Compuestos organometálicos en reacciones de acoplamiento deshidrogenante, transformación de dióxido de carbono y otros procesos catalíticos.

Tesis Doctoral Compuestos organometálicos en reacciones de acoplamiento deshidrogenante, transformación del dióxido de carbono y otros procesos catalíticos Pablo Ríos Moreno Sevilla, 2019 Departamento de Química Inorgánica Facultad de Química – Universidad de Sevilla Instituto de Investigaciones Químicas Consejo Superior de Investigaciones Científicas Fdo. Salvador Conejero Iglesias Fdo. Amor Rodríguez Iglesias Científico Titular Científico Titular Instituto de Investigaciones Químicas Instituto de Investigaciones Químicas "Nothing is so dangerous to the progress of the human mind than to assume that our views of science are ultimate, that there are no mysteries in nature, that our triumphs are complete and that there are no new worlds to conquer." Sir Humphry Davy “Don’t waste your time, or time will waste you.” Muse A mis padres Agradecimientos Bueno, pues me temo que esta etapa está llegando a su fin, y es hora de ir cerrando el chiringuito. A su vez, es bueno hacer balance de todo lo vivido durante estos 4 años y sacar conclusiones que nos hagan mejorar un poco más en lo que queda más adelante. Aunque he aprendido infinidad de cosas en el ámbito profesional, las personas que han compartido este periodo de tiempo conmigo son por supuesto lo más importante, y es algo injusto que solo tengan unas pocas páginas en todo este trabajo, ya que sin ellas habría sido impensable llevar esto a cabo. Sin embargo, el realizar la tesis me ha enseñado a ser optimista sobre todo, así que vamos a aprovechar el poquito espacio del que disponemos, ¿no? No puedo empezar de otra forma que agradeciendo a mis dos directores de tesis, Amor y Salva, la oportunidad que me brindaron al poder realizar la tesis doctoral con ellos. Confiaron en mí desde el principio, y desde entonces no han dejado de apoyarme, aconsejarme y animarme durante todo este tiempo. Ha sido todo un placer poder trabajar con vosotros, montar vuestro propio laboratorio desde 0, y haber aportado mi granito de arena para que vuestro proyecto salga hacia delante, porque no conozco a nadie que lo merezca más. Repetiría la experiencia una y mil veces. Si mi predecesor Orestes se alegraba de ser el primero de muchos Conejeros, yo me alegro enormemente de ser el primero de muchos Rodríguez-Conejero. Gracias por compartir toda vuestra experiencia conmigo, y por contagiarme vuestro entusiasmo por la química organometálica. Ojalá todo vaya bien, y podamos volver a trabajar codo con codo dentro de no mucho tiempo (¡crucemos los dedos!). Sé que sois un tándem perfecto, pero es injusto que os considere como tal yo también, y más cuando he tenido la oportunidad de conoceros mejor a cada uno por separado, así que allá voy. Amor, ¿qué voy a decir que en realidad no sepas? Sabes de sobra que detrás de todas esas bromas y chistes hay muchísimo afecto, respeto y admiración. Gracias por ser una tía tan clara, tanto para lo bueno como para lo malo, y tan buena persona. En estos tiempos donde por fin las mujeres están adquiriendo el mismo protagonismo que los hombres, tú eres un claro ejemplo de profesional en la que fijarse. Sigue así, tanto dentro como fuera del laboratorio. Salva, gracias por aportarme tantísimo también, tanto a nivel profesional como personal. Es difícil olvidar las incontables charlas sobre química en el despacho, y cómo nos veníamos arriba con un folio en blanco y un boli, sugiriendo ideas. En esos momentos intentaba encender el modo esponja y absorber toda la información posible. Aunque me alegro de no haber aprendido solo eso, ya que la ética y la moral también se pegan. Agradezco la bondad y amabilidad con la que la puerta siempre ha estado abierta para cualquier cosa, fuese de química o no. No cambies tú tampoco, porque gente tan buena como tú escasea. Los siguientes en la lista son los que desgraciadamente me han tenido que aguantar durante la tesis. Los que me sufrieron en el Laboratorio 9 al principio, y en el Laboratorio 203 en la segunda mitad, son mis compañeros de fatigas Leonardo y Jose. Leonardo es el brasileño más grande que hay, al igual que los chicharrones (según él). No conozco un tío que sea más “cacho de pan” que tú. Echando la vista atrás, creo que no ha habido ni un solo día en estos 4 años en el que no te haya visto con una sonrisa trabajando. Espero que ese detalle no cambie nunca, y que la suerte te siga en tu vuelta a Brasil. En cuanto a Jose, sabes de sobra que se te echa en falta muchísimo, tanto fuera como dentro del lab. Eres una máquina trabajando (bien lo saben los jefes), y eres otra persona buenísima de las que merece la pena conocer. Gracias por haber estado ahí día sí día también echándome un cable, o animándome el día con algún gesto o comentario positivo. Por supuesto no puedo olvidarme tampoco de Hugo, o el nota. En tan solo unos meses de estancia el tío se hizo querer muchísimo en todo el grupo y en el instituto (lo cual es muy comprensible). Además, su ayuda fue crucial en el proyecto de platino, ya que gracias a él pudimos darle un empujón bueno cuando nos faltaban manos para sacar adelante todo lo que nos estábamos proponiendo. Me alegro de no haber perdido el contacto, y de seguir visitándonos de vez en cuando. Espero que siga así! La siguiente parada obligada es el laboratorio 1-2. En primer lugar, me gustaría agradecerle a Juanjo (o GuajitoDFT, o Juanj) todos los buenos momentos disfrutados dentro y fuera de cartuja. Tal y como te he descrito alguna vez, has sido uno de los grandes descubrimientos de la tesis (y reitero la palabra grande, casi como tus tapers de comida). Sabes que he aprendido mucho de ti, tanto de química como de tu forma de ser. Tío más noble no hay, y además es el responsable de que gran parte de mi dieta en la última etapa de la tesis esté basada en pad thai de pollo al 3 (qué sudores más ricos). Te mereces que te vaya de maravilla (a ti y a Andrea, ¡que bastante te aguanta también!), y espero que así sea. Moviéndome por el laboratorio llego hasta el mítico puesto de mando del señor Carlos, el terror de las galaxias del Ogame. A ti te tengo que agradecer enormemente que le des vida a la parte inorgánica del IIQ, y la hagas un poco más “orgánica” en el sentido social. La cantidad de risas y buenos momentos echados allí han hecho las innumerables horas bastante más amenas, así que muchísimas gracias. Por otro lado, agradecer también a los otros habitantes de este lab: Mario, Juanín, Sonia, Marina, Nere (los más senior), y a la nueva generación que ha llegado ahora, como Macarena, Lola o María. No se me olvidan antiguos huéspedes de este lab como Juanje o Macarena, o antiguas doctorandas como Ángela, Laura, María y Natalia, a las cuales les agradezco todo lo que me enseñaron durante mi primer año de tesis y los buenos ratos que pasamos en diferentes comidas y escapadas. Dicho laboratorio sería un completo caos si no fuese por la presencia de Riccardo o de Jesús (antiguo compañero de lesiones futbolísticas). En el lab 3-4 se encuentra Tomás, un tío currante donde los haya y bueno a más no poder también. Espero que no pierdas ese entusiasmo que te caracteriza, y que la última recta de la tesis sea lo más suave y exitosa posible. Este lab se encontraba frecuentado anteriormente por buena gente como John, Carmen o Astrid. Destacar también los numerosos buenos momentos compartidos con Elena. Espero que esa bondad que te caracteriza siga acompañándote allá por Cambridge. Por último, agradecer también a Juan, Pilar y Antonio Rodríguez lo atentos y dispuestos que han sido siempre conmigo. Para terminar con la parte organometálica, tengo que mencionar a los residentes del Lab 10-11/207, Félix y Práxedes. A ellos, y a los asturianos Pedro y Eire (breves visitantes), tengo que agradecerles que hicieran más amenas algunas salidas y comidas de navidad. No me puedo olvidar tampoco de la parte orgánica del instituto, que con la que si bien he tenido menos contacto durante este tiempo, lo poco que he podido coincidir con ellos ha sido muy bueno: Javi Ramos, Ainhoa, Abel, Valentín, Rute, Noelia, José Alberto, Tania, Pedro Domínguez, Antonio Romero, Antonio Di Maio, José Juan, Vero, Pedro Ramírez…¡Muchas gracias a todos! Por otra parte, merece mención especial Joaquín, por su enorme paciencia y disponibilidad a la hora de enseñarme a defenderme en el mundo de los cálculos DFT, y por su paciencia corrigiendo cada espacio accidental que introducía frecuentemente en el input, haciendo que el cálculo se fuese al garete. Y merecen mención especial “las niñas”, Laura, Nuria y Patri, por estar ahí con una sonrisa siempre que he necesitado hablar con ellas, y por guiarme con nociones básicas de docencia cuando me ha tocado trabajar con estudiantes, y no con líneas de vacío ni cámaras secas. Muchas gracias a los 4 por vuestro tiempo y amabilidad. Este trabajo no sería posible si no estuviese en el Instituto de Investigaciones Químicas. Por tanto, debo agradecer que éste se sostiene en dos pilares fundamentales: Fran y Marga. Mientras que el primero es un todoterreno en cuestiones técnicas, la segunda es una máquina en cuanto a burocracia se refiere. Sin ellos, todos los que trabajamos aquí estaríamos perdidos, y tardaríamos una eternidad en conseguir lo que ellos realizan en un periquete. Así que gracias, por facilitarnos muchísimo nuestro trabajo, y por hacerlo con la mejor de las sonrisas. Me gustaría agradecer también a José Manuel la eficacia y rapidez a la hora de analizar las muestras de análisis elemental, y a Agustín Cota (CITIUS) su inagotable paciencia para medir mis cristales, y probar uno tras otro tras otro y hacer lo imposible por conseguir buenos datos de rayos X que pudiéramos utilizar, ya que ha sido de gran ayuda. Muchas gracias por hacer que los ratos de incertidumbre fueran más llevaderos. Asimismo, agradezco a la Dra. Josefina Díez la resolución de numerosas estructuras de rayos X plasmadas en esta tesis, y al Dr. Javier Borge el esfuerzo y dedicación con una de las últimas estructuras de esta tesis (bastante complicada, por cierto). Por otro lado, debo darle las gracias nuevamente a Natalia en este aspecto, debido a que ha buscado tiempo cuando no lo había para dedicarle algún rato a la resolución de estructuras, algunas de las cuales eran duras de pelar. Quisiera mencionar también a Miguel Anaya, junto con el cual me comí numerosas reuniones que acabaron culminando en las PhD Talks, las cuales van ya por la tercera edición. Espero que dicho proyecto siga adelante durante mucho más tiempo. Por otra parte, el marco de la tesis doctoral ha facilitado la realización de estancias en diferentes universidades, en la cual he conocido a gente estupenda y de la cual he podido aprender muchísimo. En primer lugar, me gustaría agradecer al Prof. Agustí Lledós que me acogiera durante un mes en su grupo en la UAB. De la misma forma, agradezco a Sergi, Gantulga y Giuseppe toda la ayuda prestada durante mi estancia. No me puedo dejar fuera a Pablo y a Anna, los cuales se hicieron querer en poquísimo tiempo. ¡Os espero por aquí abajo! En segundo lugar, agradezco al Prof. Kit Cummins el poder trabajar durante 4 meses en sus laboratorios. Fue una experiencia fantástica, y me llevo muy buenos recuerdos de allí. Asimismo, sus estudiantes Wes, Mike, Scott, Martin, Yingying, Kevin, Wendy, Liz, June y Paul crearon un ambiente de trabajo genial. Sin embargo, me veo obligado a separar de esta última estancia a un grupo de personas que hicieron de mi período en Boston una etapa inolvidable, ya que hicieron que me sintiese en familia desde el minuto 1. Steve, Isabel, Leslie, Emily y Brendan, muchísimas gracias por todo lo vivido juntos. Y por supuesto, gracias a Tim, por ser como un hermano para mí. Te veo próximamente. Volviendo a tierras sevillanas, me gustaría mencionar a Angelita, Ana, Pedro y Víctor del IBVF, por todas las cervecitas y las risas que nos hemos echado (¡y por las que quedan!). Me gustaría agradecer a mis amigos toda la comprensión que han tenido conmigo cuando no he podido reunirme con ellos debido al trabajo, y por los buenos ratos y los ánimos que me han dado cuando sí ha sido posible verlos: Porras, Pedro, Rubén, Rulo, Bastos, Guille, Davile, Fran, Antonio, Luisa, y en especial a Germán y Ángela, los cuales me siguen aguantando después de todo este tiempo. Sois tela de grandes, y yo muy afortunado por teneros. Para ir terminando, me gustaría darles las gracias a ciertos miembros de mi familia que me han apoyado especialmente durante este período de tiempo, y los cuales también van sobrados de paciencia. Entre ellos se encuentran mis titos Rosa y Manolo, mis titos Charo y Juanmi (y por supuesto mis primas Irene y Esther) y mi tía Ani. Quiero mencionar también a mi familia morilense: Manuela, Manolo, Olga, Manuel, Salud, Charo…¡Muchas gracias por acogerme tan bien, y por entenderme cuando he faltado por allí! Por último, quiero agradecerle a mi hermano Javi y a mi cuñada Rebeca el haber estado siempre ahí, animándome y entendiéndome sin dudarlo un segundo, y disfrutando conmigo todos los buenos momentos que nos hayamos podido ir encontrando. Tengo muchísima suerte de que estéis ahí. Y por supuestísimo, me siento muy orgulloso de tener los mejores padres del mundo, Carmen y Juan, por todo el cariño con el que me han tratado siempre y por comprenderme, aunque en realidad no entendieran qué eran esas “colmenitas” que dibujaba día tras día y que me robaban semanas, meses, años... Sois un ejemplo a seguir, en absolutamente todo. De mayor quiero ser como vosotros, tanto dentro como fuera del curro. Gracias por enseñarme el valor del trabajo, y a disfrutar de la vida al mismo tiempo. Os quiero mucho. Y antes de dar comienzo al contenido de esta tesis, quiero darle las gracias a Valle. A ti, que llegaste por casualidad, quiero darte mil gracias por compartir tu vida conmigo, por ser tan buena y tan dulce, por animarme los días con solo verte, y por entusiasmarte conmigo ante todo lo bueno que nos espera. Sé que aunque el futuro que se nos plantea es algo incierto, sabremos cómo superar todo lo que nos propongamos, juntos, caminando de la mano. Te quiero. And our dreams will break the boundaries of our fears… General Considerations [4] P. Ríos, J. Díez, J. López-Serrano, A. Rodríguez, S. Conejero. Cationic Platinum(II) σ-SiH Complexes in Carbon Dioxide Hydrosilation. Chem. Eur. J., 2016, 22, 16791- 16795. DOI:10.1002/chem.201603524 P. Ríos, M. Roselló-Merino, O. Rivada-Wheelaghan, J. Borge, J. López-Serrano, S. Conejero. Selective catalytic synthesis of amino-silanes at part-per million catalyst loadings. Chem. Commun., 2018, 54, 619-622. DOI: 10.1039/C7CC08530C P. Ríos, H. Fouilloux, P. Vidossich, J. Díez, A. Lledós, S. Conejero. Isolation of a Cationic Platinum(II) σ-Silane complex. Angew. Chem. Int. Ed., 2018, 57, 3217- 3221. DOI: 10.1002/anie.201712791 P. Ríos, H. Fouilloux, J. Díez, P. Vidossich, A. Lledós, S. Conejero. Manuscript in preparation. σ-Silane Platinum(II) Complexes as Intermediates in C−Si Bond Coupling Processes. Manuscript in preparation. Papers not related to the content of this Doctoral Thesis P. Rios, T. S. Carter, T. J. Mooibroek, M. P. Crump, M. Lisbjerg, M. Pittelkow, N. T. Supekar, G.