptor properties of the ligand classes utilized in this thesis. Changing the type of the heterocycles that contain N-donors leads to new properties and has received growing interest in the field of organometallic chemistry. In particular, 1,2,3- triazoles have proven to be versatile ligands due to their modular synthesis and applicability in 'Click' chemistry (II, Figure 10).[37,39,95] The five-membered heterocycle has a lower energy HOMO compared to pyridine, while the energy of the LUMO is higher, making it a weaker 𝜎 −donor and 𝜋 −acceptor, respectively.[95] 1 Introduction 20 1,2,3-Triazoles can be obtained by cycloaddition reactions of terminal alkynes and organic azides. However, thermally induced 1,3-dipolar cycloaddition (Huisgen cycloaddition) yields a mixture of the 1,4- and 1,5-regioisomers due to their similar energy profiles.[95] In 2002, Sharpless[96] and Meldal[97] independently discovered the so-called copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC), which produces exclusively the 1,4-regioisomer. The reaction proceeds under mild reaction conditions and has a very high functional group tolerance. Only three years later, Sharpless and co-workers again pioneered the development of synthetic protocol for the selective synthesis of the 1,5- regioisomer by a ruthenium-catalyzed cycloaddition.[98] Within the same decade, an alternative synthetic protocol for the 1,5-regioisomer - the metal-free approach – was discovered, paving the way for a synthesis of a wide variety of ligand systems (Scheme 5).[99] Scheme 5. Thermal and catalytic synthesis of 1,2,3-triazoles. 1,2,3-Triazoles are the most common precursor for the generation of 1,2,3-triazolylidenes (III, Figure 10), a subclass of N−heterocyclic carbenes (NHC). The term mesoionic carbene (MIC) or abnormal N−heterocyclic carbene (aNHC) was first coined by Crabtree and co-workers, who described that MICs cannot be formulated without a charge separation in their Lewis structure (Scheme 6).[37,39,100] The ligand class exhibit great 𝜎 −donor properties and moderate but tunable 𝜋 −acceptor capacities,[101] making them suitable candidates for wide-ranging applications, such as photochemistry,[34,37,40,102] electrocatalysis[103–107] and small molecule activation.[37] MICs of the 1,2,3-triazolylidene type are usually accessed by alkylation[108,109] or arylation[110] of the 1,2,3-triazole at the N3 position to generate the corresponding triazolium salt in nearly quantitative yields (IV, Scheme 6). 1 Introduction 21 Scheme 6. Synthetic strategies to obtain MICs and possible resonance structures. Direct deprotonation of the triazolium salt to obtain the free MIC is often accompanied by a methyl-shift from the N3 to the C5 atom.[108] In contrast, arylated MICs exhibit higher stability and can be stored at low temperatures under inert conditions. A common strategy to avoid decomposition of the free MIC is the addition of a metal precursor (A, Scheme 6) after deprotonation of the triazolium salt for the in-situ formation of the desired transition metal complex. Alternatively, the well-established silver transmetalation route (B, Scheme 6) can be used. The in-situ generated silver-MIC complex can be further reacted with a metal precursor of choice under mild reaction conditions.[111] The incorporation of MIC ligands into transition metal complexes is certainly interesting from many aspects, but the investigation of electronic properties is probably the most informative. Without the fundamental understanding of the electronic properties, structure-reactivity prediction for various (catalytic) applications would be challenging or even impossible. Undoubtedly, Tolman's electronic parameter (TEP) is one of the milestones to probe the net donor strength of various ligands.[112] The concept of nickel phosphine complexes [Ni(PR3)CO)3] was adapted by Gusev, by introducing the [Ni(NHC)CO)3] analogues.[113] 1 Introduction 22 In 2003, the group of Crabtree reported a less-toxic alternative of the type [M(L)(CO)2Br] (M = Rh, Ir) to convert the TEP to NHC-based systems using the average CO stretching frequencies as a probe for overall donor strength (Figure 11).[114] Although all NHC ligands are considered fairly strong 𝜎 −donors, their MIC counterparts display even higher 𝜎 −donor properties, only surpassed by the mesoionic imidazolylidene ligands.