-J. Boons, A. P. Davis. Synthetic Receptors for the High-Affinity Recognition of O-GlcNAc Derivatives. Angew. Chem. Int. Ed., 2016, 55, 3387-3392. DOI: 10.1002/anie.201510611 P. Ríos, T. J. Mooibroek, T. S. Carter, C. Williams, M. P. Crump, A. P. Davis. Enantioselective Carbohydrate Recognition by Synthetic Lectins in Water. Chem. Sci., 2017, 8, 4056-4061. DOI: 10.1039/C6SC05399H M. B. Geeson, P. Ríos, W. J. Transue, C. C. Cummins. Oxoanions of Phosphorus and Sulfur as Entryways to Carbon−Element Bond Formation. Manuscript in preparation. General Considerations [5] Patents not related to the content of this Doctoral Thesis M. B. Geeson, P. Ríos, W. J. Transue, C. C. Cummins. Method for producing trichlorosilylsulfide anion and its use in the preparation of organosulfur compounds. Application number: 62/697,037. Consideraciones Generales [6] Consideraciones generales La presente Tesis Doctoral se ha llevado a cabo en el grupo Diseño de moléculas organometálicas y aplicaciones del Instituto de Investigaciones Químicas de Sevilla (centro mixto entre la Universidad de Sevilla y el CSIC), bajo la dirección de los doctores Salvador Conejero y Amor Rodríguez (ambos Científicos Titulares del CSIC). El contenido de esta tesis se encuentra incluido en el ámbito de la química inorgánica, y más concretamente en la química organometálica y la catálisis homogénea. Los resultados que aquí se discuten tratan sobre la síntesis y caracterización de complejos de níquel y platino estabilizados por ligandos del tipo di(fosfino)borilo (PBP) y carbeno N-heterocíclico (NHC), respectivamente. La reactividad de varios de estos complejos condujo a su empleo como catalizadores en procesos de reducción del dióxido de carbono, activación C−H, y acoplamiento deshidrogenante, estudiándose el mecanismo en la mayoría de los casos. A su vez, se ha llevado a cabo la detección y el aislamiento de diversos intermedios de reacción. La caracterización de los complejos obtenidos se ha realizado a través de técnicas espectroscópicas y de difracción de rayos X. En este último aspecto, el tratamiento de datos y la resolución de estructuras cristalinas fueron realizados por la Dra. Natalia Curado, el Dr. Javier Borge (capítulos 1 y 2) y la Dra. Josefina Díez (capítulo 3). Con respecto a la reactividad de los complejos, se hicieron pruebas experimentales y estudios teóricos por medio de cálculos DFT con el fin de obtener evidencias que permitiesen proponer rutas mecanísticas plausibles en la que estas especies estuviesen involucradas. En este punto, es necesario mencionar que una parte de los cálculos DFT fueron llevados a cabo por el Dr. Joaquín López-Serrano o el Prof. Agustí Lledós. Consideraciones Generales [7] La Tesis comienza con una Introducción General, seguida por 4 capítulos que se subdividen en Introducción, Resultados y Discusión, y Referencias. Por último, la Parte Experimental se encuentra al final de este trabajo. Con el objetivo de facilitar la lectura de este trabajo, la bibliografía empleada se encuentra referenciada a pie de página y al final de cada capítulo. Además, con esta Tesis se encuentra adjunto un CD con datos espectroscópicos, estructuras cristalinas y resultados derivados de los estudios computacionales, en caso de que fuese necesaria su consulta. Mientras que la numeración de esquemas y figuras es independiente para cada capítulo, la numeración de los compuestos es común para toda la Tesis (se adjunta hoja plastificada suelta con todos los complejos numerados para una lectura más cómoda). Con objeto de obtener la Mención Internacional en el título de Doctor (RD 99/2011; BOE 20-02-2011, Artículo 15), la mayoría de la Tesis está redactada en inglés, exceptuando las Consideraciones Generales y las Conclusiones, las cuales se encuentran disponibles tanto en inglés como en español. Además, se realizó una estancia de 4 meses en los laboratorios del profesor Christopher C. Cummins en el Instituto Tecnológico de Massachusetts (MIT), en Cambridge (Estados Unidos). Dichos resultados, los cuales no se incluyen en esta memoria, consisten en el empleo de oxoaniones de azufre como materiales de partida para la formación de enlaces C−S. El contenido de la tesis se resume de la siguiente forma: en el Capítulo 1, se describe la síntesis y caracterización de complejos de níquel estabilizados por un ligando PBP, su influencia en la activación del CO2 y su posterior transformación en derivados de bis(silil)acetal. Adicionalmente, se discuten estudios experimentales y teóricos que conllevan a la propuesta de un mecanismo para la hidrosililación catalítica del CO2. Consideraciones Generales [8] En el Capítulo 2, se lleva a cabo la síntesis y caracterización de borilos de níquel, utilizando para ello la misma plataforma molecular que en el Capítulo 1. Asimismo, se discute la estructura de uno de ellos, y se demuestra el empleo de estas especies en procesos catalíticos de borilación deshidrogenante de estirenos. En el Capítulo 3, se efectúa la detección y caracterización de complejos σ-silano de platino catiónicos estabilizados por ligandos NHC. A su vez, se modulan dichos ligandos para optimizar la interacción complejo-silano, y así, poder aislar varias especies de esta clase. De manera adicional, se demuestra su papel como intermedios reactivos en la formación de enlaces C−Si y/o Pt−Si tanto de manera experimental como por medios computacionales. En el Capítulo 4, se llevan a cabo experimentos catalíticos de hidrosililación de dióxido de carbono al nivel de ácido fórmico y de acoplamiento deshidrogenante de silanos y aminas para dar mono- y disilazanos, utilizando para ello algunos complejos de platino estudiados en el Capítulo 3. Nuevamente, se llevan a cabo estudios experimentales y teóricos que permitan proponer un mecanismo en la formación de los enlaces Si−N. Parte de los resultados comentados en este trabajo se encuentra aún sin publicar, aunque hay otros fragmentos que están publicados y pueden consultarse en diversas revistas científicas. A continuación se detallan sus datos, a la vez que se indica la información pertinente a otras publicaciones no relacionadas con este trabajo, pero que se publicaron durante la realización de esta Tesis Doctoral. Consideraciones Generales [9] Publicaciones Artículos relacionados con esta Tesis Doctoral P. Ríos, N. Curado, J. López-Serrano, A. Rodríguez. Selective reduction of carbon dioxide to bis(silyl)acetal catalyzed by a PBP-supported nickel complex. Chem. Commun., 2016, 52, 2114-2117. DOI: 10.1039/C5CC09650B P. Ríos, A. Rodríguez, J. López-Serrano. Mechanistic Studies on the Selective Reduction of CO2 to the Aldehyde Level by a Bis(phosphino)boryl (PBP)- Supported Nickel Complex. ACS Catal., 2016, 6, 5715-5723. DOI: 10.1021/acscatal.6b01715 P. Ríos, J. Díez, J. López-Serrano, A. Rodríguez, S. Conejero. Cationic Platinum(II) σ-SiH Complexes in Carbon Dioxide Hydrosilation. Chem. Eur. J., 2016, 22, 16791- 16795. DOI:10.1002/chem.201603524 P. Ríos, M. Roselló-Merino, O. Rivada-Wheelaghan, J. Borge, J. López-Serrano, S. Conejero. Selective catalytic synthesis of amino-silanes at part-per million catalyst loadings. Chem. Commun., 2018, 54, 619-622. DOI: 10.1039/C7CC08530C P. Ríos, H. Fouilloux, P. Vidossich, J. Díez, A. Lledós, S. Conejero. Isolation of a Cationic Platinum(II) σ-Silane complex. Angew. Chem. Int. Ed., 2018, 57, 3217- 3221. DOI: 10.1002/anie.201712791 P. Ríos, H. Fouilloux, J. Díez, P. Vidossich, A. Lledós, S. Conejero. Manuscript in preparation. σ-Silane Platinum(II) Complexes as Intermediates in C−Si Bond Coupling Processes. Manuscript in preparation. Artículos no relacionados con el contenido de esta Tesis Doctoral P. Rios, T. S. Carter, T. J. Mooibroek, M. P. Crump, M. Lisbjerg, M. Pittelkow, N. T. Supekar, G.-J. Boons, A. P. Davis. Synthetic Receptors for the High-Affinity Consideraciones Generales [10] Recognition of O-GlcNAc Derivatives. Angew. Chem. Int. Ed., 2016, 55, 3387-3392. DOI: 10.1002/anie.201510611 P. Ríos, T. J. Mooibroek, T. S. Carter, C. Williams, M. P. Crump, A. P. Davis. Enantioselective Carbohydrate Recognition by Synthetic Lectins in Water. Chem. Sci., 2017, 8, 4056-4061. DOI: 10.1039/C6SC05399H M. B. Geeson, P. Ríos, W. J. Transue, C. C. Cummins. Oxoanions of Phosphorus and Sulfur as Entryways to Carbon−Element Bond Formation. Manuscript in preparation. Patentes no relacionadas con el contenido de esta Tesis Doctoral M. B. Geeson, P. Ríos, W. J. Transue, C. C. Cummins. Method for producing trichlorosilylsulfide anion and its use in the preparation of organosulfur compounds. Application number: 62/697,037. Abbreviations [11] Abbreviations η Hapticity κ Denticity 9-BBN 9-borabicyclo[3.3.1]nonane APCI Atmospheric Pressure Chemical Ionization BAr4F Tetrakis(3,5-bis(trifluoromethyl)phenyl)borate BCF Tris(pentafluorophenyl)borane BCP Bond Critical Point BDE Bond dissociation energy BICAAC Bicyclic Alkyl Amino Carbene Bu Butyl CAAC Cyclic Alkyl Amino Carbene Cat Catechol CBC Covalent Bond Classification COD 1,5-Cyclooctadiene Cp Cyclopentadienyl Cp* Pentamethylcyclopentadienyl Cy Cyclohexyl DCM Dichloromethane DFT Density Functional Theory DHB Dehydrogenative borylation Dme 1,2-dimethoxyethane DMF N,N-dimethylformamide ESI Electrospray Ionization Et Ethyl Abbreviations [12] FAB Fast Atom Bombardment FLP Frustrated Lewis Pair GC Gas Chromatography HMDS Hexamethyldisilazide HOMO Highest Occupied Molecular Orbital HPLC High Performance Liquid Chromatography HRMS High Resolution Mass Spectrometry IMes 1,3-Bis(2,4,6-trimethylphenyl)imidazol-2-ylidene IMes* 4,5-Dimethyl-1,3-bis-(2,4,6-trimethylphenyl)-imidazol- 2-ylidene IiPr 1,3-diisopropyl-4,5-dimethylimidazol-2-ylidene IPr 1,3-bis-(2,6-diisopropylphenyl)-imidazol-2-ylidene IR Infra-red IRC Intrinsic Reaction Coordinate ItBu 1,3-di-tert-butylimidazol-2-ylidene KIE Kinetic Isotope Effect LA Lewis Acid LMO Localized Molecular Orbitals LUMO Lowest Unoccupied Molecular Orbital Me Methyl Mes Mesityl NBE 2-norbornene NBO Natural Bond Orbital NHC N-Heterocyclic Carbene ORTEP Oak Ridge Thermal Ellipsoid Plot PBP Bis(posphino)boryl Abbreviations [13] PES Potential Energy Surface Ph Phenyl Pin Pinacol Pr Propyl QTAIM Quantum Theory of Atoms In Molecules RCP Ring Critical Point r.t. Room temperature SN2 Bimolecular nucleophilic substitution TEP Tolman Electronic Parameter THF Tetrahydrofuran TM Transition metal TOF Turnover frequency TON Turnover number TS Transition state %Vbur Percent Buried Volume VBE Vinyl Boronate Ester VSEPR Valence Shell Electron Pair Repulsion WBI Wiberg Bond Index ZPE Zero-point energy NMR abbreviations δ Chemical Shift COSY Correlation Spectroscopy HMBC Heteronuclear Multiple Bond Correlation HMQC Heteronuclear Multiple-Quantum Correlation General Introduction [20] these two opposite scenarios. Fortunately, numerous pieces of information are available at our disposal for describing this kind of bonds, derived from experimental (NMR spectroscopy, X-Ray crystallography) and theoretical data (Topological analysis, NBO and/or molecular orbital studies). 7 Figure 1. a) Molecular orbital diagrams displaying the different metal−ligand bonding situations according to the Covalent Bond Classification (adapted from reference 3). b) Limiting bonding scenarios for a M−Z bond (adapted from references 4 and 7). Figure 1a exemplifies a scenario where a single pair of electrons is shared between the metal and the ligand. However, ligands can interact with a metal atom through a combination of the 3 situations described above (e.g. a benzene ring and a cyclopentadienyl anion can behave as L3 and a L2X ligands, respectively, Figure 2A). Additionally, in some cases the metal−ligand bond can 7 G. Bouhadir, D. Bourissou in The Chemical Bond III. Structure and Bonding (2006), vol. 171. Springer, Cham. Chapter 5: Coordination of Lewis Acids to Transition Metals: Z-Type Ligands. General Introduction [21] be strengthened because of electron density donation from the metal through its filled d orbitals (if they are high enough in energy), provided the already bound ligands possess energetically accessible π* orbitals of the right symmetry. This phenomenon is termed back-bonding, and it gives multiple bond character to the metal−ligand fragment (Figure 2B). 8 This should not be mistaken with the M−Z bonding situation, where a σ-bond is formed. In back-bonding, the orbitals involved possess π character. Figure 2. a) Different coordination modes depending on the nature of the ligand. b) Orbital description of the back-bonding phenomenon in a CO ligand (adapted from references 2 and 8). Ligand Design The reactivity and stability of the transition metal complexes is greatly influenced by the nature of the ligands surrounding the metal atom. In fact, there is a wide variety of parameters of the ligands (involving steric and/or electronic properties, in most of the cases) that can be modified and modulated so the resulting complexes possess certain properties. In this way, as it will be seen 8 R. J. Lundgren, M. Stradiotto in Ligand Design in Metal Chemistry: Reactivity and Catalysis. 1st Edition (2016). John Wiley & Sons, Ltd. Chapter 1: Key Concepts in Ligand Design: An Introduction. General Introduction [22] throughout this thesis, ligand design can be used to produce complexes capable of achieving the activation of a priori inert molecules (Chapter 1), catalytically obtaining a specific product in a selective manner (Chapters 1 and 4), or detecting and isolating intermediates whose nature is elusive and highly reactive (Chapters 2 and 3). The following sections will describe the ligands employed during this thesis. At the same time, such description will be leveraged to explain some of the ligand properties (highlighted in bold) that are fundamental in the design process and that were taken into account when synthesizing the corresponding transition metal complexes. Bis(phosphino)boryl (PBP) ligand General features The ligand utilized in the first 2 chapters of this thesis is the bis(phosphino)boryl (PBP) ligand, which is shown in Figure 3 in its hydroborane form (C6H4{N(CH2PR2)}2BH). Figure 3. General structure of the bis(phosphino)boryl (PBP) ligand in its hydroborane form. Originally developed by Yamashita and Nozaki in 2009, 9 the first appreciable feature that comes to mind at first glance is the presence of various donor atoms such as the phosphine groups. This makes this molecule a 9 Y. Segawa, M. Yamashita, K. Nozaki, J. Am. Chem. Soc., 2009, 131, 9201-9203. General Introduction [23] polydentate ligand. Thus, its coordination to a metal group will take advantage of the chelate effect, which consists of the increased stability of polydentate ligands in comparison to monodentate ones when they are coordinated to a metal atom. This can be easily illustrated with the example of the displacement of ammonia (monodentate) ligands by the coordination of ethylenediamine (polydentate, chelating ligand), as shown in Scheme 1. It can be clearly seen that this displacement changes the number of free molecules from 4 to 7. This great increase in entropy makes the process thermodynamically favourable.2 Scheme 1. Substitution of ammonia by ethylenediamine, illustrating the chelate effect (adapted from reference 2). The designers of the PBP ligand observed activation of the B−H bond upon coordination to the transition metal center.9 This reaction converts this molecule into an anionic tridentate ligand with adjacent donor atoms (Scheme 2). 10 For this reason, the PBP fragment is a pincer ligand. This type of ligands has been appearing in the literature for around 40 years and they are widely used in homogeneous catalysis, given that pincer-based transition metal complexes are capable of efficiently catalyse transformations such as cross-coupling, dehydrogenation or polymerization reactions as well as transforming challenging small molecules like CO2, H2 or O2. 11 10 The authors conclude that they observe the boryl hydride arrangement displayed in Scheme 2, given that NMR, DFT and X-Ray studies rule out a σ-BH coordination mode. 11 a) M. E. van der Boom, D. Milstein, Chem. Rev., 2003, 103, 1759-1792; b) G. van Koten, D. Milstein, Organometallic Pincer Chemistry, 2012, Springer, Heidelberg. General Introduction [24] Scheme 2. Coordination of the PBP molecule to give the first transition metal complex possessing a tridentate boryl ligand (adapted from reference 9). The majority of the pincer ligands described to date contain carbon or nitrogen as the central atom that establishes a σ-bond with the metal. Whereas the presence of boryl ligands in the coordination sphere of transition metal complexes was limited to 1,2-dioxo derivatives such as catechol (Bcat) or pinacol (Bpin), the utilization of trivalent nitrogen to form B−N bonds has opened the door to the possibility of obtaining branched boryl groups, giving rise to pincer ligands like the one used in this thesis.9, 12 However, alternative strategies have been developed to obtain pincer ligands with boryl functionalities in the central position. Some of them include the utilization of the ambiphilic (containing L and Z type ligands) diphosphanyl borane developed by Bourissou et al. in 2006 13 (Scheme 3a), which undergoes C−B oxidative addition in the presence of the iridium dimer [(COE)2IrCl]2 to give complex (PBP)Ir(Ph)(Cl), as demonstrated by the group of Ozerov. 14 In some cases, 2nd row metals like rhodium yield mixtures of PBP compounds (featuring metal−borane and metal−boryl interactions), given the lower tendency of 4d metals to undergo oxidative addition reactions against 5d ones.14, 15 Other procedures take advantage of the rigidity of a 12 H. Braunschweig, M. Colling, Coord. Chem. Rev., 2001, 223, 1-51. 13 S. Bontemps, H. Gornitzka, G. Bouhadir, K. Miqueu, D. Bourissou, Angew. Chem. Int. Ed., 2006, 45, 1611-1614. 14 W-C. Shih, W. Gu, M. C. MacInnis, S. D. Timpa, N. Bhuvanesh, J. Zhou, O. V. Ozerov, J. Am. Chem. Soc., 2016, 138, 2086-2089. 15 W-C. Shih, W. Gu, M. C. MacInnis, D. E. Herbert, O. V. Ozerov, Organometallics, 2017, 36, 1718-1726. General Introduction [25] carborane unit, like the ones shown in Scheme 3b, where the incorporation of the “arms” produces the corresponding tridentate scaffolds. 16 Scheme 3. Alternative methods to obtain transition metal complexes stabilized by boryl-based pincer ligands (adapted from references 13 and 14). The many reasons that explain the success of pincer ligands are related to the high level of steric and electronic modulability that these ligands can impart on the metal complexes, leading to numerous opportunities in the field of catalysis. 17 They are summarized in Figure 4, and some of them will be described 16 a) A. M. Spokoyny, M. G. Reuter, C. L. Stern, M. A. Ratner, T. Seiderman, C. A. Mirkin, J. Am. Chem. Soc., 2009, 131, 9482-9483; b) M. E. El-Zaria, H. Arii, H. Nakamura, Inorg. Chem., 2011, 50, 4149-4161; c) B. J. Eleazer, M. D. Smith, A. A. Popov, D. V. Peryshkov, J. Am. Chem. Soc., 2016, 138, 10531-10538. 17 E. Peris, R. H. Crabtree, Chem. Soc. Rev., 2018, 47, 1959-1968. General Introduction [26] in detail in this section, given their implications in the results of this work. One of the most important features of these frameworks is the stabilization they confer to the metal complexes, preventing metal leaching during the catalytic processes, which sometimes require harsh temperature and pressure conditions. 18 Figure 4. General structure of a pincer ligand bound to a metal atom and description of the parameters that can be modified upon modification of its structure (adapted from reference 17). Boryl groups: electronic properties Boryl ligands possess the general formula –BR2, and their main feature is the presence of an anionic, sp2-hybridized boron atom. These compounds will behave as X ligands upon coordination to a transition metal center. As the majority of the ligands, this will take place through a σ-bond (as depicted in Figure 5a), yet the presence of an a priori empty p orbital in the boron atom makes the boryl group a π-acceptor ligand, and may lead to back-bonding phenomena (Figure 5b). However, this will depend on the R substituents on the 18 M. A. W. Lawrence, K-A. Green, P. N. Nelson, S. C. Lorraine, Polyhedron, 2018, 143, 11-27. General Introduction [27] boron atom, since π-donor groups like oxygen or nitrogen can also donate electron density to the boron p orbital (Figure 5c), increasing the energy of the boryl acceptor π* orbital and diminishing the degree of back-bonding from the metal. 19 For this reason, the contribution of back-bonding to the overall M−B bond density is small (even in B(C6F5)2 compounds, Table 1), 20 with values of around 15-20% at the most. 21 Although this might be seen as a competition between both parts for the p orbital of boron, the truth is that there is a synergy between them, because the lack of electron density on B must be compensated from the substituents (if the metal is not able to undergo back-bonding) or vice versa. Thus, both fragments will play a role in the stability of the boryl ligand, which will in turn have a great influence on the reactivity of the complex. 22 Figure 5. Orbital interactions during the metal−boryl coordination. a) σ-bonding. b) π backbonding. c) Competitive π interactions between the metal and the R groups (adapted from reference 19). 19 D. L. Kays, S. Aldridge in Contemporary Metal Boron Chemistry I, 2008. Springer-Berlag Berlin Heidelberg. Chapter 2: Transition metal boryl complexes. Pages 29-122. 20 A. Al-Fawaz, S. Aldridge, D. L. Coombs, A. A. Dickinson, D. J. Willock, L-L. Ooi, M. E. Light, S. J. Coles, M. B. Hursthouse, Dalton Trans., 2004, 4030-4037. 21 A. A. Dickinson, D. J. Willock, R. J. Calder, S. Aldridge, Organometallics, 2002, 21, 1146- 1157. 22 G. J. Irvine, M. J. G. Lesley, T. B. Marder, N. C. Norman, C. R. Rice, E. G. Robins, W. R. Roper, G. R. Whittell, L. J. Wright, Chem. Rev., 1998, 98, 2685- 2722. General Introduction [28] Complex (η5-C5H5)Fe(CO)2BR2 R Calculated d(Fe−B)/Åa Orbital contribution to Fe−B bond (%) σ π C6F5 2.006 (1.965) 81.9 18.0 C6H5 2.107 (2.034) 90.4 9.5 H 1.988 84.1 15.8 F 2.010 94.0 6.0 Cl 2.008 (1.942) 82.1 17.8 Table 1. Influence of the boron substituents on the Fe−B bond in complex (η5-C5H5)Fe(CO)2BR2. a Crystallographically determined distances in parentheses (adapted from reference 20). Another feature of the boron atom that must be taken into account when designing complexes bearing boryl functionalities is related to its electropositivity (χ = 2.0) 23 and its strong σ-donation ability. Because boryl groups are powerful σ-donors, they also are among the strongest trans influence ligands. This means that a ligand with a strong trans influence will weaken the metal−ligand bond of a ligand located trans to it because both ligands are sharing the same metal orbital for establishing their corresponding bonds. In other words, strong trans influence ligands “monopolize” the metal orbital due to their σ-donation ability, labilizing the metal−ligand bond trans to it. 24 , 25 This is a thermodynamic property, and the consequences of this labilization can be observed by elongation of bond lengths through X-Ray crystallography, by a decrease in certain coupling constants through NMR spectroscopy or the reduction of some stretching frequencies through IR spectroscopy. DFT methods 23 N. N. Greenwood, A. Earnshaw, Chemistry of the Elements, 2nd ed. Butterworth- Heinemann: Woburn, MA, 1997. 24 K. M. Anderson, A. G. Orpen, Chem. Commun., 2001, 2682-2683. 25 This is a qualitative description of the concept of trans influence. For quantitative parameters that contribute to this property, see: J. K. Burdett, T. A. Albright, Inorg. Chem., 1979, 18, 2112-2120. General Introduction [29] are accurate enough to reproduce these structural changes and to correlate them to the trans influence of the ligands involved. In fact, Marder et al. ordered a large amount of ligands according to their trans influence, concluding that boryl groups were on top of this scale, way above hydride and alkyl ligands for example. 26 For the reasons mentioned above, complexes bearing boryl functionalities will likely be highly reactive, which opens the door to the possibility of activating inert and thermodynamically stable species. Steric properties of pincer ligands Steric effects can be determinant when designing a pincer complex due to the fact that they can either protect the metal so much that the compound becomes unreactive or expose it to an extent that the molecule turns out unstable and very reactive. Thus, an intermediate situation would be ideal to achieve an appropriate robustness/activity proportion. Modification of the ancillary ligands (Figure 4, highlighted in blue) is a straightforward method to do this. Phosphine groups are extensively used in transition metal chemistry, and their employment as ancillary ligands in pincer chemistry is extremely common.11b This is due to their wide availability and their easy modification, which allows the introduction of phosphine groups of many different sizes into the pincer scaffold. As a matter of fact, there are parameters which classify such species according to their steric volume, so the resulting cavity in the transition metal complexes can be predicted and designed to some extent. A good example of these parameters is the Tolman cone angle (θ), which is the apex angle of a cylindrical cone centered at the metal of a M(PR3) fragment that touches the van 26 J. Zhu, Z. Lin, T. B. Marder, Inorg. Chem., 2005, 44, 9384-9390. General Introduction [36] Scheme 7. a) and c) Examples of ligand cooperativity exhibited by PBP pincer ligands. b) General scheme of ligand-mediated heterolytic activation of small molecules (adapted from references 37a, 38 and 39). In the first half of this thesis, the outstanding ability of nickel to activate inert substrates 40 in combination with the great reactivity of boryl ligands will be tamed by making use of some of the concepts of ligand design described above 40 a) S. Z. Tasker, E. A. Standley, T. F. Jamison, Nature, 2014, 509, 299-309; b) V. P. Ananikov, ACS Catal., 2015, 5, 1964-1971. General Introduction [37] so as to produce stable complexes or intermediates that are able to effectively perform diverse catalytic reactions. N-heterocyclic carbenes (NHCs) Initial studies Some pioneer works in the chemistry of stable carbenes started to develop around the 1960s with the work of Wanzlick, who claimed that imidazol- 2-ylidenes were intermediates in the formation of tetraaminofulvenes and that there was an equilibrium between such species (Wanzlick equilibrium, Scheme 8a). 41 However, one of the first experimental evidences was provided in 1988 by the group of Bertrand, who isolated the first (phosphino)(silyl)carbene shown in Scheme 8b. 42 A few years later, Arduengo and co-workers synthesized the first isolable N-heterocyclic carbene thanks to the steric stabilization of adamantyl groups (Scheme 8c). 43 The stability of this molecule (whose integrity remained untouched after 7 years at room temperature dissolved in THF under a CO atmosphere) 44 demonstrated that these species could indeed be very stable depending on the surroundings around the carbene moiety. These breakthroughs were the beginning of a period, which spans until today, in which the chemistry of NHCs has advanced tremendously, transforming these compounds from mere laboratory curiosities to useful tools that are applied as 41 H. W. Wanzlick, Angew. Chem. Int. Ed., 1962, 1, 75-80. This equilibrium remained controversial for some years (see D. M. Lemal, R. A. Lovald, K. I. Kawano, J. Am. Chem. Soc., 1964, 86, 2518-2519), but subsequent experiments proved Wanzlick right in his statement (a) Y. Liu, P. E. Lindner, D. M. Lemal, J. Am. Chem. Soc., 1999, 121, 10626-10627; b) V. P. W. Böhm, W. A. Herrmann, Angew. Chem. Int. Ed., 2000, 39, 4036-4038). 42 A. Igau, H. Grutzmacher, A. Baceiredo, G. Bertrand, J. Am. Chem. Soc., 1988, 110, 6463- 6466. 43 A. J. Arduengo, R. L. Harlow, M. Kline, J. Am. Chem. Soc., 1991, 113, 361-363. 44 A. J. Arduengo, Acc. Chem. Res., 1999, 32, 913-921. General Introduction [38] catalysts, in the chemical industry or in the stabilization of unusual species or materials. 45 Scheme 8. Initial studies on carbene chemistry (adapted from references 41-43). General features As their own name indicate, NHCs are composed of a heterocycle ring containing a carbene unit adjacent to at least one nitrogen atom. Among these, imidazole-based carbenes (imidazol-2-ylidenes) are the most commonly used and studied.45 The carbene moiety is a neutral fragment with 6 electrons in its valence shell. In a NHC, the lone pair of electrons on the carbon atom is usually located in a sp2 orbital (HOMO), whereas the remaining empty p orbital 45 a) W. A. Herrmann, C. Köcher, Angew. Chem. Int. Ed., 1997, 36, 2162-2187; b) P. de Frémont, N. Marion, S. P. Nolan, Coord. Chem. Rev., 2009, 253, 862-892; c) M. N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature, 2014, 510, 485-496. General Introduction [39] constitutes the LUMO. The energy difference between them (HOMO-LUMO gap, or ES-T) is influenced by several factors that depend on geometrical and electronic parameters. As a consequence, the reactivity of the resulting carbene will also depend on these aspects. For instance, Figure 8 shows that linear arrangements give rise to a carbon atom with a sp hybrid orbital and 2 degenerate p orbitals. 46 Bending the carbene will favour a sp2 hybridization on carbon, as illustrated in the figure. The magnitude of ES-T (the difference between the σ and pπ orbitals) will therefore determine whether the carbene is in a singlet or triplet ground state. 47 Figure 8. Distribution on the frontier orbitals depending on the geometry of the carbene (adapted from reference 46). Apart from geometry, electronic factors also affect the HOMO-LUMO gap. In the specific case of NHCs (imidazol-2-ylidenes as a model species, since these are the carbenes employed throughout this thesis), the presence of two electronegative nitrogen atoms adjacent to the carbene fragment will stabilize the σ orbital by increasing its s character (inductive effects). At the same time, both N atoms possess filled π orbitals with the right symmetry to overlap with the empty pπ orbital of the carbene. Figure 9 shows how this overlap increases even more the value of ES-T (mesomeric effects). For these reasons, NHCs are 46 D. Bourissou, O. Guerret, F. P. Gabbai, G. Bertrand, Chem. Rev., 2000, 100, 39-91. 47 R. Gleiter, R. Hoffmann, J. Am. Chem. Soc., 1968, 90, 5457-5460. General Introduction [40] usually singlet species, and they possess a filled σ non-bonding orbital and an empty pπ orbital.46 This makes NHCs very good σ donors and poor π-acceptors. Nonetheless, this latter aspect is not negligible given that in some cases the metal−carbene π back-bonding interaction contributes up to 30% of the overall bonding energy. 48 In a different case, π-acidic NHC ligands have proven crucial for the selectivity of gold catalysed cycloaddition reactions due to their ability to remove electron density of some intermediates, favouring one mechanistic pathway. 49 Figure 9. Molecular orbitals diagram illustrating mesomeric effects in NHCs (adapted from reference 46). Similarly to pincer ligands, there are several parameters that can be modulated on an N-heterocyclic carbene. In the same way, Figure 10 shows every part of its structure that can be tuned in order to achieve certain stereoelectronic 48 a) X. Hu, Y. Tang, P. Gantzel, K. Meyer, Organometallics, 2003, 22, 612-614; b) X. Hu, I. Castro-Rodriguez, K. Olsen, K. Meyer, Organometallics, 2004, 23, 755-764. 49 M. Alcarazo, T. Stork, A. Anoop, W. Thiel, A. Furstner, Angew. Chem. Int. Ed., 2010, 49, 2542-2546. General Introduction [41] properties.45c As it was done in the previous section, some of these parameters will be key in the design of the complexes synthesized in this thesis and so, they will be described in detail. Figure 10. General structural parameters of imidazol-2-ylidenes (NHC models) (adapted from reference 45c). Electronic properties of NHCs N-heterocyclic carbenes are among the strongest neutral bases and nucleophiles, and they surpass strong bases like DBU or DABCO. As illustrated above, these properties can be easily adjusted by simple modification of the substituents on the nitrogen atoms, which is very convenient for organocatalytic reactions, for example. 50 As stated above, the σ-donor and π-acceptor properties of a NHC depend on their relative energy of their frontier orbitals. However, it is not straightforward to quantify these capabilities separately, as they are not 50 A. Levens, F. An, M. Breugst, H. Mayr, D. W. Lupton, Org. Lett., 2016, 18, 3566-3569. General Introduction [42] completely independent. Nonetheless, numerous experimental methods have been developed so as to determine their overall donor-acceptor abilities. 51 One of these methods is the Tolman electronic parameter (TEP), which gives the donor-acceptor capability of a ligand according to the A1 CO stretching of a [Ni(CO)3(NHC)] complex. Electron-rich carbenes will donate electron density to the nickel atom, which will in turn undergo π back-bonding with the empty π* orbitals of the CO ligands, strengthening the M−C bonds and weakening the C−O ones. This will have an impact on the CO stretching frequency (the more electron donating carbene, the smaller CO wavenumber in the IR spectrum), which can be correlated to the donor-acceptor properties of the ligand. 52 Figure 11 shows the CO stretch for Ni complexes containing some of the NHCs used in this thesis. It is important to note that this parameter has several disadvantages such as the solvent dependence or the small range of wavenumbers (~40 cm-1) 53 in which the values of the CO vibrations for [Ni(CO)3(NHC)] fall. Besides, the toxicity of [Ni(CO)4] led to the utilization of different complexes so as to establish a TEP scale, like cis-[IrCl(CO)2(NHC)] or cis-[RhCl(CO)2(NHC)].51, 54 51 H. V. Huynh, Chem. Rev., 2018, 118, 9457-9492. 52 C. A. Tolman, J. Am. Chem. Soc., 1970, 92, 2953-2956. 53 D. Munz, Organometallics, 2018, 37, 275-289. 54 D. J. Nelson, S. P. Nolan, Chem. Soc. Rev., 2013, 42, 6723-6753. General Introduction [43] Figure 11. Top: general scheme of the influence of the electron density of a NHC ligand on the CO stretching frequency. Bottom: Tolman electronic parameters of some NHC ligands employed in this thesis. Values obtained from reference 54. Some other authors have been able to quantify the π acidity of the NHC ligands by means of NMR techniques, so both σ and π contributions can be known by comparison with other scales. Whereas Bertrand et al. developed a scale based on 31P NMR chemical shifts of carbene-phosphinidene adducts, 55 the group of Ganter established a different one based on the 77Se NMR chemical shift of selenium carbene compounds. 56 The idea of these scales lies on the deshielding that the 31P or 77Se undergo upon π back-bonding to the carbene moiety: the higher degree of backbonding, the lower chemical field in the NMR spectrum (Scheme 9). Apart from the deconvolution contribution, another advantage of these methods is the width of the different scales in comparison to that obtained for the TEP values, since the 31P and the 77Se scales cover a range of around 150 and 850 ppm, respectively. 55 O. Back, M. Henry-Ellinger, C. D. Martin, D. Martin, G. Bertrand, Angew. Chem. Int. Ed., 2013, 52, 2939-2943. 