[115] In recent years, spectroscopic methods have emerged as a tool to differentiate between 𝜎 −donor and 𝜋 −acceptor contributions of moieties. Huynh's group introduced a physical parameter named as Huynh's electronic parameter (HEP) for the determination of the 𝜎 −donor strength of the bound carbene in square planar palladium or linear gold NHC- and MIC-complexes (Figure 11). The negligible 𝜋 −backbonding in late transition metals facilitates the determination of the 𝜎 −donor ability of the ligand. As a result, the chemical shift of the trans-positioned benzimidazolylidene carbene in the 13C NMR acts as an internal probe and is directly influenced by the change in electron density caused by the trans-influence of the ligand. Notably, the HEPs for NHCs and MIC show a good correlation with the previously described 𝜎 −donor strength observed for the TEP.[116] Based on these reports, Ganter and co-workers used the 1J coupling constant of the C-H bond in the cationic NHC as an inexpensive and simple strategy to determine the 𝜎 −donor strength of the singlet carbene.[117] Beerhues et al. demonstrated the spectroscopic approach for MICs which showed that the 𝜎 −donor strengths of the investigated triazoliums salts were in a similar range, irrespective of their substituents (Figure 11).[101] Along this line, the group of Bertrand and Ganter introduced a method to determine the 𝜋 −acceptor capacities of the ligand by 31P and 77Se NMR spectroscopy. The chemical shift of the main group adduct originates from the E(py)→E−NHC(𝜋∗) (E = P, Se) backbonding and is in good agreement with the expected 𝜋 −acceptor properties of the incorporated ligand (Figure 11).[117,118] In 2020, Sarkar and co-workers investigated the mesoionic selones of the 1,2,3-triazolylidene type by 77Se NMR spectroscopy. The 𝜋 −acceptor ability of the corresponding MIC ligands is shown to be significantly influenced by the substituents at the N1 and C4 positions.[101] 1 Introduction 23 Figure 11. Spectroscopic methods for the determination of the 𝜎 −donor strength (left) and 𝜋 −acceptor capacity (right) of L = NHC, MIC. The probe functionalities are highlighted in red. The feasibility of incorporating a great variety of substituents on the 1,2,3-triazole and 1,2,3-triazolylidene moieties opened up a large toolbox for combining the electronic properties of each of the aforementioned classes of compounds in bi- and tridentate ligands. Despite the great developments of tridentate 1,2,3-triazole and 1,2,3- triazolylidene ligands, the following discussion will focus mainly on bidentate ligands.[37] In 1990, Lever introduced an alternative method for determining the ligand redox properties of RuII/III metal complexes using cyclic voltammetry. The ligand electrochemical parameter (LEP) describes the relative ability of ligands to stabilize a metal in a certain oxidation state and can be correlated with the overall 𝜎 −donor strength of the ligand. With the unambiguous assignment of a dominant ligand-centered reduction, the methodology can be applied to the 𝜋 −acceptor capacities of the ligand. Dominantly metal-centered oxidation or ligand-centered reduction can be assigned by EPR- and IR-SEC in combination with theoretical calculations.[86] In 2017, Suntrup et al. investigated fac-[(L−L)ReCl(CO)3] complexes, bearing bidentate ligands with at least one 1,2,3-triazole (triaz) or 1,2,3-triazolylidene (MIC) containing moiety by EPR-, IR- and UV/vis/NIR-SEC, supported by DFT-calculations.[119] The results indicate a predominantly metal-centered oxidation and ligand-based reduction, which allows the determination of the 𝜎 −donor strengths by IR-spectroscopy and the 𝜋 −acceptor properties by cyclic voltammetry (Figure 12). 1 Introduction 24 Figure 12. Comparison of 𝜎 −donor and 𝜋 −acceptor properties in fac-[(L−L)ReCl(CO)3]. Consistent with the aforementioned discussion, the incorporation of MICs drastically improved the 𝜎 −donor strength of the chelating ligands. The MIC-MIC ligand imparts greater stabilization of the oxidized metal-center, leading to an unusual reversible oxidation of the Re(I) central ion. Similar to its monodentate analogue, the triaz-triaz ligand shows the weakest 𝜎 −donor ability, followed by the well-established bpy ligand. While bpy exhibits the highest 𝜋 −acceptor capacity, the triaz-triaz ligand shows the poorest 𝜋 −acceptor properties among the ligands presented. Importantly, the inclusion of a pyridine ring drastically increases the 𝜋 −acceptor ability of the bidentate ligands. This observation is significant in many ways, since the interplay between excellent 𝜋 −acceptor and strong 𝜎 −donor properties leads to an increase in the thermal activation barrier (∆𝐸) of octahedral transition metal-based photocatalysts, resulting in a prolonged excited state lifetime and enhanced catalytic activity in a push-pull system (see section 1.1). In addition, according to the "iron law" (see section 1.2), an increased 𝜋 −acceptor ability lowers the overpotential of the electrocatalyst, while improving 𝜎 −donor strength enhances metal-substrate reactivity and hence catalytic activity. The ideal candidate to fulfill these requirements is the pyridiyl-MIC ligand (py-MIC). The bidentate ligand combines the strong 𝜋 −acceptor properties of the pyridine functionality with the excellent 𝜎 −donating ability of the MIC. The balanced synergy between the 𝜋 −accepting pyridyl-moiety and the strong 𝜎 −donating MIC formed the basis for the present thesis. 