56 A. Liske, K. Verlinden, H. Buhl, K. Schaper, C. Ganter, Organometallics, 2013, 32, 5269-5272. General Introduction [44] The aforementioned experimental methods are the most common ones to study the electronic properties of carbene molecules. However, there exist several others such as pKa values, redox potentials or the Huynh’s electronic parameter (HEP), among others.51 Scheme 9. Summary of the NMR scales developed by Bertrand and Ganter for the determination of the π acidity of carbene ligands (adapted from reference 53). Steric properties of NHCs The kinetic stability of the NHCs is greatly influenced by the nature of the substituents on the nitrogen atoms, which has a crucial role in their reactivity patterns or coordination modes in a transition metal complex. However, choosing the right carbene for a specific purpose is not straightforward. The sometimes asymmetric character of NHCs in combination with their conformational flexibility makes difficult to quantify with a single number their General Introduction [45] steric properties in the same way as many phosphines are defined by their Tolman cone angle. This is also due to the fact that phosphine substituents point away from the metal, as illustrated in Figure 6, whereas N-heterocyclic carbenes form a cavity between the metal atom and the substituent on nitrogen (Figure 12, space between the green cones and the grey circle). 57 Thus, Nolan, Cavallo et al. defined the percent buried volume (%Vbur) as the percentage of volume occupied by the NHC ligand in a sphere centered at the metal atom. This parameter is usually calculated by fixing a M−Ccarbene distance of 2 Å and a sphere radius of 3 or 3.5 Å. 58 Figure 12. Representation of the steric impact of a NHC ligand. The easiest way to calculate the percent buried volume is to utilize the SambVca software, which is a free online tool developed by the group of Cavallo that performs the calculation of %Vbur on CIF or XYZ files. This algorithm can also be employed for calculating the steric impact of other groups, such as phosphines or cyclopentadienyl-based ligands. 59 Although this gives an idea of how bulky a 57 A. Gómez-Suárez, D. J. Nelson, S. P. Nolan, Chem. Commun., 2017, 53, 2650-2660. 58 A. C. Hillier, W. J. Sommer, B. S. Yong, J. L. Petersen, L. Cavallo, S. P. Nolan, Organometallics, 2003, 22, 4322-4326. 59 A. Poater, B. Cosenza, A. Correa, S. Giudice, F. Ragone, V. Scarano, L. Cavallo, Eur. J. Inorg. Chem., 2009, 1759-1766. General Introduction [52] Figure 14. Examples of several families of carbenes different to NHCs (adapted from references 76 and 78). NHC-stabilized Pt(II) complexes Previously in our group, the tunable character of NHCs was leveraged so several Pt(II) complexes were synthesized and able to fulfil some specific needs, taking into account many of the concepts detailed in this introduction. In this way, carbenes bearing mesityl groups made possible intermolecular stoichiometric C−H activation reactions of benzene, toluene and fluoroarene derivatives (Scheme 12). Experimental and computational studies led to the conclusion that careful design of the steric and electronic properties of the complexes were crucial for this process to take place, since bulkier carbenes like ItBu did not give any C−H activation product. In addition, subtle electronic effects were also important for the success of this reaction, given that only IMes* ligand afforded some Pt−aryl compounds, unlike less basic IMes ligand (traces were observed in this case). 79 79 O. Rivada-Wheelaghan, M. A. Ortuño, J. Díez, A. Lledós, S. Conejero, Angew. Chem. Int. Ed., 2012, 51, 3936-3939. General Introduction [53] Scheme 12. C−H activation reactions of NHC-stabilized Pt(II) complexes previously explored in our group (adapted from reference 79). On the other hand, these type of complexes made possible the stabilization and isolation of a rare and elusive cationic mononuclear Pt(III) alkyl species by making use of IPr ligands. Addition of 0.5 equivalents of Br2 to a dichloromethane solution of the cationic cyclometalated IPr complex gave rise to the Pt(III) derivative (Scheme 13), which could be characterized by X-Ray diffraction studies. Interestingly, the uncommon seesaw structure found around the platinum center was due to electronic reasons, and not steric hindrance. In fact, computational studies shed light about this phenomenon, and concluded that the trans influence of the ligand trans to the bromide ligand was responsible for the observed geometry. 80 80 O. Rivada-Wheelaghan, M. A. Ortuño, J. Díez, S. E. García-Garrido, C. Maya, A. Lledós, S. Conejero, J. Am. Chem. Soc., 2012, 134, 15261-15264. General Introduction [54] Scheme 13. Synthesis of a NHC-stabilized Pt(III) complex (adapted from reference 80). All the examples mentioned above in this introduction reflect the importance of ligand design in transition metal complexes, as their chemical behaviour is greatly influenced by the properties conferred by these species. In the second half of this thesis, different NHCs will be used in the synthesis of Pt(II) complexes for the isolation of elusive intermediates and the development of catalytically active and selective species. General Introduction [55] References 1 F. A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry, 3rd Edition (1972). Wiley-Interscience. 2 R. H. Crabtree, The Organometallic Chemistry of the Transition Metals, 4th Edition (2005). Wiley-Interscience. 3 M. L. H. Green, G. Parkin, J. Chem. Educ., 2014, 91, 807-816. 4 A. Amgoune, D. Bourissou, Chem. Commun., 2011, 47, 859-871. 5 A. F. Hill, Organometallics, 2006, 25, 4741-4743. 6 G. Parkin, Organometallics, 2006, 25, 4744-4747. 7 G. Bouhadir, D. Bourissou in The Chemical Bond III. Structure and Bonding (2006), vol. 171. Springer, Cham. Chapter 5: Coordination of Lewis Acids to Transition Metals: Z-Type Ligands. 8 R. J. Lundgren, M. Stradiotto in Ligand Design in Metal Chemistry: Reactivity and Catalysis. 1st Edition (2016). John Wiley & Sons, Ltd. Chapter 1: Key Concepts in Ligand Design: An Introduction. 9 Y. Segawa, M. Yamashita, K. Nozaki, J. Am. Chem. Soc., 2009, 131, 9201-9203. 10 The authors conclude that they observe the boryl hydride arrangement displayed in Scheme 2, given that NMR, DFT and X-Ray studies rule out a σ-BH coordination mode. 11 a) M. E. van der Boom, D. Milstein, Chem. Rev., 2003, 103, 1759-1792; b) G. van Koten, D. Milstein, Organometallic Pincer Chemistry, 2012, Springer, Heidelberg. 12 H. Braunschweig, M. Colling, Coord. Chem. Rev., 2001, 223, 1-51. 13 S. Bontemps, H. Gornitzka, G. Bouhadir, K. Miqueu, D. Bourissou, Angew. Chem. Int. Ed., 2006, 45, 1611-1614. 14 W-C. Shih, W. Gu, M. C. MacInnis, S. D. Timpa, N. Bhuvanesh, J. Zhou, O. V. Ozerov, J. Am. Chem. Soc., 2016, 138, 2086-2089. 15 W-C. Shih, W. Gu, M. C. MacInnis, D. E. Herbert, O. V. Ozerov, Organometallics, 2017, 36, 1718-1726. General Introduction. References [56] 16 a) A. M. Spokoyny, M. G. Reuter, C. L. Stern, M. A. Ratner, T. Seiderman, C. A. Mirkin, J. Am. Chem. Soc., 2009, 131, 9482-9483; b) M. E. El-Zaria, H. Arii, H. Nakamura, Inorg. Chem., 2011, 50, 4149-4161; c) B. J. Eleazer, M. D. Smith, A. A. Popov, D. V. Peryshkov, J. Am. Chem. Soc., 2016, 138, 10531-10538. 17 E. Peris, R. H. Crabtree, Chem. Soc. Rev., 2018, 47, 1959-1968. 18 M. A. W. Lawrence, K-A. Green, P. N. Nelson, S. C. Lorraine, Polyhedron, 2018, 143, 11-27. 19 D. L. Kays, S. Aldridge in Contemporary Metal Boron Chemistry I, 2008. Springer-Berlag Berlin Heidelberg. Chapter 2: Transition metal boryl complexes. Pages 29-122. 20 A. Al-Fawaz, S. Aldridge, D. L. Coombs, A. A. Dickinson, D. J. Willock, L-L. Ooi, M. E. Light, S. J. Coles, M. B. Hursthouse, Dalton Trans., 2004, 4030-4037. 21 A. A. Dickinson, D. J. Willock, R. J. Calder, S. Aldridge, Organometallics, 2002, 21, 1146-1157. 22 G. J. Irvine, M. J. G. Lesley, T. B. Marder, N. C. Norman, C. R. Rice, E. G. Robins, W. R. Roper, G. R. Whittell, L. J. Wright, Chem. Rev., 1998, 98, 2685- 2722. 23 N. N. Greenwood, A. Earnshaw, Chemistry of the Elements, 2nd ed. Butterworth- Heinemann: Woburn, MA, 1997. 24 K. M. Anderson, A. G. Orpen, Chem. Commun., 2001, 2682-2683. 25 This is a qualitative description of the concept of trans influence. For quantitative parameters that contribute to this property, see: J. K. Burdett, T. A. Albright, Inorg. Chem., 1979, 18, 2112-2120. 26 J. Zhu, Z. Lin, T. B. Marder, Inorg. Chem., 2005, 44, 9384-9390. 27 C. A. Tolman, Chem. Rev., 1977, 77, 313-348. 28 V. Pandarus, D. Zargarian, Organometallics, 2007, 26, 4321-4334. 29 A. B. Salah, D. Zargarian, Dalton Trans., 2011, 40, 8977-8985. 30 K. Tanoue, M. Yamashita, Organometallics, 2015, 34, 4011-4017. 31 Y. Segawa, M. Yamashita, K. Nozaki, Organometallics, 2009, 28, 6234-6242. 32 J. R. Khusnutdinova, D. Milstein, Angew. Chem. Int. Ed., 2015, 54, 12236-12273. General Introduction [57] 33 S. W. Kohl, L. Weiner, L. Schwartsburd, L. Konstantinovski, L. J. W. Shimon, Y. Ben-David, M. A. Iron, D. Milstein, Science, 2009, 324, 74-77. 34 P. M. P. García, P. Ren, R. Scopelliti, X. Hu, ACS Catal., 2015, 5, 1164-1171. 35 O. Vechorkin, D. Barmaz, V. Proust, X. Hu, J. Am. Chem. Soc., 2009, 131, 12078- 12079. 36 T. J. Schmeier, G. E. Dobereiner, R. H. Crabtree, N. Hazari, J. Am. Chem. Soc., 2011, 133, 9274-9277. 37 a) W. H. Harman, J. C. Peters, J. Am. Chem. Soc., 2012, 134, 5080-5082; b) M. Devillard, G. Bouhadir, D. Bourissou, Angew. Chem. Int. Ed., 2015, 54, 730-732. 38 G. R. Owen, Chem. Commun., 2016, 52, 10712-10726. 39 T-P. Lin, J. C. Peters, J. Am. Chem. Soc., 2014, 136, 13672-13683. 40 a) S. Z. Tasker, E. A. Standley, T. F. Jamison, Nature, 2014, 509, 299-309; b) V. P. Ananikov, ACS Catal., 2015, 5, 1964-1971. 41 H. W. Wanzlick, Angew. Chem. Int. Ed., 1962, 1, 75-80. This equilibrium remained controversial for some years (see D. M. Lemal, R. A. Lovald, K. I. Kawano, J. Am. Chem. Soc., 1964, 86, 2518-2519), but subsequent experiments proved Wanzlick right in his statement (a) Y. Liu, P. E. Lindner, D. M. Lemal, J. Am. Chem. Soc., 1999, 121, 10626-10627; b) V. P. W. Böhm, W. A. Herrmann, Angew. Chem. Int. Ed., 2000, 39, 4036-4038). 42 A. Igau, H. Grutzmacher, A. Baceiredo, G. Bertrand, J. Am. Chem. Soc., 1988, 110, 6463-6466. 43 A. J. Arduengo, R. L. Harlow, M. Kline, J. Am. Chem. Soc., 1991, 113, 361-363. 44 A. J. Arduengo, Acc. Chem. Res., 1999, 32, 913-921. 45 a) W. A. Herrmann, C. Köcher, Angew. Chem. Int. Ed., 1997, 36, 2162-2187; b) P. de Frémont, N. Marion, S. P. Nolan, Coord. Chem. Rev., 2009, 253, 862-892; c) M. N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature, 2014, 510, 485-496. 46 D. Bourissou, O. Guerret, F. P. Gabbai, G. Bertrand, Chem. Rev., 2000, 100, 39- 91. 47 R. Gleiter, R. Hoffmann, J. Am. Chem. Soc., 1968, 90, 5457-5460. General Introduction. References [58] 48 a) X. Hu, Y. Tang, P. Gantzel, K. Meyer, Organometallics, 2003, 22, 612-614; b) X. Hu, I. Castro-Rodriguez, K. Olsen, K. Meyer, Organometallics, 2004, 23, 755- 764. 49 M. Alcarazo, T. Stork, A. Anoop, W. Thiel, A. Furstner, Angew. Chem. Int. Ed., 2010, 49, 2542-2546. 50 A. Levens, F. An, M. Breugst, H. Mayr, D. W. Lupton, Org. Lett., 2016, 18, 3566- 3569. 51 H. V. Huynh, Chem. Rev., 2018, 118, 9457-9492. 52 C. A. Tolman, J. Am. Chem. Soc., 1970, 92, 2953-2956. 53 D. Munz, Organometallics, 2018, 37, 275-289. 54 D. J. Nelson, S. P. Nolan, Chem. Soc. Rev., 2013, 42, 6723-6753. 55 O. Back, M. Henry-Ellinger, C. D. Martin, D. Martin, G. Bertrand, Angew. Chem. Int. Ed., 2013, 52, 2939-2943. 56 A. Liske, K. Verlinden, H. Buhl, K. Schaper, C. Ganter, Organometallics, 2013, 32, 5269-5272. 57 A. Gómez-Suárez, D. J. Nelson, S. P. Nolan, Chem. Commun., 2017, 53, 2650- 2660. 58 A. C. Hillier, W. J. Sommer, B. S. Yong, J. L. Petersen, L. Cavallo, S. P. Nolan, Organometallics, 2003, 22, 4322-4326. 59 A. Poater, B. Cosenza, A. Correa, S. Giudice, F. Ragone, V. Scarano, L. Cavallo, Eur. J. Inorg. Chem., 2009, 1759-1766. 60 A. Poater, F. Ragone, R. Mariz, R. Dorta, L. Cavallo, Chem. Eur. J., 2010, 16, 14348-14353. 61 L. Falivene, R. Credentino, A. Poater, A. Petta, L. Serra, R. Oliva, V. Scarano, L. Cavallo, Organometallics, 2016, 35, 2286-2293 (also see https://www.molnac.unisa.it/OMtools/sambvca2.0). 62 Some recent examples include: a) A. Al-Harbi, B. Kriegel, S. Gulati, M. J. Hammond, G. Parkin, Inorg. Chem., 2017, 56, 15271-15284; b) Y. Shikata, R. Yasue, K. Yoshida, Chem. Eur. J., 2017, 23, 16806-16812; c) D. Hueber, M. Teci, E. General Introduction [59] Brenner, D. Matt, J-M. Weibel, P. Pale, A. Blanc, Adv. Synth. Catal., 2018, 360, 2453- 2459. 63 N. Fey, M. F. Haddow, J. N. Harvey, C. L. McMullin, A. G. Orpen, Dalton Trans., 2009, 8183-8196. 64 D. G. Gusev, Organometallics, 2009, 28, 6458-6461. 65 A. Poater, F. Ragone, S. Giudice, C. Costabile, R. Dorta, S. P. Nolan, L. Cavallo, Organometallics, 2008, 27, 2679-2681. 66 D. J. D. Wilson, S. A. Couchman, J. L. Dutton, Inorg. Chem., 2012, 51, 7657-7668. 67 J. Huang, E. D. Stevens, S. P. Nolan, J. L. Petersen, J. Am. Chem. Soc., 1999, 121, 2674-2678. 68 M. Scholl, T. M. Trnka, J. P. Morgan, R. H. Grubbs, Tetrahedron Lett., 1999, 40, 2247-2250. 69 L. Ackermann, A. Fürstner, T. Weskamp, F. J. Kohl, W. A. Herrmann, Tetrahedron Lett., 1999, 40, 4787-4790. 70 M. Scholl, S. Ding, C. W. Lee, R. H. Grubbs, Org. Lett., 1999, 1, 953-956. 71 SIMes stands for 1,3-bis(2,4,6-trimethylphenyl)-4,5-dihydroimidazol-2- ylidene. 72 G. C. Fortman, S. P. Nolan, Chem. Soc. Rev., 2011, 40, 5151-5169. 73 a) D. Enders, O. Niemeier, A. Henseler, Chem. Rev., 2007, 107, 5606-5655; b) D. M. Flanigan, F. Romanov-Michailidis, N. A. White, T. Rovis, Chem. Rev., 2015, 115, 9307-9387. 74 a) Y. Wang, G. H. Robinson, Inorg. Chem., 2014, 53, 11815-11832; b) S. Würtemberger-Pietsch, U. Radius, T. B. Marder, Dalton Trans., 2016, 45, 5880- 5895; c) V. Nesterov, D. Reiter, P. Bag, P. Frisch, R. Holzner, A. Porzelt, S. Inoue, Chem. Rev., 2018, 118, 9678-9842. 75 a) J. Vignolle, T. D. Tilley, Chem. Commun., 2009, 7230-7232; b) C. M. Crudden, J. H. Horton, I. J. Ebralidze, O. V. Zenkina, A. B. McLean, B. Drevniok, Z. She, H-B. Kraatz, N. J. Mosey, T. Seki, E. C. Keske, J. D. Leake, A. Rousina-Webb, G. Wu, Nat. Chem., 2014, 6, 409-414; c) A. V. Zhukhovitskiy, M. J. MacLeod, J. A. Johnson, General Introduction. References [60] Chem. Rev., 2015, 115, 11503-11532; d) G. Wang, A. Rühling, S. Amirjalayer, M. Knor, J. B. Ernst, C. Richter, H-J. Gao, A. Timmer, H-Y. Gao, N. L. Doltsinis, F. Glorius, H. Fuchs, Nat. Chem., 2017, 9, 152-156. 76 A. Vivancos, C. Segarra, M. Albrecht, Chem. Rev., 2018, 118, 9493-9586. 77 M. Melaimi, R. Jazzar, M. Soleilhavoup, G. Bertrand, Angew. Chem. Int. Ed., 2017, 56, 10046-10068. 78 E. Tomás-Mendivil, M. M. Hansmann, C. M. Weinstein, R. Jazzar, M. Melaimi, G. Bertrand, J. Am. Chem. Soc., 2017, 139, 7753-7756. 79 O. Rivada-Wheelaghan, M. A. Ortuño, J. Díez, A. Lledós, S. Conejero, Angew. Chem. Int. Ed., 2012, 51, 3936-3939. 80 O. Rivada-Wheelaghan, M. A. Ortuño, J. Díez, S. E. García-Garrido, C. Maya, A. Lledós, S. Conejero, J. Am. Chem. Soc., 2012, 134, 15261-15264. Chapter 1 Nickel(II) catalysed Carbon Dioxide Hydrosilation Chapter 1. Introduction [68] on the reaction conditions. Since then, organometallic complexes have been employed in combination with various hydride sources like hydrogen, boranes or silanes so as to overcome the kinetic barriers of this transformation (Scheme 1). Additionally, selective synthesis of any of the intermediates of this synthetic pathway can be obtained in a few cases by careful design of the catalyst, thus precluding the presence of mixtures of compounds which need to be purified afterwards. Scheme 1. Overview of the sequential steps in CO2 reduction assisted by hydride sources (vertical reduction in Figure 2). Conversion to methane from methoxyborane is not known. 19 In the case of hydrogen, the main handicap is its lack of reactivity due to its absence of polarity and the high stability of its σ bond (BDEH-H = 104.2 kcal/mol). 20 This becomes an additional problem, because that implies there would be two inert molecules (H2 and CO2) that need to be activated instead of one. Thus, very high pressure of both gases needs to be applied to get some conversion. These disadvantages, along with some complications inherent to the nature of the reduced products remain as problems still to be solved. For instance, reduction of CO2 to formic acid is not favourable thermodynamically due to entropic factors (ΔHr = -7.5 kcal/mol and ΔGr = 7.6 kcal/mol) 21 and the 19 S. Bontemps, Coord. Chem. Rev., 2016, 308, 117-130. 20 G. Herzberg, A. Monfils, J. Mol. Spect., 1960, 5, 482-498. 21 W. Leitner, Angew. Chem. Int. Ed. Engl., 1995, 34, 2207-2221. Nickel(II) catalysed Carbon Dioxide Hydrosilation [69] most common solution to displace the equilibrium to the products is the addition of Lewis basic additives (usually amines) or solvents so the product can be obtained as a salt or it can be stabilized by hydrogen-bonding, respectively. 22 The use of this type of additives is also extensively used in the methanol synthesis, but in this case it is employed to circumvent kinetic issues: carbon dioxide can be captured by alcohols or amines (Figure 2, diagonal and horizontal reactions) in order to give substrates that are easier to hydrogenate to the methanol stage (Figure 5, pathway B). 