1 Introduction 25 1.4 Photochemistry Inspired by Mesoionic Carbenes In the broad history of transition metal-based complexes in photochemistry, the conceptual design of the incorporated ligand has become increasingly important in recent decades due to the possibility of tailor-made tunability of the excited state properties.[34,36,40,120,121] Mesoionic carbenes have shown to be particularly fruitful in increasing the excited state lifetimes of transition metal complexes, as their exceptional 𝜎 −donating properties are capable of increasing the ligand field strength in octahedral complexes and consequently the thermal barrier ∆𝐸 (see section 1.1). The shift of MC states to higher energies becomes particularly important in transition metal complexes of the 3rd period, as the intrinsic weak ligand field splitting leads to a rapid population of non-radiative 3/5MC states.[34,36,40,44,121] Despite the large developments of main group adducts in photochemistry,[122] as well as in early[123] and late[124] transition metal complexes, the following discussion will focus on octahedral transition metal complexes of group 7-9 in context of this thesis. In 2018, Wärnmark and co-workers reported an octahedral [Fe(btz)3]2+ (btz = 3,3- dimethyl-1,1-bis(p-tolyl)-4,4-bis(1,2,3-triazol-5-ylidene)) with an excellent excited-state lifetime of 528 ps (4th generation, Figure 13).[125] The complex shows a low oxidation potential of the Fe(II)/Fe(III) redox couple at −0.58 V, demonstrating significant impact of the strongly electron-donating MIC units. Indeed, the electron-rich nature of the MIC ligands enables the isolation of the oxidized [Fe(btz)3]3+ complex with an unusual 2LMCT and an excited state lifetime of 100 ps at room temperature (3rd generation, Figure 13).[126] The oxidized Fe(III) complex shows great potential as a photooxidant with potentials of +1.5 V (2LMCT) and +2.1 V (2MLCT) vs. Fc+/0, while the Fe(II) complex is a strong photoreductant with a potential of −1.6 V vs. Fc+/0.[125,126] Figure 13. Recent developments in bidentate iron-MIC complexes (R = p-tolyl). 1 Introduction 26 Replacing a btz-ligand with a 𝜋 −acceptor ligand, such as bpy, leads to a push-pull system that has proven to be a key concept for enhancing the ligand field strength in various transition metal complexes.[34,40] The 3MLCT excited state lifetime in [Fe(btz)2(bpy)]2+ (2nd generation, Figure 13)[127] could be increased by three orders of magnitude (13 ps) compared to the 'classical' [Fe(bpy)3]2+ complex (13 fs, 1st generation, Figure 13)[128] and shows a remarkable photochemical stability in overnight laser experiments. The photostability of transition metal complexes is often correlated with the distortion of the excited state. A common strategy to suppress excited-state distortion is to increase the structural rigidity by tridentate ligands and to expand the 𝜋 −conjugation of the ligand framework. In addition, the higher symmetry of the chelating ligand optimizes the bite angle of the system, leading to increased overlap of the metal-ligand orbitals.[120] Recently, a Fe(II) complex with two tridentate ligands containing two MIC units and a central pyridyl moiety with an extended 𝜋 −system was studied photochemically (I, Figure 14). However, the tridentate ligands lead to stronger distortion of the ideal octahedral geometry compared to their bidentate counterparts, resulting in reduced ligand field splitting and excited state lifetimes (𝜆𝑒𝑥 650 𝑛𝑚 = 3.8−120 ps; 𝜆𝑒𝑥 400 𝑛𝑚 = 3.7−1910 ps). At 400 nm, the population of an unusually long-lived dissociative 5MC occurs, while excitation at 650 nm leads to the population of the 3MLCT state.[129] Photochemical investigations of Ru(II) MIC complexes are reported in great detail. Replacement of the one terpyridine ligand (tpy) with the MIC-derived pincer ligand produces a push-pull system II (Figure 14) with an excited state lifetime of 𝜏 > 600 ns, 2500 times longer than observed for [Ru(tpy)2]2+ and similar to that reported for [Ru(bpy)3]2+. The 1MLCT excitation is associated with the tpy ligand, whereas the 3MLCT emission is attributed to the MIC ligand after the electronic redistribution during vibrational relaxation and ISC.