23 Figure 5. Synthetic pathways for obtaining methanol from carbon dioxide (adapted from reference 23). Some of the few examples to date in which methanol can be directly obtained by hydrogenation include the one reported by Sanford et al. (Scheme 2A), where a system of 3 different catalysts achieve effective cascade transformation to sequentially afford formic acid, methyl formate and finally methanol (Figure 5, pathway B). 24 On the other hand, Klankermayer and Leitner 22 a) J. Klankermayer, S. Wesselbaum, K. Beydoun, W. Leitner, Angew. Chem. Int. Ed., 2016, 55, 7296-7343; b) W. H. Bernskoetter, N. Hazari, Acc. Chem. Res., 2017, 50, 1049-1058. 23 K. Sordakis, C. Tang, L. K. Vogt, H. Junge, P. J. Dyson, M. Beller, G. Laurenczy, Chem. Rev., 2018, 118, 372-433. 24 C. A. Huff, M. S. Sanford, J. Am. Chem. Soc., 2011, 133, 18122-18125. Chapter 1. Introduction [70] in 2012, 25 and Beller in 2017 26 described very similar conditions for CO2 hydrogenation to methanol, given the similitude in solvent, ligand and acidic additive for the reaction to proceed (Scheme 2B and 2C). In both cases, analogous reaction times (24 h) and considerably harsh conditions are required for achieving low amounts of product. A significant advance was achieved in 2016 by the group of Prakash and Olah. 27 In this work, the authors employed pentaethylenehexamine to capture carbon dioxide, after which the resulting formate was hydrogenated by using a ruthenium catalyst (Scheme 2D). The reported system proved to be robust enough to be recycled 5 times giving TON values superior to 2000, and it turned out to be active enough to capture carbon dioxide from air, transforming it into methanol with good yields. 25 S. Wesselbaum, T. vom Stein, J. Klankermayer, W. Leitner, Angew. Chem. Int. Ed., 2012, 51, 7499-7502. 26 J. Schneidewind, R. Adam, W. Baumann, R. Jackstell, M. Beller, Angew. Chem. Int. Ed., 2017, 56, 1890-1893. 27 J. Kothandaraman, A. Goeppert, M. Czaun, G. A. Olah, G. K. S. Prakash, J. Am. Chem. Soc., 2016, 138, 778-781. Nickel(II) catalysed Carbon Dioxide Hydrosilation [71] Scheme 2. Current methods to directly obtain methanol from carbon dioxide hydrogenation (adapted from references 24-27). The obtainment of formaldehyde from CO2 hydrogenation is even more underdeveloped, given that there are only a few examples reported in the literature. 28 These include a stoichiometric reaction described by Floriani in 1985, in which Schwartz’s reagent reacts with CO2 to afford an oxygen-bridged dimer and formaldehyde, 29 and a catalytic reaction also reported by 28 L. E. Heim, H. Konnerth, M. H. G. Prechtl, Green. Chem., 2017, 19, 2347-2355. 29 S. Gambarotta, S. Strologo, C. Floriani, A. Chiesi-Villa, C. Guastini, J. Am. Chem. Soc., 1985, 107, 6278-6272. Chapter 1. Introduction [72] Klankermayer and Leitner in 2016, where their same ruthenium complex described above was combined with aluminium triflate in methanol to yield the acetal (Scheme 1, formaldehyde stage) and the methyl formate derivatives in a 2:1 ratio (Scheme 3, equations A and B respectively). 30 Scheme 3. Different methods to obtain formaldehyde from carbon dioxide hydrogenation (adapted from references 29 and 30). The scarcity of results (with low turnover numbers and poor selectivity sometimes) from direct hydrogenation of CO2 to reduced derivatives along with the harsh reaction conditions required reflect once again the challenge in using hydrogen as a reductant, showing that further work needs to be done in this field. An alternative approach would be the employment of reductants which possess less energy demanding activation barriers, such as boranes or silanes. The polarity of these substrates makes them hydridic enough (BDEB-H = 82.6 kcal/mol and BDESi-H = 70.2 kcal/mol) to attack the carbon atom of CO2 once it has been activated by a transition metal. Moreover, the strength of the new resulting bonds (BDEB-O = 193.5 kcal/mol and BDESi-O = 191.4 kcal/mol) acts as a thermodynamic driving force for the reduction when using these reagents instead of hydrogen (BDEH-O = 102.8 kcal/mol). 31 All these data seems to suggest 30 K. Thenert, K. Beydoun, J. Wiesenthal, W. Leitner, J. Klankermayer, Angew. Chem. Int. Ed., 2016, 55, 12266-12269. 31 Y-R. Luo, Comprehensive Handbook of Chemical Bond Energies, CRC Press: Boca Raton, FL, 2007. Nickel(II) catalysed Carbon Dioxide Hydrosilation [73] that carbon dioxide should undergo reduction reactions under mild conditions when using these hydride sources, which is desirable from the point of view of safety and economy, given that the energy required would be much lower. The only drawback to this method is the generation of boryl ethers or siloxanes; in order to regenerate the corresponding hydroborane or hydrosilane, the employment of a hydride source would be necessary.16 Nonetheless, siloxanes are very valuable in the polymer industry, and “what may be by-products to organic chemists could be the synthetic target for a polymer chemist”. 32 Therefore, the by-products of CO2 hydrosilation could be useful in the silicone chemistry. 1.3.1 Formaldehyde derivatives Formaldehyde is synthesized in the chemical industry by means of the Formox process. It consists of a 3-step process in which natural gas is transformed to syngas by reaction with water at very high temperatures. Then, the syngas is reduced to methanol, which is later oxidized, affording formaldehyde as the main product at the end of the process (Scheme 4).28 Scheme 4. Current industrial synthesis of formaldehyde (adapted from reference 28). 32 J. B. Grande, “The Piers-Rubinsztajn Reaction: New Routes to Structured Silicones”. PhD Thesis, McMaster University, 2013. Chapter 1. Introduction [74] Following the previous examples, the conditions employed in this process call for an alternative method, less energy-demanding, more economically viable and environmentally friendly. As it will be shown below, the chemistry of hydroboration and hydrosilation to the formaldehyde stage is still somewhat primitive, with only a few systems described on each case. 33 Bis(boryl)acetal derivatives Reduction to formatoborane or methoxyborane derivatives is wellknown, and there are numerous examples describing such transformations. However, chemoselectivity towards the intermediate bis(boryl)acetal is rare.33 Hydroboration to the formaldehyde level was firstly described by Sabo-Etienne et al. in 2012. In their work, they showed that the ruthenium complex [RuH2(η2- H2)2(PCy3)2] was able to afford the formaldehyde surrogate in the presence of pinacolborane, albeit in very low yield and poor selectivity (26% of methylene products) due to the presence of numerous species in the reaction mixture (Scheme 5). 34 In a subsequent article, they could trap both methylene species by reaction with CD3OD, yielding the corresponding hemiacetal CD3OCH2OD. 35 Scheme 5. Hydroboration of carbon dioxide catalyzed by a ruthenium complex (adapted from reference 34). 33 For more examples of carbon dioxide hydroboration, see reference 19. For more examples of carbon dioxide hydrosilation, see: F. J. Fernández-Álvarez, L. A. Oro, ChemCatChem, 2018, 10, 4783-4796. 34 S. Bontemps, L. Vendier, S. Sabo-Etienne, Angew. Chem. Int. Ed., 2012, 51, 1671-1674. 35 S. Bontemps, S. Sabo-Etienne, Angew. Chem. Int. Ed., 2013, 52, 10253-10255. Nickel(II) catalysed Carbon Dioxide Hydrosilation [75] In 2014, the same group improved the synthesis of the bis(boryl)acetal by slightly modifying the structure of the ligands and the reaction conditions (solvent, catalyst loading, CO2 pressure), and they trapped it by means of a bulky amine such as 2,6-diisopropylaniline, achieving yields up to 74% of the resulting imine (Scheme 6A). Such a high yield is in contrast with the observed yield for the acetal species, but the authors attribute the high conversion obtained due to the role of pinacolborane as a dehydrating agent, displacing the equilibrium to the products. In the final step, the imine can undergo hydrolysis to produce formalin. 36 A year later, they explored this type of transformation by using an iron catalyst. This time, the resulting acetal was used as a CH2 transfer agent in a wide variety of molecules, affording C−C, C−O and C−N bonds (Scheme 6B). 37 A similar strategy was employed by Turculet et al. after transforming carbon dioxide to the acetal level with group 10 metals (Scheme 7). 38 Scheme 6. Different strategies followed by Sabo-Etienne et al. to capture formaldehyde from hydroboration of carbon dioxide (adapted from references 36 and 37). 36 S. Bontemps, L. Vendier, S. Sabo-Etienne, J. Am. Chem. Soc., 2014, 136, 4419-4425. 37 G. Jin, C. G. Werncke, Y. Escudié, S. Sabo-Etienne, S. Bontemps, J. Am. Chem. Soc., 2015, 137, 9563-9566. 38 L. J. Murphy, H. Hollenhorst, R. McDonald, M. Ferguson, M. D. Lumsden, L. Turculet, Organometallics, 2017, 36, 3709-3720. Chapter 1. Introduction [76] Scheme 7. Approach followed by Turculet et al. to transfer a methylene fragment into organic molecules from carbon dioxide (adapted from reference 38). Another example of CO2 hydroboration to the formaldehyde level was reported by Berthet and Cantat in 2016. The authors used a set of 1st row transition metal complexes (Cu, Co and Fe) stabilized by silylphosphine ligands as catalysts in the reaction between carbon dioxide and 9-BBN (9- borabicyclo[3.3.1]nonane). Whereas the Co and Fe compounds showed chemoselectivity towards the methoxyborane (up to 100%) at 60 °C, the Cu derivatives gave the acetal species in yields up to 92% in less forcing reaction conditions and faster reactions (Scheme 8). 39 Scheme 8. Catalytic systems developed by Berthet and Cantat for CO2 hydroboration (adapted from reference 39). 39 A. Aloisi, J-C. Berthet, C. Genre, P. Thuéry, T. Cantat, Dalton Trans., 2016, 45, 14774-14788. Nickel(II) catalysed Carbon Dioxide Hydrosilation [77] Bis(silyl)acetal derivatives In the same way as hydroboration, chemoselective partial reduction to intermediate species is also challenging in the case of hydrosilation. In fact, various reports have described the synthesis of silyl formates or the full reduction to methane, yet the controlled reduction to bis(silyl)acetal or methoxysilane derivatives is still uncommon, with only a few examples described (Scheme 9). 40 Scheme 9. Overview of the current situation in CO2 hydrosilation chemistry. NB: The last step (i.e. over-reduction of alkoxysilanes to alkanes mediated by silanes and B(C6F5)3) is called Piers- Rubinsztajn reaction. 41 In the hydrosilation situation, it is first necessary to mention that activation of the Si−H bond by strong Lewis acids can take place, increasing the hydricity of the reducing agent. As a result, these species are powerful reductants that can achieve full conversion to methane from silyl formates (Scheme 9). One of the most common compounds of this sort is that composed of triethylsilane and tris(pentafluorophenyl)borane (BCF). 42 In fact, these borane-silane complexes have been extensively studied in the hydrosilation of unsaturated molecules such as carbonyl, imine, or alkene functionalizations (Scheme 10), the 40 T. T. Metsänen, M. Oestreich, Organometallics, 2015, 34, 543-546. 41 M. A. Brook, Chem. Eur. J., 2018, 24, 8458-8469. 42 D. J. Parks, J. M. Blackwell, W. E. Piers, J. Org. Chem., 2000, 65, 3090-3098. Chapter 1. Results and discussion [84] In this chapter, the synthesis of a Ni(II) complex stabilized by a bis(phosphino)boryl ligand and experimental data showing its performance as a catalyst for carbon dioxide hydrosilation are described. Experimental and theoretical mechanistic studies are also shown, as well as a proposal for a catalytic cycle. 2. Results and discussion The aim of this work is to exploit the strong trans influence exerted by the PBP ligand and apply it in CO2 activation. A common strategy for activating this molecule involves nucleophilic attack from a ligand of a metal complex to the carbon atom of carbon dioxide, which possesses a slight electrophilic character (the higher contribution of the carbon atom to the LUMO can be observed in Figure 6). If the C−O bond weakens enough, CO2 can insert into the metal-ligand bond. Figure 6. Frontier molecular orbitals of carbon dioxide, showing its ambiphilic character; it can behave as a nucleophile through the oxygen atoms (HOMO) or it can act as an electrophile through the carbon atom (LUMO). Taking this information into consideration, a hydride ligand in trans to the boron atom of the PBP framework would be strongly nucleophilic, given that the M−H bond is weakened by the strong trans influence of the boryl moiety. In this way, the hydride could attack a carbon dioxide molecule to yield the Nickel(II) catalysed Carbon Dioxide Hydrosilation [85] corresponding formate. 50 Therefore, a complex containing a hydride ligand in trans to the PBP scaffold was envisaged as a potential compound for activating carbon dioxide. 2.1. Synthesis of Ni−PBP complexes The synthesis of Ni(II) complexes stabilized by the tBuPBP ligand 1 has been previously described in our group: NiBr2(dme) can react with 1 in toluene at room temperature to afford the bromide derivative 2 in very high yields. Et3N is needed in order to neutralize the hydrobromic acid resulting from the activation of the B−H bond (Scheme 17). 51 Scheme 17. Utilization of the tBuPBP ligand 1 in the synthesis of Ni(II) complexes. In order to test the σ-donation capability of the boryl ligand, the carbonyl derivative 3 was synthesized in dichloromethane by using the halogen abstractor NaBAr4F (sodium tetrakis(3,5-bis(trifluoromethyl)phenyl)borate) in the presence of CO (Scheme 18). The poor coordination ability of the borate anion 52 makes possible the obtainment of a clean product after a few minutes at room temperature. After work-up, X-Ray quality crystals (Figure 7) can be grown by slow diffusion of hexamethyldisiloxane into a dichloromethane solution (10:1 ratio) of 3. 50 J. E. Heimann, W. H. Bernskoetter, N. Hazari, J. M. Mayer, Chem. Sci., 2018, 9, 6629-6638. 51 N. Curado, C. Maya, J. López-Serrano, A. Rodriguez, Chem. Commun., 2014, 50, 15718- 15721. 52 N. A. Yakelis, R. G. Bergman, Organometallics, 2005, 24, 3579-3581. Chapter 1. Results and discussion [86] Scheme 18. Synthesis of carbonyl derivative 3. The structure of 3 corresponds to a somewhat distorted square-planar Ni(II) complex in which the geometrical features follow the same trend as the previous complexes of this type reported by our group, such as Ni−P distances or P−Ni−P angles.51 Figure 7. X-Ray structure of complex 3. Anion and hydrogen atoms have been omitted for clarity. This complex gives a singlet at 114.5 ppm in the 31P{1H} NMR spectrum, indicative of the symmetric nature of the resulting square-planar complex in which both phosphine ligands are equivalent. A broad and a sharp peak at 38 and -6.6 ppm in the 11B{1H} spectrum reveals the presence of a boryl and borate functionalities respectively, in agreement with the structure of 3 (Figure 8). Nickel(II) catalysed Carbon Dioxide Hydrosilation [87] Unlike the case of some cationic Ni−PCP carbonyl derivatives, 53 we were able to observe the resonance of the CO ligand by 13C{1H} NMR, showing a singlet at 193.6 ppm. The observation of a singlet (and not a triplet due to coupling with the P atoms) might be attributable to the weak signal to noise ratio because of sensitivity reasons. An IR spectrum (Nujol) was also recorded on the sample, and it showed a ν(CO) = 2049 cm-1, indicative of an appreciable degree of backbonding to the carbonyl ligand (ν(CO) = 2143 cm-1 for free CO). 54 As it can be seen in table 1, compound 3 seems to show a higher debilitation of the carbon monoxide fragment in comparison to the aryl based PCP,53d pointing out a greater trans influence of the PBP ligand. Nevertheless, the alkyl-based PCP ligand displays a similar value,53c in accordance with their analogous overall electronic donating abilities.13 A closer look to both crystal structures reveals some information about the nature of the ligand in trans to the carbonyl moiety: whereas both complexes show a similar C−O distance to that observed in free CO (~1.128 Å),54 the alkyl-based pincer complex possesses a shorter Ni−Ccarbonyl bond (1.78 Å) compared to 3 (1.81 Å), denoting a lower trans influence than the PBP framework. 53 a) C. J. Moulton, B. L. Shaw, J. Chem. Soc. Dalton Trans., 1976, 1020-1024; b) P. Giannoccaro, G. Vasapollo, A. Sacco, J. Chem. Soc. Chem. Commun., 1980, 1136-1137; c) A. Castonguay, A. L. Beauchamp, D. Zargarian, Inorg. Chem., 2009, 48, 3177-3184; d) S. J. Connelly, A. C. Zimmerman, W. Kaminsky, D. M. Heinekey, Chem. Eur. J. 2012, 18, 15932- 15934. 54 N. N. Greenwood, A. Earnshaw, Chemistry of the Elements, 2nd ed. Butterworth- Heinemann: Woburn, MA, 1997. Chapter 1. Results and discussion [88] Compound ν(CO)/ cm-1 C−O distance/Å Ni−C distance/ Å δ(13C)/ ppm Reference 2049 1.131(6) 1.810(4) 193.6 This work 2065 Not given Not given Not given 53d 2040 1.139 1.780 Not found 53c Table 1. Comparison of some structural data of cationic Ni pincer carbonyl complexes. Nickel(II) catalysed Carbon Dioxide Hydrosilation [89] Figure 8. Relevant NMR (400 MHz, CD2Cl2) features of complex 3. Bromide derivative 2 can serve as a precursor to alkyl compounds such as methyl complex 4 (Scheme 19). Our group previously reported the synthesis of this species by making use of an excess (10 equivalents) of the Grignard reagent MeMgBr in order to achieve full conversion to the desired compound.51 The need for such a high amount of the reagent might be due to the strong alkylating character of complex 4 because of the presence of the boryl ligand trans to it, 55 which would favour the opposite reaction (i.e. methylation of MgBr2). 55 J. Zhu, Z. Lin, T. B. Marder, Inorg. Chem., 2005, 44, 9384-9390. Chapter 1. Results and discussion [90] Scheme 19. Synthesis of methyl complex 4. Following the synthesis of 4, our group studied its reactivity towards dihydrogen. In this work, cooperativity of the PBP ligand was demonstrated in the heterolytic cleavage of the hydrogen molecule to yield hydride 6 (Scheme 20): upon coordination of H2, the methyl group can attack one of the hydrogen atoms (H+) in order to form methane, whereas the remaining hydrogen atom (hydride) attacks the boron atom of the ligand (transition state of Scheme 20). Thus, the latter behaves as a Lewis acid in this mechanism, in which the Ni(0) intermediate 5 stabilized by a η2 B−H interaction is formed. Finally, 5 undergoes oxidative addition to give hydride species 6. Experimental evidence for such intermediate and theoretical calculations support this mechanism.51 This method is an alternative way to the introduction of hydride ligands in a metal center, given that the strategy of ligand substitution by using numerous hydride reagents proved unfruitful. Scheme 20. Hydrogenolysis of the Ni−Me bond assisted by the PBP ligand (adapted from reference 51). Nickel(II) catalysed Carbon Dioxide Hydrosilation [91] With hydride 6 in hand, reactivity tests against carbon dioxide could be performed. In this way, exposure of 6 to an atmosphere of CO2 (1 bar) resulted in the instantaneous clean formation of a new species showing a resonance at 9.11 ppm in the 1H NMR spectrum which correlates in the HSQC spectrum with a new carbon peak in 13C NMR at 167.2 ppm. Given that the hydride signal disappeared in this reaction, the new signals were assigned to the newly formed formate group (Scheme 21). Scheme 21. Formation of formate 7 from hydride 6 upon exposure to CO2. Complex 7 also shows a singlet at 85.2 ppm and a broad peak at 37.5 ppm in the 31P{1H} and 11B{1H} NMR spectra, respectively (Figure 9). It is necessary to mention that by the time these studies were taking place in our laboratory, the group of Prof. Jonas C. Peters reported the isolation and characterization of formate 7. 56 56 T-P. Lin, J. C. Peters, J. Am. Chem. Soc., 2014, 136, 13672-13683. Chapter 1. Results and discussion [92] Figure 9. NMR spectra (400 MHz, C6D6) of the reaction mixture of the synthesis formate 7. 2.2. Theoretical study of CO2 insertion in Ni−H pincer complexes The influence of the PBP ligand in the carbon dioxide insertion into the Ni−H bond was studied in more detail by means of DFT calculations. 57 For this purpose, several Ni−H complexes were compared (Figure 10), among which some of them have been experimentally synthesized (PCP, 58 POCOP 59 and PSiP 60 derivatives), so the reported results can be compared to the ones we produced 57 Insertion of CO2 into different types of metal-element σ bonds has been studied in detail. See: N. Hazari, J. E. Heimann, Inorg. Chem., 2017, 56, 13655-13678. 58 T. J. Schmeier, N. Hazari, C. D. Incarvito, J. A. Raskatov, Chem. Commun., 2011, 47, 1824- 1826. 59 a) S. Chakraborty, J. Zhang, J. A. Krause, H. Guan, J. Am. Chem. Soc., 2010, 132, 8872-8873 ; b) S. Chakraborty, Y. J. Patel, J. A. Krause, H. Guan, Polyhedron, 2012, 32, 30-34. 60 W-H. Suh, M. L. Guard, N. Hazari, Polyhedron, 2014, 84, 37-43. Nickel(II) catalysed Carbon Dioxide Hydrosilation [93] in the laboratory using hydride 6. 61 In this work, model complexes were calculated with methyl moieties on the phosphine or amine groups, and the energy for the transition state (Figure 10, red) of each complex was calculated as well as that corresponding to the resulting formates (Figure 10, pink). Figure 10. ΔG⧧ (red, top) and ΔG0 (pink, bottom) for the insertion of CO2 into the Ni−H bond of several pincer complexes. Energies in kcal/mol. As expected, complexes 6PSiP,Me and 6Me possess the lowest energy barriers for CO2 insertion, with Gibbs free energies of 6.1 and 6.3 kcal/mol, respectively. This is in excellent agreement with the results published by Marder and Lin in 2005,55 where they observed the strong trans influence exerted by silyl and boryl ligands. Unlike the PNP derivative, all the complexes give stable formates, indicating the thermodynamic feasibility of the process. Cationic 61 Some studies previously reported the energies for CO2 insertion in Ni−R or Pd−R groups. See: a) T.J. Schmeier, A. Nova, N. Hazari, F. Maseras, Chem. Eur. J., 2012, 18, 6915-6927; b) W. H. Suh, T. J. Schmeier, N. Hazari, R. A. Kemp, M. K. Takase, Organometallics, 2012, 31, 8225-8236. Chapter 1. Results and discussion [100] Experimental value Calculated value Distance (Å) B1−Ni1 1.903(2) 1.90874 O1−Ni1 1.978(1) 1.99717 B2−O2 1.552(2) 1.55803 O1−C25 1.219(2) 1.23942 O2−C25 1.283(2) 1.27708 Angle (°) B1−Ni1−O1 171.34(6) 167.393 O1−C25−O2 124.1(1) 124.604 C25−O2−B2 128.0(1) 123.263 P1−Ni1−P2 158.61(2) 158.124 Table 3. Comparison of experimental and theoretical geometry data of 8. With pure complex 8 in hand, catalytic hydrosilation of carbon dioxide was explored by using Et3SiH under 4 bar of CO2 at 70 °C in C6D6. Selective reduction to the bis(silyl)acetal level was observed during the whole process, as in Figure 12. The catalyst was efficient enough to allow us to lower the catalytic loading down to 0.05% molar (relative to Et3SiH), achieving turnover numbers (TON) and frequencies (TOF) of 1200 and 56 h-1, respectively. Such a remarkable selectivity can be observed in Figure 15, where only the silane and acetal signals are observed by 1H NMR: Nickel(II) catalysed Carbon Dioxide Hydrosilation [101] Figure 15. Representative example of catalysis (400 MHz, C6D6) with 8 (0.05% mol) and Et3SiH after 21 hours at 70 °C. The scope of this catalytic reaction was studied by employing several other silanes such as Ph2MeSiH (table 4), which also gives the acetal exclusively, with negligible amounts of silyl methyl ether (Figure 16). Faster catalysis can be achieved with the smaller silane Me2PhSiH, albeit the selectivity is compromised, since some formation of the methoxysilane and methane is observed (Figure 17). It was observed that an increase in the concentration of the reactants was beneficial to the reaction rate, and from table 4 we can conclude that catalyst 8 gives similar turnover numbers and higher turnover frequencies in comparison to the most efficient transition-metal based catalysts reported in the literature.46 Chapter 1. Results and discussion [102] TOF / h-1 55.8 26.8 9.2 1.3 16.5 18.6 18.3 2 40 TON b 1200 910 143 90 620 410 137 90 121 t / h 21.5 34 15.5 70 37.5 22 7.5 45 3 Bis(silyl) acetal / % > 99 > 99 > 99 > 99 > 99 > 99 > 99 > 99 73 (11% silyl methyl ether / 16 % methane) Conversion / % 60 91 86 90 62 41 82 90 90 c Total volume / μLa 304 S + 100 B 304 S + 100 B 101 S + 250 B 30 S + 600 B 378 S + 100 B 378 S (Neat) 126 S + 225 B 38 S + 600 B 97 S + 250 B [Silane] / M 4.72 4.72 1.81 0.3 4 5.04 1.81 0.3 1.83 [8] / mM 2.36 4.77 10.9 3 3.98 5 10.9 3 10.9 Catalyst loading / % mol 0.05 0.1 0.6 1 0.1 0.1 0.6 1 0.6 Silane Et3SiH Ph2MeSiH Me2PhSiH Table 4. Summary of the results of the carbon dioxide hydrosilation catalyzed by 8. Every catalytic experiment has been repeated at least 3 times. Partial decomposition of the catalyst was observed after the conversion values indicated above. a S = silane; B = d6-Benzene b (μmoles of silane/ μmoles of catalyst) * acetal ratio c Toluene was used as internal standard. Nickel(II) catalysed Carbon Dioxide Hydrosilation [103] Figure 16. Representative example of catalysis (400 MHz, C6D6) with 8 (0.1% mol) and Ph2MeSiH after 21 hours at 70 °C. Chapter 1. Results and discussion [104] Figure 17. Representative example (400 MHz, C6D6) of catalysis with 8 (0.6% mol) and Me2PhSiH after 3 hours at 70 °C. Toluene was added after the experiment. 2.4. Experimental mechanistic studies In order to elucidate the mechanism by which the catalytic reactions take place, some experiments were carried out so as to explain the extraordinary selectivity and the good catalytic data. Given that free B(C6F5)3 in contact with the silane is responsible for the over-reduction of the intermediates to methane, the first experiment involved the assessment of the equilibrium displayed in scheme 23 in the catalytic conditions. Scheme 23. Equilibrium between complexes 6, 7 and 8. Nickel(II) catalysed Carbon Dioxide Hydrosilation [105] For this purpose, a sample of 8-13CO2 (prepared using 13C-labelled carbon dioxide) was heated up to 70 °C in C6D6 in a sealed NMR tube. After 4 hours, no evidence of free borane was observed by NMR spectroscopy (13C, 19F, 11B). Nonetheless, when the same NMR tube is charged with 12CO2 (1 bar), the insertion of non-labelled carbon dioxide is observed in the 1H NMR spectrum in a 12% ratio after 22 hours at 70 °C (Figure 18), which indicates that the equilibrium shown in scheme 23 is actually taking place at the catalytic conditions, although it is strongly shifted to the right. Therefore, the amount of Lewis acid in the reaction medium is minimal. Figure 18. Exchange experiment: 1H NMR spectrum (400 MHz, C6D6) of 8-13CO2 after 22 h heating at 70 °C in the presence of 12CO2 (1 bar). Incorporation of 12CO2 in a 12% ratio. These experimental results seem to agree with the Gibbs energy difference found in our calculations for the model complex 8Me (-27 kcal/mol) Chapter 1. Results and discussion [106] and 7 + B(C6F5)3 (-11.8 kcal/mol). 66 In comparison to Cp*2Sc,45 the PBP-Ni framework is more efficient in capturing the Lewis acid, which is consistent with the selectivity observed for both systems (Schemes 12 and 24). Scheme 24. Comparison of the energies of Ni-PBP and Cp*2Sc complexes and their B(C6F5)3 adducts (adapted from reference 45). 13C-labelled carbon dioxide was also employed in a monitoring experiment, where the formation of the formaldehyde derivative as well as the presence of intermediates was analysed over time (Figure 19): 66 Complex 6 + CO2 + B(C6F5)3 were used as the energy reference. Nickel(II) catalysed Carbon Dioxide Hydrosilation [107] Figure 19. 13C (fully coupled) NMR spectra (400 MHz, C6D6) of the monitoring and overreduction experiment using 8 and Et3SiH in the presence of 13CO2 (4 bar). It can be observed that no other product but the bis(silyl)acetal (triplet at 85.5 ppm with 1JCH = 161 Hz) is generated throughout the experiment. After 8 hours, the carbon dioxide peak diminished to some extent and the main observable compound is the formaldehyde derivative. At this point, 4 equivalents of the Lewis acid were added in order to over-reduce the product to methane, to confirm that an excess of tris(pentafluorophenyl)borane with respect to the Ni complex should be avoided if selectivity is desired. Nevertheless, to make sure that the Ni complex was not responsible for this overreduction and it was only due to the formation of the borane-silane complex, the bis(silyl)acetal was isolated (Figure 20). Chapter 1. Results and discussion [108] Figure 20. NMR characterization (400 MHz, C6D6) of the bis(silyl)acetal (Et3SiO)2CH2. This product can be purified under air using column chromatography techniques, affording a colorless liquid after solvent removal. Once it was isolated, it was mixed with C6D6, Et3SiH and B(C6F5)3 and it was analysed by NMR techniques. Figure 21 shows the evolution of the process by using 29Si{1H} NMR, where the formation of the siloxane (Et3Si)2O is observed upon heating up to 50 °C. No silyl methyl ether is detected, indicating formation of methane. Nickel(II) catalysed Carbon Dioxide Hydrosilation [109] Figure 21. 29Si{1H} NMR spectra (400 MHz, C6D6) of the over-reduction of the isolated acetal species (Et3SiO)2CH2 in the absence of Ni complexes. Finally, it is necessary to mention that studies in which CO2 is consumed completely, over-reduction to methane is observed and the catalyst changes its nature, since there is no more carbon dioxide to regenerate the formatoborate unit (Figure 22). Chapter 1. Results and discussion [116] unlikely, and the rest of the calculations will be performed for pathway b with the real complex. Calculations with real catalyst 8 The employment of the real catalyst seems to have a negligible effect in the dissociation of the Lewis acid, since similar energy values (30.9 kcal/mol) were obtained (Figure 26). Similarly, this stage is the most energy-demanding of the whole mechanism. No transition state could be found after this step, because the system would either evolve to the reagents (7 + borane-silane complex) or the products (A). The results of a relaxed potential energy surface scan in which the Si−O distance is modified and the energy is evaluated accordingly (Figure 27) along with the almost non-existent barrier in the case of the model complex suggest that the attack of the formate to the activated silane occurs with practically no energy cost. Figure 26. Energy profile of the hydrosilation to the formate stage using real catalyst 8. ΔZPE in kcal/mol (ΔG50 in parentheses). Nickel(II) catalysed Carbon Dioxide Hydrosilation [117] Figure 27. Relaxed Potential Energy Surface scan for the Me3Si+ transfer from the borane-silane complex to the formate group. The transition state would be located in the flat region in the middle of the curve. 2) Second Reduction Step Hydride transfer It can be observed in Figure 26 that the molecule rearranges after the attack from A to a more stable geometry A’ 8.2 kcal/mol below the origin. At this point, two possibilities arise (Scheme 26): the first one (i) involves the displacement of the silylformate unit from the coordination sphere by attack of the hydridoborate anion [HB(C6F5)3]-, and the second possibility (ii) consists of the hydride transfer to the carbonyl atom to yield species B·B(C6F5)3. Both options are equal in terms of energy (around 25 kcal/mol above A’), although the Chapter 1. Results and discussion [118] experimental results showed no trace of silylformate during catalysis and/or stoichiometric reactions, which seems to point out that step ii is much faster than i. Scheme 26. Possible scenarios of intermediate A’ for the attack of the hydridoborate unit. ΔZPE in kcal/mol from A’. Again, the transition state for this process using the real catalyst could not be found, but the model system showed that going from B·B(C6F5)3 to A’ required no energy to occur. Moreover, a relaxed potential energy surface scan also supports the kinetic ease of the reaction. Silyl transfer The interaction that holds the borane and the Ni fragment in B·B(C6F5)3 is weak, so the Lewis acid can interact with another silane molecule in the reaction medium to give rise to another borane-silane complex (Figure 28, B·[Me3Si−H···B(C6F5)3]). Then, the oxygen atom bound to the metal can attack the activated silane to yield Ni hydridoborate 9 and the bis(silyl)acetal product. It can be observed in Figure 28 that this silylium transfer also proceeds Nickel(II) catalysed Carbon Dioxide Hydrosilation [119] kinetically swift since the energy for T2b is minuscule. 70 The final product of this reaction is the pair C [(PBP)NiO(SiMe3)CH2OSiMe3][HB(C6F5)3]. Figure 28. Energy profile of the second reduction stage. 