[130] Modification of the tpy ligand in III, via the insertion of an electron-withdrawing groups to increase the 𝜋 −acceptor properties and the incorporation of long alkyl chains on the MIC-ligand, to reduce recombination reactions in dye-sensitized solar cells, leads to lifetimes of up to 410 ns (Figure 14).[131] Further improvement of the MIC pincer ligand by replacing the substituents at the N1-position of the MIC with aryl-units and extending the 𝜋 −conjugation with furanyl- substituents in the para-position of the tpy ligand could drastically enhance the excited state lifetimes up to 7.9 𝜇s (IV, Figure 14).[132] 1 Introduction 27 Figure 14. Excited state lifetimes of MIC-pyridyl-MIC pincer ligands in Fe(II)/Ru(II) complexes. Ruthenium complexes with bidentate MIC ligands and their higher osmium analogues exhibit less impressive photophysical properties (Figure 15).[37] The excited state lifetimes usually do not exceed values of 𝜏 > 300 ns and the quantum yields of 5.9% which are lower than those reported for [Ru(bpy)3]2+. However, natural transition orbitals analysis of [Ru(di-MIC)(bpy)2]2+ reveals that the Ru-bpy centered excited state can be significantly influenced by the MIC moieties. The influence of the MIC moiety is particularly evident in the case of the photoredox properties, as observed in the Latimer diagrams of the investigated [M(di-MIC)(bpy)2]2+ and [M(py-MIC)(bpy)2]2+ complexes (M = Ru, Os). In comparison to the well-established [M(bpy)3]2+ (M = Ru, Os) complexes, the oxidative quenching potential ( 𝐸 ∗ 𝑟𝑒𝑑) is increased by −400 mV after the incorporation of two MIC units, making MIC-based complexes attractive candidates for photoreductive applications.[133] Indeed, the MLCT lifetimes of the reported complexes are long enough to enable photoinduced electron transfer reactions in photocatalysis or electron injection into semiconductors.[134] Figure 15. Investigated bidentate Ru(II) and Os(II) MIC complexes. 1 Introduction 28 Iridium complexes play a prominent role in the area of photochemistry and photophysics, due to their high ligand field splitting, the large size of the d-orbitals and the higher ionic charge precluding non-emissive states and photodissociation and leading to a significant increase in excited state lifetimes. Cyclometallation of the chelating ligand, such as 2-phenylpyridine (= ppy), in [Ir(ppy)3], increases the covalency of the metal-ligand bond and contributes significantly to the metal-centered HOMO, while the LUMO is predominantly localized on the 𝜋 −system of the ligand.[135] In 2018, E. Matteucci et al. investigated a series of cyclometalated MIC ligands in [Ir(L−MIC)(ppy)2]0/+ (L = py, triaz, triazolide) as an alternative to standard chelating systems, such as 1,10-phenanthroline (= phen) and bpy (Figure 16). Unfortunately, the complexes [Ir(py−MIC)(ppy)2]+ and [Ir(triaz−MIC)(ppy)2]+ show very poor quantum yields of only 1% in acetonitrile. Theoretical calculations revealed that the luminescence is quenched by the lowest 3MC state, leading to a reversible detachment of the metal- nitrogen bond, similar to the observations noted for the NHC counterparts.[136] In contrast, [Ir(triazolid−MIC)(ppy)2] shows quantum yields of up to 12% in acetonitrile, which could be attributed to the increased metal-triazolide bond strength. Accordingly, the beneficial bonding situation in [Ir(triazolid−MIC)(ppy)2] suppresses the non-radiative deactivation pathway via the dissociative 3MC state.[137] Earlier reports by Baschieri et al. with two py-MIC ligands and two chloride ligands in [Ir(py−MIC)2Cl2]+ show luminescence from the LC state after excitation to the MLCT state. The replacement of the two chloride ligands by a bi-tetrazolate ligand (= bi-tetr) causes a blue shift of the emission (Figure 16). Both complexes exhibit quantum yields comparable to the archetypal [Ir(ppy)2(bpy)]+ and other chloride-containing Ir(III) complexes ranging from 1% to 12%.[138] High quantum yields were obtained by Barnard and co-workers. They presented a new series of Ir(III) complexes, combining MICs with NHCs in [Ir(NHC−MIC)(ppy)2]+ (Figure 16). The complexes investigated showed quantum yields of up to 57%. Furthermore, preliminary studies as luminescent probes for cell imaging were also conducted, demonstrating the great potential of MIC-based Ir(III) complexes for biological applications.