3) Activation of CO2 and catalyst regeneration The Ni···O interaction in C can be cleaved by the hydridoborate [HB(C6F5)3]- to give complex 9 and the free acetal derivative. It can be seen in Figure 28 that this step is also thermodynamically favourable. Given that during catalysis a homogeneous solution is always present, the lifetime of 9 must be rather short so as not to precipitate from the reaction medium. Two very close minima (1.5 kcal/mol away from each other) were found when calculating the geometry of complex 9, in which the hydridoborate unit can 70 NB: Unless dispersion and entropy corrections are applied, ΔG⧧ would be above 50 kcal/mol with respect to the origin. Considering only dispersion corrections, ΔG⧧ = 36.9 kcal/mol. This would mean the unfeasibility of this mechanism. Moreover, if it were to be considered as a possibility, it would imply the accumulation of the silyl formate rather than the bis(silyl)acetal, which opposes the experimental facts. This is a clear example that remarks the importance of considering entropy and dispersion effects in the calculations. Chapter 1. Results and discussion [120] interact with Ni via κ1-F or κ1-H arrangements (Figure 29), as in the scandocinium system reported by Piers et al.45 Figure 29. Optimized structure of complex 9. The two possible interactions with the Ni framework are shown. In the presence of 9, carbon dioxide can insert into the cavity created between the 2 ions by stablishing an interaction (2.08 Å) with Ni through one of the oxygen atoms (D, Figure 30). At this point, the carbon dioxide molecule is still rather linear (O−C−O angle = 173 °). Then, a first transition state is found (TS3, 0.3 kcal/mol away) in which the hydride transfer to the carbon atom takes place. Immediately after, intermediate E features a weak H···B interaction (2.36 Å) and a closer O−C−O angle (129 °). Finally, attachment of the borane to the free oxygen of the formate regenerates catalyst 8 after surpassing a second transition state (2.3 kcal/mol from E). The whole operation requires practically no energy to overcome the two transition states, and it is quite exothermic, since catalyst 8 is more stable than 9 by 34.9 kcal/mol. These values account for the instantaneous formation of the catalyst after one catalytic cycle and the low degree of CO2 exchange when using a mixture of labelled and non-labelled carbon dioxide, since the reverse step requires a considerable amount of energy. Nickel(II) catalysed Carbon Dioxide Hydrosilation [121] Figure 30. Energy profile for the regeneration of catalyst 8 from hydridoborate 9. 4) How is selective CO2 hydrosilation possible? Efficient sequestration of BCF has been previously postulated as the reason for the selectivity in carbon dioxide hydrosilation.46 In the Ni-PBP case, liberation of BCF is difficult in all the possible scenarios during the catalytic process: -Dissociation of the Lewis acid from the catalyst (Scheme 27, equation 1) requires a high amount of energy and the equilibrium is therefore displaced to the left, as well as in the case of the release of B(C6F5)3 from complex 9 (equation 2). Besides, 9 reacts more favourably with CO2 and Me3SiOCHO. -B·B(C6F5)3 could also release borane given that it interacts weakly with the Ni framework. Nonetheless, this process is also energy demanding, with ΔG50 = 22.3 kcal/mol (considering 8 + 2 Me3SiH as the reference), as observed in equation 3. Combination of the silyl formate generation (equation 4) with the latter process would give rise to equation 5, which is also extremely energydemanding. Chapter 1. Results and discussion [122] -Even if the silyl formate was produced and released during catalysis, its reduction to the bis(silyl)acetal stage with the borane-silane complex [Me3Si−H···B(C6F5)3] would not be irreversible, since formation of the formaldehyde derivative would be only moderately exoergic (ΔG50 = -17.7 kcal/mol), i.e. the reverse step would be energetically feasible (equation 5). Scheme 27. Possible scenarios in which BCF could be released during catalysis (species with free BCF highlighted in red). ΔG50 in kcal/mol. For all the reasons stated above, the (PBP)Ni complexes are necessary in the proposed mechanism for carbon dioxide reduction, since they are able to activate CO2 efficiently, they mediate the hydrosilation processes and they control the release of the Lewis acid, avoiding in this way over-reduction phenomena. A scheme of the catalytic cycle (Scheme 28) and an overall energy profile (Figure 31) are shown below. Nickel(II) catalysed Carbon Dioxide Hydrosilation [123] Scheme 28. Proposed catalytic cycle for the carbon dioxide hydrosilation to the bis(silyl)acetal stage based on the experimental and theoretical data. Chapter 1. Results and discussion [124] Figure 31. Energy profile for the carbon dioxide hydrosilation to the bis(silyl)acetal stage catalyzed by (PBP)Ni complex 8. ΔZPE in kcal/mol (ΔG50 in parentheses). Nickel(II) catalysed Carbon Dioxide Hydrosilation [125] It can be observed in Figure 31 that there is a pattern in the reduction process, in which a silylium transfer always precedes the hydride attack, in a similar fashion to the systems described by Oestreich40 and Müller 71 (Scheme 29), where they describe carbon dioxide hydrosilation by using Ru−S systems and silyl cations, respectively. Scheme 29. Reduction steps in the hydrosilation of CO2 as proposed by Oestreich and Müller (adapted from references 40 and 71). 71 A. Schäfer, W. Saak, D. Haase, T. Müller, Angew. Chem. Int. Ed., 2012, 51, 2981-2984. Chapter 2. Introduction [132] 1. Introduction 1.1. Transition metal boryl complexes (M−BY2, Y = OR, NR2, alkyl…) Although the first studies on transition metal boryl complexes date back to the 1960s, 1 there has been a recent increasing interest on these compounds in the last couple of decades, since they have been widely proposed as reactive intermediates in the catalytic borylation or hydroboration of alkenes, alkanes and arenes. Given the high versatility of these resulting organoboron species for making a great diversity of new bonds, it is not surprising that studies on TM boryl complexes are becoming more frequent, so the better understanding on the bonding and reactivity can help in developing more selective and efficient catalytic processes. 2 1.1.1. Synthesis Structural features Before describing the most common approaches to generate metal-boryl bonds, it is worth mentioning that the strong σ-donor properties and trans influence of these ligands have consequences on the structure and geometry of the resulting complexes. The main reason for this phenomenon is the tendency of the ligands to rearrange inasmuch as possible to avoid steric hindrance and/or electronic repulsion (e.g. two strong trans influence ligands in trans position to each other). 3 1 H. Nöth, G. Schmid, Angew. Chem. Int. Ed. 1963, 2, 623. 2 a) S. Aldridge, D. L. Coombs, Coord. Chem. Rev., 2004, 248, 535-559; b) I. A. I. Mkhalid, J. H. Barnard, T. B. Marder, J. M. Murphy, J. F. Hartwig, Chem. Rev. 2010, 110, 890-931. 3 Z. Lin in Computational Studies in Organometallic Chemistry, S. A. McGregor & O. Eisenstein (Editors), Springer International Publishing, 2016. Chapter 2: Reactivities and Electronic Properties of Boryl Ligands, pages 39-58. Nickel(II) boryl complexes as reactive intermediates [133] Figure 1. Representative examples of the trans influence of boryl ligands and its repercussion on the geometry of transition metal complexes (adapted from reference 3). Bcat = catecholboryl. BCat’ = 4-tert-butylcatecholboryl. Figure 1 shows some cases where the presence of boryl ligands forces a certain geometry in a transition metal complex. For instance, platinum square planar complex I 4 displays a ligand arrangement in which the boryl moiety is located trans to the weakest trans influence ligand of the complex (i.e. chloride ligand). In this way, the two bulky phosphine groups are far enough from each other not to be sterically constrained. This is not the case for complex II, 5 where the electronic effects force the phosphines to be in a cis position regardless of the steric bulk. Octahedral complexes III 6 and IV 7 show the same phenomenon: in III, the boryl fragments adopt again a cis position, even though the carbonyl ligands are also strong trans influence groups. The same effect can be observed 4 W. Clegg, F. J. Lawlor, G. Lesley, T. B. Marder, N. C. Norman, A. G. Orpen, M. J. Quayle, C. R. Rice, A. J. Scott, F. E. S. Souza, J. Organomet. Chem., 1998, 550, 183-192. 5 G. Lesley, P. Nguyen, N. J. Taylor, T. B. Marder, A. J. Scott, W. Clegg, N. C. Norman, Organometallics, 1996, 15, 5137-5154. 6 X. He, J. F. Hartwig, Organometallics, 1996, 15, 400-407. 7 C. Dai, G. Stringer, T. B. Marder, A. J. Scott, W. Clegg, N. C. Norman, Inorg. Chem., 1997, 36, 272-273. Chapter 2. Introduction [134] in IV, where the BCat moieties are oriented in a fac arrangement. In square pyramidal complex V 8 the boryl fragment is placed in the apical position given that there is a vacant trans to it. Nonetheless, they occupy basal positions in the 17-electron complex VI, 9 since the cobalt atom possesses an electron in the apical position which is located in an orbital with antibonding character, precluding the presence of a boryl ligand in such zone. To the best of our knowledge, there is no example reported yet for a transition metal complex with two boryl units trans to each other. Synthetic approaches The most common strategies for the construction of metal−boryl bonds include the following: 10 - Attack of an anionic transition metal complex to a boron halide to give the corresponding metal−boryl fragment and a salt. - σ bond metathesis reaction between a M−R bond (e.g. M−alkyl, M−alkoxide) and a borane reagent. - Oxidative addition of a B−X bond to a transition metal, where X = H, Cl, Br, Sn or B. The last method, and especially the use of diboron(4) reagents is gaining increasing interest in the chemistry community because these species have proven very useful in synthetic chemistry, and they possess attractive features 8 a) C. E. F. Rickard, W. R. Roper, A. Williamson, L. J. Wright, Organometallics, 1998, 17, 4869- 4874; b) G. R. Clark, G. J. Irvine, W. R. Roper, L. J. Wright, J. Organomet. Chem., 2003, 680, 81-88. 9 a) C. Dai, G. Stringer, J. F. Corrigan, N. J. Taylor, T. B. Marder, N. C. Norman, J. Organomet. Chem., 1996, 513, 273-275; b) K. C. Lam, W. H. Lam, Z. Lin, T. B. Marder, N. C. Norman, Inorg. Chem., 2004, 43, 2541-2547. 10 C. S. Cundy, H. Nöth, J. Organomet. Chem., 1971, 30, 135-143. Nickel(II) boryl complexes as reactive intermediates [135] such as their ease to manipulate and their stability. 11 Specifically, the stability of diboron(4) Y2B−BY2 reagents against disproportionation decreases in the following order: Y = NR2, OR > aryl, alkyl, Cl. The high stability shown by tetraamino and tetra(alkoxy)diborons is due to the π orbital overlap and partial filling of the empty p-orbital of the boron atom. Whereas, B2Cl4 is extremely reactive (pyrophoric, and it undergoes partial decomposition to B4Cl4 above 0 °C), 12 B2(NR2)4 or B2(OR)4 possess an appropriate stability/reactivity ratio for synthetic purposes. 13 This gradient in reactivity and stability can be observed in Figure 2. Figure 2. Comparison of the chemical behaviour of some diboron(4) reagents depending on the substituents (adapted from reference 13). Depending on the environment surrounding them, diboron(4) reagents also react in a variety of modes so the B−B bond undergoes full cleavage. In this way, the presence of a M−R fragment in a transition metal complex can give rise to a metathetic pathway in which there is a ligand exchange, ending up with a metal−boryl bond (Figure 3A). Usually the R group is an alkoxide fragment, so the generation of a new B−O bond drives the reaction towards the products.13 Alternatively, the Lewis acidity of the boron atoms in these reagents make the diboron(4) compounds prone to react with Lewis bases. This 11 E. C. Neeve, S. J. Geier, I. A. I. Mkhalid, S. A. Westcott, T. B. Marder, Chem. Rev. 2016, 116, 9091-9161. 12 G. Urry, T. Wartik, R. E. Moore, H. I. Schlesinger, J. Am. Chem. Soc., 1954, 76, 5293-5298. 13 S. A. Westcott, E. Fernández, Adv. Organomet. Chem., 2015, 63, 39-89. Chapter 2. Introduction [136] interaction polarizes the B−B bond, so the resulting adduct is composed of a sp3 hybridized boron atom and a boryl-type fragment, characterized by a nucleophilic sp2 boron (Figure 3B). Several Lewis bases are known to react with these compounds, such as pyridines, alkoxides, amidines, phosphines or NHCs.11 Figure 3. Different methods of activating diboron(4) species (adapted from reference 13). If the diboron(4) is in the presence of a low-valent transition metal with labile ligands that can dissociate easily, oxidative addition events may take place, affording a metal bis(boryl) complex that might isomerize accordingly so as to avoid steric and electronic repulsions as stated above (Figure 3C). According to some authors, oxidative addition of B−X bonds is the most common strategy for the synthesis of transition metal boryl compounds and ‘this route has been exploited for elements of virtually all of the transition metal groups and in the synthesis of mono-, bis- and tris(boryl) derivatives.’2a, 14 14 According to some references (e.g. J. F. Hartwig, Organotransition Metal Chemistry: From Bonding to Catalysis, University Science Books, Sausalito, CA, 2009. Chapter 4, page 186), boryl complexes of all transition metals have been isolated except for group 3 and group 4 elements. However, recent works have reported on the syntheses of group 3 and group 4 boryl complexes. For group 3 boryl compounds, see: L. M. A. Saleh, K. H. Birjkumar, A. V. Protchenko, A. D. Schwarz, S. Aldridge, C. Jones, N. Kaltsoyannis, P. Mountford, J. Am. Chem. Soc., 2011, 133, 3836-3839. For group 4 boryl species, see: T. Terabayashi, T. Kajiwara, M. Yamashita, K. Nozaki, J. Am. Chem. Soc., 2009, 131, 14162-14163. Nickel(II) boryl complexes as reactive intermediates [137] Nonetheless, there are obviously some exceptions to this latter statement. For instance, the oxidative addition process is the reason why Pt(0) complexes can catalyse alkyne diboration reactions whereas Pd(0) cannot, even though they are highly active in silyl or stannylmetalation reactions. Following the work of Sakaki, 15 Musaev and Morokuma 16 carried out a DFT study in 1998 where they found a plausible explanation for this discrepancy in reactivity: in both cases the diboration reaction can occur by following the proposed cycle in Scheme 1, in which the first step involves the oxidative addition of the B−B bond. The activation energy for this step is 14.0 kcal/mol for Pt(0) and 8.6 kcal/mol for Pd(0). Although the transition state is easier to access for palladium, the resulting product (i.e. bis boryl complex) is 8.5 kcal/mol in energy above the origin, whereas in the case of platinum it is -7.2 kcal/mol. Due to the negligible reverse energy barrier (0.1 kcal/mol), the process does not take place and no catalysis is observed experimentally, given that this difference in energy remains for every intermediate of the catalytic cycle. 15 S. Sakaki, T. Kikuno, Organometallics, 1997, 36, 226-229. 16 Q. Cui, D. G. Musaev, K. Morokuma, Organometallics, 1998, 17, 742-751. Chapter 2. Introduction [138] Scheme 1. One of the proposed catalytic cycles for cis alkyne diborylation, as proposed by Musaev and Morokuma. Inset: scheme of the energies required for the oxidative addition of the B−B bond (adapted from reference 16). For this reason, it is more common to find σ-metathesis processes in palladium complexes than in platinum ones. In fact, going up from Pt in group 10 shows a trend in reactivity against tetra(alkoxy)diboron reagents: whereas Pt complexes usually react through oxidative addition processes, Pd 17 and Ni 18 often undergo σ-bond metathesis.13 The last approach for activating diboron(4) species by transition metal complexes involve the cooperativity of the ligand, and it is named 1,2 – addition (Figure 3D). In this strategy, the B−B bond is added across a metal-ligand, so each 17 a) J. Takaya, N. Kirai, N. Iwasawa, J. Am. Chem. Soc., 2011, 133, 12980-12983; b) N. Kirai, J. Takaya, N. Iwasawa, J. Am. Chem. Soc., 2013, 135, 2493-2496. Oxidative addition events usually take place when a B−X (X= halide, Sn) containing reagent is used. See: c) S. Onozawa, Y. Hatanaka, T. Sakakura, S. Shimada, M. Tanaka, Organometallics, 1996, 15, 5450-5452; d) S. Onozawa, M. Tanaka, Organometallics, 2001, 20, 2956-2958; e) H. Braunschweig, K. Gruss, K. Radacki, K. Uttinger, Eur. J. Inorg. Chem., 2008, 1462-1466. 18 B. L. Tran, D. Adhikari, H. Fan, M. Pink, D. J. Mindiola, Dalton Trans., 2010, 39, 358-360. Nickel(II) boryl complexes as reactive intermediates [139] fragment of the bond possesses one boryl moiety bound to it. The only examples in the literature of activation of B−X bonds by this mechanism are displayed in Scheme 2. The first two reactions involve the activation of hydroboranes across a metal−oxygen bond to yield the rare peroxoborate (A) 19 and borate (B) 20 species and metal−hydride unions in each case. It can be seen in all of the reactions that monoactivation of B−H species across M−L bonds proceed under mild conditions unlike reaction C, in which B−B cleavage requires some energy to take place. In this example (which is the first case where 1,2 - addition occurs across a M−N bond) it is possible to obtain in one step two boryl functionalities in different regions of the same complex. 21 19 M. A. Salomon, T. Braun, A. Penner, Angew. Chem. Int. Ed., 2008, 47, 8867-8871. 20 A. C. Fernandes, J. A. Fernandes, F. A. A. Paz, C. C. Romão, Dalton Trans., 2008, 6686-6688. 