[139] 1 Introduction 29 Figure 16. Selected Ir(III) complexes with MIC ligands. Rhenium complexes of the fac-[Re(CO)3X] (X = halide) type are excellent candidates for the photocatalytically conversion of CO2,[140] since the metastable 3MLCT state can induce the dissociation of the halide co-ligand by the thermal population of the 3MC. However, the photo-induced ligand dissociation strongly depends on the excitation energy, temperature and the choice of solvent, giving access to long-lived emissive states.[141] Recently, Suntrup et al. reported a series of py-MIC-based fac-[ReL(CO)3Cl] complexes with different substituents at the N1-positon (Scheme 7, section 1.5). Irradiation at 360 nm in DMF at room temperature leads to excited state lifetimes of up to 56 ns, similar to those reported for [Re(bpy)(CO)3Cl]. The insertion of the electron-donating MIC moiety shifts the emission to higher energies (550 nm) compared to the Re(I)-bpy counterpart (574 nm), demonstrating the appreciable influence of the MIC ligand.[106] In addition, the investigated fac-[Re(py−MIC)(CO)3Cl] complexes are catalytically active in the electrochemical conversion of CO2 and show a high selectivity for the formation of CO under protic conditions, which will be the subject of the following section. 1 Introduction 30 1.5 Recent Developments in MIC-based Electrocatalysis Probably one of the best known molecular electrocatalysts reported so far is the so-called Lehn catalyst. The fac-[Re(bpy)(CO)3Cl] complex shows an excellent Faradaic efficiency (FE) of 98% for the selective conversion of CO2 to CO in a 9:1 DMF/H2O mixture at mild potentials of −1.5 V vs. NHE with a turnover frequency of 21.4 h-1.[142] Nearly 40 years later, after Lehn's seminal discovery, Suntrup et al. investigated the analogous pyridyl-MIC fac-[Re(py−MIC)(CO)3Cl] in electrochemical CO2 reduction.[106] Earlier reports by the group of Sarkar and co-workers on the electronic structures of fac-[Re(py−MIC)(CO)3Cl] allowed an in-depth characterization of the redox stability in different redox states by cyclic voltammetry, IR-, EPR- and UV/vis/NIR-SEC, combined with (TD-)DFT calculations.[119] At a later stage, IR-SEC measurements were performed in a CO2-saturated DMF/Bu4NPF6 solution to obtain detailed information on the catalytic intermediates (Scheme 7). Scheme 7. Proposed pathways of fac-[Re(py−MIC)(CO)3Cl] (R = dipp) in the absence (top, first reduction) and in the presence of CO2 (bottom). 1 Introduction 31 The first reduction was assigned to a ligand-centered reduction, as shown by the EPR-, UV/vis/NIR- and IR-SEC measurements and was further supported by TD-DFT calculations. The scan rate dependency in cyclic voltammetry reveals an EC mechanism in the reduction, that can be attributed to the dissociation of chloride from the Re(I) metal center. Accordingly, the newly generated species detected at −1.83 V originates from either the solvent adduct fac-[Re(py−MIC)(CO)3DMF] or the coordinatively unsaturated fac-[Re(py-MIC)(CO)3] complex. In addition, IR-SEC measurements supported the presence of an EC mechanism. The metal-bound CO ligands are ideal probes to provide detailed information about the electronic structure of the complexes, as their position is directly influenced by the changes in electron density. The small shift of the CO stretching frequencies of about 30 cm-1 to lower wavenumbers observed in fac-[Re(py−MIC)(CO)3] indicates a ligand- centered reduction, since the increase in electron density is not directly induced by the metal center. However, prolonged electrolysis leads to dissociation of the chloride ligand, as shown by the absence of the isosbestic points and the appearance of several new CO bands. The identification of an EC mechanism becomes particularly important in the electrochemical conversion of CO2. Dissociation of the chloride ligand facilitates the metal-substrate interaction in fac-[Re(py−MIC)(CO)3]. In the presence of CO2, IR-SEC measurements showed exceptional reactivity of the singly-reduced species. In contrast to the archetypal Lehn catalyst, fac-[Re(py-MIC)(CO)3Cl] is potentially capable of catalyzing CO2 in a one-electron pathway without the need to generate the doubly reduced fac-[Re(py−MIC)(CO)3]− species (two-electron pathway). The intense CO2 band at 2338 cm-1 disappeared completely during the spectroelectrochemical measurement at −2.1 V vs. Fc/Fc+ and the obtained CO bands are not identical with the aforementioned singly reduced species in the absence of CO2. Spectral similarity was observed only after almost all CO2 has been converted during electrolysis. The difference in the IR bands formed during the reduction in CO2 and argon atmosphere suggests the formation of a CO2-containing intermediate, such as fac-[Re(CO2)(py−MIC)(CO)3] or a CO2-bridged dimer.[63,143] Other reactions, such as the Re-Re dimerization or the formation of the solvent adduct fac-[Re(py−MIC)(CO)3DMF], are presumably suppressed in the presence of CO2. 1 Introduction 32 The electrochemical reduction of CO2 was tested with two differently substituted fac-[Re(py−MIC)(CO)3Cl] complexes in the presence and absence of methanol as a proton source in DMF at −2.3 V vs. Ag/AgNO3. For comparison, the Lehn catalyst fac-[Re(bpy)(CO)3Cl] was investigated under identical conditions at −1.9 V vs. Ag/AgNO3. All complexes investigated exhibit high selectivity for the formation of CO. However, substitution at the MIC unit has a drastic effect on the catalytic activity, as indicated by the FE. The substitution with −CH2PhF shows a FE of only 64%, while the dipp-substituted fac-[Re(py−MIC)(CO)3Cl] complex shows an excellent FE of 99%, outperforming the Lehn catalyst with 71% FE. The influence of the substitution at the MIC unit is also evident in the TON and TOF of the complexes studied. The fluorinated fac-[Re(py−MIC)(CO)3Cl] complex shows a TON of 110 and a TOF of 0.03 s−1, similar to the values observed for the fac-[Re(bpy)(CO)3Cl] complex. In contrast, the dipp-substituted fac-[Re(py−MIC)(CO)3Cl] complex shows a TON of 191 and a TOF of 0.08 s−1, almost three times higher of that observed for archetypal Lehn catalyst.[106] The groups of Piers and Royo investigated the lower homologues with the constitutional pyridyl-MIC isomers of the fac-[Mn(py−MIC)(CO)3Br] type. In addition, Royo and co-workers studied the influence of the chelating ligand by replacing the pyridyl N-heterocycle with a 1,2,3-trizole and a MIC moiety in fac-[Mn(MIC−MIC)(CO)3Br] and fac-[Mn(triaz−MIC)(CO)3Br] (Figure 17).[104,105] Figure 17. Investigated MIC-Mn(I) complexes (top) and isolated intermediates (bottom). 1 Introduction 33 Manganese precatalysts are known to dimerize rapidly upon one-electron reduction.[144] The dimer itself can react catalytically by cleaving the Mn−Mn bond and subsequent insertion of CO2.[145] However, dimerization of the Mn(0) species leads to an increase in the overpotential.[91,146] A common strategy to suppress dimerization is to incorporate bulky substituents near the metal center to slow the rate of dimerization and consequently to reduce the overpotential. The electrocatalysts are extremely photosensitive because of the weak ligand field splitting, leading to rapid CO dissociation, as previously discussed in section 1.4. The insertion of strongly 𝜎 −donating ligands, such as MICs facilitate the compensation of the electron deficiency of the formally oxidized metal center and increases the ligand field splitting, which is essential for the stability of the electrocatalyst.[147] In the series of fac-[Mn(L)(CO)3Br] complexes investigated by Royo and co-workers, only moderate catalytic activity was observed for the electrochemical reduction of CO2 in MeCN and in the presence of H2O. The low efficiency was attributed to the electrodeposition of the catalysts during the electrolysis experiment. The highest FE was observed for the fac-[Mn(MIC−MIC)(CO)3Br] complex with 70%. In contrast, Piers and co-workers investigated the Cpy−NMIC linked constitutional isomer and demonstrated that the fac-[Mn(py−MIC)(CO)3Br] complex is robust under the experimental conditions and can operate at two well-separated potentials (ΔE = 400 mV) for electrochemical CO2 reduction.[105] In the low operating regime at −1.54 V, CO2 is converted by a Mn(0) species to CO and CO3 2− at a maximum rate of 7 s−1 for nearly 30.7 hours. The product analysis at a higher potential of −1.94 V displays the selective formation of CO and H2O with a TOF of 200 s−1. The catalytically active species was assigned to the fac-[Mn(py−MIC)(CO)3]− species. However, the operating time is limited to 6.7 hours, which demonstrates the synergy in the structure-reactivity relationship of the electrocatalyst used. The isolation of various predicted intermediates, such as the Mn−Mn dimer, the cationic fac-[Mn(py−MIC)(CO)4]+ complex and the two-electron reduced species, allowed an in-depth (spectro-)electrochemical analysis of the (pre-)catalytic activation pathway at the different operating potentials (Scheme 8). 1 Introduction 34 Scheme 8. Illustration of the redox chemistry in the fac-[Mn(py−MIC)(CO)3Br] complex based on isolated intermediates (adapted from Piers and co-workers).[105] Cyclic voltammetry fac-[Mn(py−MIC)(CO)3Br] reveals rich redox chemistry. In particular, the cyclic voltammogram in MeCN shows a 120 mV shift in the first reduction (𝐸𝑝𝑐,1 = −1.57 V) to a more anodic potential compared to the cyclic voltammogram recorded in DMF (𝐸𝑝𝑐,1 = −1.69 V), indicating a rapid halide/solvent dissociation in MeCN. Accordingly, the first reduction can be assigned to the halide/solvent dissociation of the fac-[Mn(py−MIC)(CO)3Br] and/or fac-[Mn(py−MIC)(CO)3S] complex (S = MeCN, DMF), leading to the radical Mn(0) species. The radical Mn(0) species can undergo rapid dimerization to the Mn−Mn complex. Electrochemical investigation of the isolated dimer confirmed this observation, as indicated by the consistent reduction potential at −1.87 V, leading to the monoanionic complex [Mn(py−MIC)(CO)3]−. Reverse anodic scan leads to the formation of the Mn(0) radical species, followed by subsequent dimerization. Finally, oxidation of the dimer leads to the formation of the solvent adduct fac-[Mn(py−MIC)(CO)3S] (S = MeCN, DMF) or the respective starting complex fac-[Mn(py−MIC)(CO)3Br] in the presence of an excess of bromide. IR-SEC in the presence of CO2 and a proton source indicated the formation of CO and carbonates at lower reduction potentials, which was further confirmed by the precipitation of MgCO3 after prolonged electrolysis. In contrast, electrolysis at higher potentials shows selective formation of CO and H2O via an EECC mechanism, yielding fac-[Mn(py−MIC)(CO)4] after the consumption of CO2, as mentioned earlier. 1 Introduction 35 The exceptional properties of the pyridyl-MIC ligand were also investigated by Sarkar and co-workers in the electrochemical H+ reduction with a [Co(Cp*)(py−MIC)Cl]+ complex (Scheme 9). The complex exhibits a low overpotential (130 mV) and a TOF of 4 ∙ 102 s-1 with a glassy carbon electrode. The TON was determined at about 650 000 during the 30 min bulk electrolysis experiment and the Co(III) complex shows remarkable stability towards acidic acid due to the unique robustness of the metal-MIC bond. On basis of the electrochemical data, a (catalytic) mechanism was proposed.[107] Scheme 9. Proposed mechanism for electrochemical H+ reduction for a Co(III)-MIC catalyst.[107] Upon reduction, an electron-transfer/chemical reaction (EC) mechanism, accompanied by dissociation of the chloride ligand, is proposed. Similar observations have been made for [Co(Cp*)(bpy)Cl]+ complexes.[148] The first reduction at −1.1 V vs. Fc/Fc+ is metal-centered, as indicated by the similar potential for the investigated Co(III) MIC−MIC and triaz−py counterparts. The second reduction shifts according to the 𝜋 −acceptor capacities of the ligands, indicating a 'non-innocent' nature of the ligand, followed by an electron density distribution to produce an active Co(I) metal center. In the presence of a proton source, the formation of a Co(III)-hydride species is postulated, which can facilitate dihydrogen formation by a proton-coupled electron transfer (PCET) to the Co(II) species or regeneration of the starting complex after chloride coordination. The electrochemical studies revealed first-order dependence on the catalyst and a second- order for acetic acid, further corroborating the proposed mechanism. References 36 References [1] Nobel Prize Outreach AB 2023, "Carolyn Bertozzi – Facts – 2022", can be found under https://www.nobelprize.org/prizes/chemistry/2022/bertozzi/interview/, 2023. [2] a) N. Armaroli, V. Balzani, Energy for a Sustainable World. From the Oil Age to a Sun- Powered Future, Wiley-VCH, Weinheim, Germany, 2011; b) T. B. Johansson, Global Energy Assessment. Toward a Sustainable Future, Cambridge Univ. Press, Cambridge, 2012. [3] J. Lelieveld, A. Pozzer, U. Pöschl, M. Fnais, A. Haines, T. Münzel, Cardiovasc. Res. 2020, 116, 1910–1917. [4] a) M. Dittmar, Energy 2012, 37, 35–40; b) J. W. Eerkens, The Nuclear Imperative, Springer, Dordrecht, 2010; c) R. Wengenmayr, Renewable Energy. Sustainable Concepts for the Energy Change, Wiley-VCH, Weinheim, 2013. [5] A. Kumar, P. Daw, D. Milstein, Chem. Rev. 2022, 122, 385–441. [6] K. Sordakis, C. Tang, L. K. Vogt, H. Junge, P. J. Dyson, M. Beller, G. Laurenczy, Chem. Rev. 2018, 118, 372–433. [7] B. Milani, G. Licini, E. Clot, M. Albrecht, Dalton Trans. 2016, 45, 14419–14420. [8] P. W. Roesky, A. R. Fout, Inorg. Chem. 2021, 60, 13757–13758. [9] R. Poli, Comments Inorg. Chem. 2009, 30, 177–228. [10] M. Y. Darensbourg, A. Llobet, Inorg. Chem. 2016, 55, 371–377. [11] a) N. Lehnert, J. C. Peters, Inorg. Chem. 2015, 54, 9229–9233; b) S. Amanullah, P. Saha, A. Nayek, M. E. Ahmed, A. Dey, Chem. Soc. Rev. 2021, 50, 3755–3823; c) C. W. Koo, A. C. Rosenzweig, Chem. Soc. Rev. 2021, 50, 3424–3436; d) N. Lehnert, B. W. Musselman, L. C. Seefeldt, Chem. Soc. Rev. 2021, 50, 3640–3646; e) C. J. Reed, Q. N. Lam, E. N. Mirts, Y. Lu, Chem. Soc. Rev. 2021, 50, 2486–2539; f) X.-P. Zhang, A. Chandra, Y.-M. Lee, R. Cao, K. Ray, W. Nam, Chem. Soc. Rev. 2021, 50, 4804–4811; g) M. R. A. Blomberg, Chem. Soc. Rev. 2020, 49, 7301–7330. [12] F. Meyer, W. B. Tolman, Inorg. Chem. 2015, 54, 5039. [13] a) G. T. Babcock, M. Wikström, Nature 1992, 356, 301–309; b) A. J. Gow, H. Ischiropoulos, J. Cell Physiol. 2001, 187, 277–282; c) M. Wikström, V. Sharma, V. R. I. Kaila, J. P. Hosler, G. Hummer, Chem. Rev. 2015, 115, 2196–2221. [14] J. Barber, P. D. Tran, J. R. Soc. Interface 2013, 10, 20120984. [15] a) C. Descorme, P. Gallezot, C. Geantet, C. George, ChemCatChem 2012, 4, 1897–1906; b) I. Fechete, Y. Wang, J. C. Védrine, Catal. 2012, 189, 2–27; c) X. Jiang, X. Nie, X. Guo, C. References 37 Song, J. G. Chen, Chem. Rev. 2020, 120, 7984–8034; d) J. M. Thomas, ChemSusChem 2014, 7, 1801–1832; e) S. Liu, M. Wang, Q. Cheng, Y. He, J. Ni, J. Liu, C. Yan, T. Qian, ACS Nano 2022, 16, 17911–17930. [16] a) D. Ar, A. F. R. Kilpatrick, B. Cula, C. Herwig, C. Limberg, Inorg. Chem. 2021, 60, 13844–13853; b) L. Chen, G. Chen, C.-F. Leung, C. Cometto, M. Robert, T.-C. Lau, Chem. Soc. Rev. 2020, 49, 7271–7283; c) M. Gennari, C. Duboc, Acc. Chem. Res. 2020, 53, 2753–2761; d) R. Latifi, T. D. Palluccio, W. Ye, J. L. Minnick, K. S. Glinton, E. V. Rybak-Akimova, S. P. de Visser, L. Tahsini, Inorg. Chem. 2021, 60, 13821–13832; e) L. C. Lewis, H. S. Shafaat, Inorg. Chem. 2021, 60, 13869–13875; f) Q. Liang, J. C. DeMuth, A. Radović, N. J. Wolford, M. L. Neidig, D. Song, Inorg. Chem. 2021, 60, 13811–13820; g) M. Springborg, J.-O. Joswig (Eds.) Chemical Modelling, Royal Society of Chemistry, Cambridge, 2019; h) J. I. van der Vlugt, Eur. J. Inorg. Chem. 2012, 2012, 363–375; i) R. L. Meyer, P. Miró, W. W. Brennessel, E. M. Matson, Inorg. Chem. 2021, 60, 13833–13843. [17] F. Möller, S. Piontek, R. G. Miller, U.-P. Apfel, Chem. Eur. J. 2018, 24, 1471–1493. [18] L. Lindh, P. Chábera, N. W. Rosemann, J. Uhlig, K. Wärnmark, A. Yartsev, V. Sundström, P. Persson, Catalysts 2020, 10, 315. [19] a) Y. Wang, H. Suzuki, J. Xie, O. Tomita, D. J. Martin, M. Higashi, D. Kong, R. Abe, J. Tang, Chem. Rev. 2018, 118, 5201–5241; b) Y. Yao, X. Gao, Z. Li, X. Meng, Catalysts 2020, 10, 335. [20] D.-M. Feng, Y.-P. Zhu, P. Chen, T.-Y. Ma, Catalysts 2017, 7, 373. [21] a) C. Finn, S. Schnittger, L. J. Yellowlees, J. B. Love, Chem. Commun. 2012, 48, 1392– 1399; b) N. Kaeffer, W. Leitner, JACS Au 2022, 2, 1266–1289; c) C. Tang, Y. Zheng, M. Jaroniec, S.-Z. Qiao, Angew. Chem. Int. Ed. Engl. 2021, 60, 19572–19590; d) X. Yin, J. R. Moss, Coord. Chem. Rev. 1999, 181, 27–59; e) L.-H. Zhang, S. Mathew, J. Hessels, J. N. H. Reek, F. Yu, ChemSusChem 2021, 14, 234–250; f) R. Zhang, J. J. Warren, ChemSusChem 2021, 14, 293–302. [22] W.-H. Wang, Y. Himeda, J. T. Muckerman, G. F. Manbeck, E. Fujita, Chem. Rev. 2015, 115, 12936–12973. [23] R. Francke, B. Schille, M. Roemelt, Chem. Rev. 2018, 118, 4631–4701. [24] a) M. E. Ali, M. M. Rahman, S. M. Sarkar, S. B. A. Hamid, J. Nanomater. 2014, 2014, 1– 23; b) A. Z. Fadhel, P. Pollet, C. L. Liotta, C. A. Eckert, Molecules 2010, 15, 8400–8424. [25] A. J. Esswein, D. G. Nocera, Chem. Rev. 2007, 107, 4022–4047. [26] T. Stoll, C. E. Castillo, M. Kayanuma, M. Sandroni, C. Daniel, F. Odobel, J. Fortage, M.-N. Collomb, Coord. Chem. Rev. 2015, 304-305, 20–37. References 38 [27] G. Knör, Coord. Chem. Rev. 2015, 304-305, 102–108. [28] K. P. S. Cheung, S. Sarkar, V. Gevorgyan, Chem. Rev. 2022, 122, 1543–1625. [29] L. Guillemard, J. Wencel-Delord, Beilstein J. Org. Chem. 2020, 16, 1754–1804. [30] a) C. K. Prier, D. A. Rankic, D. W. C. MacMillan, Chem. Rev. 2013, 113, 5322–5363; b) D. M. Schultz, T. P. Yoon, Science 2014, 343, 1239176; c) F. Strieth-Kalthoff, F. Glorius, Chem 2020, 6, 1888–1903. [31] J.-H. Shon, T. S. Teets, Comments Inorg. Chem. 2020, 40, 53–85. [32] B. O'Regan, M. Grätzel, Nature 1991, 353, 737–740. [33] a) R. Ham, C. J. Nielsen, S. Pullen, J. N. H. Reek, Chem. Rev. 2023, 9, 5225–5261; b) M. Schulz, M. Karnahl, M. Schwalbe, J. G. Vos, Coord. Chem. Rev. 2012, 256, 1682–1705; c) H. Kumagai, Y. Tamaki, O. Ishitani, Acc. Chem. Res. 2022, 55, 978–990. [34] C. Förster, K. Heinze, Chem. Soc. Rev. 2020, 49, 1057–1070. [35] S. Treiling, C. Wang, C. Förster, F. Reichenauer, J. Kalmbach, P. Boden, J. P. Harris, L. M. Carrella, E. Rentschler, U. Resch-Genger, C. Reber, M. Seitz, M. Gerhards, K. Heinze, Angew. Chem. Int. Ed. Engl. 2019, 58, 18075–18085. [36] C. Wegeberg, O. S. Wenger, JACS Au 2021, 1, 1860–1876. [37] R. Maity, B. Sarkar, JACS Au 2022, 2, 22–57. [38] P