21 Y. Zhu, C-H. Chen, C. M. Farfard, B. M. Foxman, O. V. Ozerov, Inorg. Chem., 2011, 50, 7980- 7987. Chapter 2. Introduction [140] Scheme 2. Examples of B−H or B−B activation by following a 1,2 - addition mechanism (adapted from references 19, 20 and 21). Some other examples of metal−ligand cooperativity and 1,2 - addition of small molecules like dihydrogen or hydrosilanes can be found in the literature, 22 albeit no B−B activation mediated by a M−B bond has been described to date. 22 a) J. R. Khusnutdinova, D. Milstein, Angew. Chem. Int. Ed., 2015, 54, 12236-12273; b) G. R. Owen, Chem. Commun., 2016, 52, 10712-10726. Nickel(II) boryl complexes as reactive intermediates [141] Nickel boryl complexes The first examples of Nickel boryl complexes were synthesized in the 1960s and 1970s, when Nöth and co-workers obtained the first Ni derivatives of this kind by transmetalation reactions with cobalt complexes 23 and by mixing Ni(0) precursors with alkyl- and haloboranes.10 However, these compounds lacked exhaustive characterization, and analytical data that could provide conclusive structural confirmation was scarce. Approximately ten years ago, the group of Prof. Mindiola synthesized the first Ni boryl complex whose structure was completely confirmed by several methods, including X-Ray crystallography. This complex is stabilized by the same PNP scaffold as the Pd species developed by Ozerov et al. (Scheme 2C).21 By mixing catecholborane with a (PNP)Ni−H complex they could obtain VII, releasing hydrogen in the process. DFT studies revealed that no backbonding from the metal to the boron atom was occurring, what made the authors think that the boryl fragment could be transferred to other chemicals. In this fashion, they could employ VII as a recyclable boryl reagent for the borylation of bromobenzene (Scheme 3). 24 23 a) G. Schmid, H. Nöth, Chem. Ber., 1967, 100, 2899-2907; b) G. Schmid, P. Powell, H. Nöth, Chem. Ber., 1968, 101, 1205-1214. 24 D. Adhikari, J. C. Huffman, D. J. Mindiola, Chem. Commun., 2007, 4489-4491. Chapter 3. Results and discussion [244] Figure 23. 1H NMR spectrum (400 MHz, CD2Cl2, -30 °C) of complex 19·Et3SiH. Inset (bottom right): HMQC 1H−29Si experiment, where the cross-peak of the bridging hydride with the Si atom can be observed. Complex 19·Et3SiH is an excellent example to be compared to the previously reported complex Ir·HSiEt3 described by Brookhart et al.,27 given that in both cases Et3SiH is employed as silane. The 1H−29Si HMQC spectrum shows a cross-peak of the bridging hydride with a signal at 11.5 ppm, which is shifted 11.5 ppm downfield with respect to that of the free silane. 56 This phenomenon reflects the deshielding of the Si atom in the presence of the Lewis-acidic Pt(II) species, which enhances the electrophilicity of the silane.32 In addition, both Pt and Ir systems display the same 1H chemical shifts for the bridging hydride (-4.9 56 Free Et3SiH: 1H NMR (600 MHz, 213 K, CD2Cl2): δ 3.50 (sept, 1 H, 1JSi-H = 173.6 Hz) ppm, 29Si NMR (119 MHz, 213 K, CD2Cl2): δ 0.0 (s) ppm. See: M. Hamdaoui, C. Desrousseaux, H. Habbita, J-P. Djukic, Organometallics, 2017, 36, 4864-4882. Cationic Pt(II) σ-SiH complexes as reactive intermediates [245] ppm) with identical coupling constants to 29Si (1JSi-H = 79 Hz), which points out to a η1 coordination mode, since a smaller J value would suggest a stronger debilitation of the Si−H bond resulting from a higher degree of backbonding from the metal atom, typical for η2 Si−H compounds. 57 These adducts are stable at low temperatures (usually below 0 °C). Therefore, in order to isolate any of these species and to confirm their geometry, numerous crystallization attempts were performed by trying different silanes, solvents and crystallization conditions. Fortunately, crystals of 19·Et3SiH were stable enough to be measured by X-Ray diffraction methods (Figure 24). Figure 24. X-Ray and calculated structures of 19·Et3SiH. BAr4F anion and all hydrogen atoms except the bridging hydride have been omitted for clarity. As previously observed by NMR experiments, the crystal structure shows a trans arrangement of both NHCs, being one of them cyclometalated. The methylene moiety is located trans to the hydrogen atom of Et3SiH (C7−Pt1−H1Si: 173.1(1)°), which was located in the Fourier map. It is bound to Si (H1Si−Si1: 57 W. Scherer, P. Meixner, K. Batke, J. E. Barquera-Lozada, K. Ruhland, A. Fischer, G. Eickerling, K. Eichele, Angew. Chem. Int. Ed., 2016, 55, 11673-11677. Chapter 3. Results and discussion [246] 1.66(3) Å) and Pt (Pt1−H1Si: 1.58(3) Å) forming an angle of 103(2)°. This value suggests that 19·Et3SiH possesses a silane geometry intermediate between η1 and η2 coordination modes. However, there seems to be no formal Pt−Si bond since the distance found (2.53(1) Å) is longer than the sum of the covalent radii for Pt and Si (2.47 Å)45 and that found for Pt(II)−silyl derivatives (~2.3 Å, Figure 14). Moreover, the formation of an η2 complex would place the C−Pt bond in the bisection of the Si−H linkage (Figure 25, right). Instead, the experimental structure shows the initial stage of this process (given the C−Pt−Si angle), since the first step involves the approach of the silane through the hydrogen atom (with hydride character, Figure 3) prior to the pivoting motion around such bridging hydride (Figure 25).15 Figure 25. Schematic view of the motion between η1 and η2 coordination modes in a σ-SiH complex. The detection and isolation of this type of complexes demonstrates that ligand design and the Lewis acidity of the metal are crucial. In fact, this contradicts several other strategies employed for avoiding η2 geometries and/or Si−H bond cleavage, such as a) the use of 1st row transition metals (since their contracted d-orbitals do not overlap enough with the ligand orbitals); b) the presence of π acidic auxiliary ligands like CO or PF3 (which delocalize the metal electrons); c) high oxidation states on the metal (which contract the d-orbitals).15 Cationic Pt(II) σ-SiH complexes as reactive intermediates [247] Computational studies Nonetheless, DFT studies 58 were carried out on this system in order to gather additional data that could help to explain the interaction between Et3SiH and complex 19. In a first approach, the geometry of 19·Et3SiH was optimized, and an excellent agreement between the calculated and the experimental structure was observed (Figure 24 and Table 5), with slight disparity in the Pt−Si and Pt−H bond lengths. Thus, the selected functional and basis set accurately reproduce this system. The presence of the non-coordinating anion BAr4F produces an insignificant effect on the geometrical parameters of the sigma complex (Table 5). X-Ray DFT (no BAr4F) DFT (BAr4F) C−Pt (Å) 2.082(2) 2.084 2.082 Pt−H (Å) 1.58(3) 1.716 1.726 Si−H (Å) 1.66(3) 1.651 1.642 Pt−Si (Å) 2.532(2) 2.631 2.642 Pt−H−Si (°) 102.9 102.8 103.3 C7−Pt−H (°) 173.1(11) 171.3 172.2 C−Pt−C (°) 166.74(9) 166.9 167.3 Table 5. Comparison of the experimental and calculated geometrical parameters of complex 19·Et3SiH in the presence and absence of the BAr4F anion. The QTAIM methodology (Quantum Theory of Atoms In Molecules) was applied to perform a topological analysis of the electron density of this complex. However, studying the electron density function is not appropriate, since it does not define spatial regions beyond the atomic center. Instead, the laplacian of the electron density (∇2ρ(r)) is usually interrogated, given that it measures its local 58 Calculations were carried out at the TPSSh-D3/6-31g(d,p)/SDD//TPSSh-D3/6- 311++G(2d,p)/def2QZVP level of theory. SMD (Dichloromethane) continuum was used to model the solvent. Chapter 3. Results and discussion [248] concentration or depletion. Thus, it can be employed to identify electron pair localization. In fact, correlation between QTAIM analysis and the VSEPR model (Valence Shell Electron Pair Repulsion) is accurate enough (yet not perfect) to explain different chemical phenomena. 59 Therefore, the laplacian of the electron density was studied on 19·Et3SiH and the two other crystallographically characterized σ-SiH complexes Ru·HSiCl3 and Ir·HSiEt3 for comparison reasons. The outcome of this investigation is shown in Figure 26: Figure 26. DFT calculated ∇2ρ(r) maps of σ-SiH complexes 19·Et3SiH, Ir·Et3SiH and Ru·Cl3SiH in the plane defined by the metal, silicon, and hydrogen atoms. Bond paths are displayed as black solid lines and bond critical points (BCP) as blue circles. As it can be seen above, the Pt and Ir complexes contain atomic interaction lines between Pt and H, and Si and H, each of them containing in turn a bond critical point (BCP). 60 On the other hand, there is neither bond path nor critical point between Pt and Si. 61 The electron density at the BCPs (M−H: ρ(r) = 1.06/0.62 e·Å-3, Si−H: ρ(r) = 0.65/0.70 e·Å-3 for 19·Et3SiH and Ir·Et3SiH, respectively), combined with the aforementioned data, point out to η1 59 P. L. A. Popelier, Coord. Chem. Rev., 2000, 197, 169-189. 60 A BCP indicates a bonding situation between two nuclei. See reference 59 for more details. 61 The same study was performed on the experimental structure derived from X-Ray diffraction studies, and the same topological map was obtained (i.e. no BCP was found between Pt and Si). Cationic Pt(II) σ-SiH complexes as reactive intermediates [249] coordination modes in both complexes. However, the Pt complex shows the initial stage of the pivoting motion leading to a η2 binding fashion, which is in agreement with the experimentally found bond lengths and angles, like the one formed by Pt, H and Si (102.9 °), for example. In the Ru complex, the opposite situation is found: there is no BCP or bond path between Si and H, and there seems to be a Pt−Si and a Pt−H bond. This suggests a complete cleavage of the Si−H bond (i.e. oxidative addition), and not a η2 σ-SiH complex. All these data are in accordance with the calculated Wiberg Bond Indices and natural charges resulting from NBO analysis (Table 6), which clearly show that coordination modes close to η1 geometries enhance the electrophilicity of the silane. Wiberg Bond Index Natural Charge 19·Et3SiH Si−H: 0.45 Si: 1.682 Si−Pt: 0.53 Pt: -0.078 Pt−H: 0.58 H: -0.123 Ir·Et3SiH Si−H: 0.58 Si: 1.738 Si−Ir: 0.23 Ir: -0.334 Ir−H: 0.39 H: -0.266 Ru·Cl3SiH Si−H: 0.24 Si: 1.341 Si−Ru: 1.22 Ru: -1.406 Ru−H: 0.70 H: 0.140 Table 6. Wiberg Bond Indices and Natural Charges of the atoms involved in the interactions describing the σ-SiH complex formation of the different compounds. Influence of the silane The influence of the silane in the coordination mode was also explored by testing the coordination capabilities of Ph3SiH to complex 19. Single crystals of the σ-SiH complex 19·Ph3SiH could be grown by solvent diffusion techniques (dichloromethane/pentane) at -30 °C. The crystals were of enough quality to perform X-Ray diffraction studies. Interestingly, overlap of two different Chapter 3. Results and discussion [250] structures was found in the unit cell: sigma complex 19·Ph3SiH and silyl derivative 21 [Pt(SiPh3)(ItBuiPr)2][BAr4F] 62 (Figure 27, left). 62 Formation of silyl complex 21 is a consequence of the reactivity patterns displayed by the σ-silane complexes studied in this thesis. This behaviour will be described in detail in the next section of the chapter. Cationic Pt(II) σ-SiH complexes as reactive intermediates [251] Figure 27. Left: Overlaid experimental crystal structure of complexes 19·Ph3SiH and 21. Center: Top view of both complexes, comparison of dihedral angles and Pt−Si bond lengths. Right: ORTEP representations of complex 19·Ph3SiH (after removal of atoms C5 and C6) and complex 21 (independently prepared and crystallized). Chapter 3. Results and discussion [252] Although the crystal structure might look as crystallographically disordered at a first glance, a closer look seems to point out to the presence of two overlaid structures, in which the only variations are the presence or absence of the bridging hydride and the labeled carbon atoms in Figure 27 (left) corresponding to one of the tert-butyl groups of one NHC ligand. A careful examination of the angles between them suggests that C5, C6 and C7 belong to one of the non-cyclometalated carbenes of complex 21, and C5A, C6A and C7 belong to the cyclometalated arm of the σ-SiH complex 19·Ph3SiH. Analysis of the independently synthesized and crystallized silyl complex 21 (Figure 27, right) provides additional support to this theory. Three other structural parameters back the hypothesis of having two overlaid geometries: a) The bridging hydride could be found in the Fourier map, trans to the methylene group (C5A−Pt−H angle: 177.99°); b) Both Pt−Si distances only differ in 0.05 Å; c) The dihedral angles between the NHC ligands are almost identical. It can be seen that the Pt−Si distance and the Pt−H−Si angle in 19·Ph3SiH (2.357 Å and 87.7°) considerably vary from those obtained by DFT methods (2.534 Å and 97.5°, Table 7), probably because of the overlap effect of both structures. In any case, topological analysis was performed on the experimental geometry to check the presence of BCP between Pt−Si, in order to discern the binding mode of the silane. It is interesting to notice that despite having the same Pt−Si bond length, experimental 19·Ph3SiH does not show any BCP between the metal and silicon, but complex 21 does (Figure 28). 63 63 A BCP is observed between Pt and Si in the calculated structure of 19·Ph3SiH upon further closing the Pt−H−Si angle to around 80°, which suggests that the situation observed above is close to the threshold for a BCP to appear. Cationic Pt(II) σ-SiH complexes as reactive intermediates [253] Figure 28. DFT calculated ∇2ρ(r) maps of σ-SiH complexes 19·Ph3SiH and 21 in the plane defined by the metal, silicon, and hydrogen atoms. Bond paths are displayed as black solid lines and bond critical points (BCP) as blue circles. Following the study of the influence of the silane in the coordination mode of the resulting σ-SiH complexes, DFT calculations on different 19·R3SiH complexes (R = Et, Ph or Cl) showed that electron-withdrawing substituents on the Si atom induce a higher degree of backbonding from the metal (following the trend SiCl3 > SiPh3 > SiEt3), hence there is a lower positive charge on Si and the Si−H bond is weakened in favour of the formation of Pt−Si and Pt−H bonds. This transition can be observed in Table 7, where smaller Pt−H−Si angles and Pt−Si distances and longer Si−H bond lengths can be found with increasing electronegative substituents on the silane. This situation is extreme in the case Chapter 3. Results and discussion [260] Figure 33. Relaxed Potential Energy Surface (PES) Scan of complex Ir·Et3SiH along the Ir−H−Si coordinate. The energy minimum is located at 169° (The X-Ray structure showed a Ir−H−Si angle of 157°, which requires less than 1 kcal/mol). Figure 34. Relaxed Potential Energy Surface (PES) Scan of complex Ru·Cl3SiH along the Ru−H−Si coordinate. The energy minimum is located at 83°. (The X-Ray structure showed a Ru−H−Si angle of 87°, which requires less than 1 kcal/mol). The absence of data between 120° and 160° is due to the fact that convergence criteria were not reached during the PES calculation. Cationic Pt(II) σ-SiH complexes as reactive intermediates [261] This section has shown how minor and subtle modifications on the structure of the NHC ligands can dramatically change the stability of the corresponding Pt(II) σ-SiH species. However, these compounds are only stable at low temperatures (usually below 0 °C), and they are reactive intermediates that evolve at room temperature to different products depending on the silane used. The following sections will describe in detail the nature of these resulting derivatives (Scheme 13). Although Scheme 13 shows the example of cyclometalated ItBuiPr complexes, it will be shown below that different compounds such as hydride species containing different types of NHC ligands can also give rise to σ-SiH complexes which undergo subsequent transformations. Scheme 13. σ-silane complexes as reactive intermediates in the formation of different siliconcontaining Pt(II) species. 2.3. Reactivity of the σ-SiH complexes using tertiary silanes Substitution of the ItBu carbene for the smaller version ItBuiPr gave rise to more stable σ-SiH complexes, which allowed their study and characterization at temperatures below 0 °C. In agreement with the theoretical predictions, tertiary silane-based σ-complexes rearrange above this temperature into the 14- electron Pt(II) silyl species (Scheme 14), in the same way as 13 reacts with Et2SiH2 to yield complex 16. Chapter 3. Results and discussion [262] Scheme 14. Evolution of the tertiary silane-based σ-SiH complexes 19 into the corresponding 14-electron Pt(II) silyl derivatives 20 and 21. When 19·Et3SiH was left to warm to r.t., a color change to intense yellow was observed. This change in appearance was accompanied by the emergence of a new set of signals by 1H NMR, which corresponded to a complex possessing equivalent NHC ligands (Figure 35). Complex 20 can be made by stirring complex 19 with 1.1 equivalents of Et3SiH in dichloromethane at room temperature for 5 hours, and it can be purified by crystallization techniques through solvent diffusion at 0 °C. Unfortunately, the resulting crystals were not adequate for X- Ray diffraction studies. Cationic Pt(II) σ-SiH complexes as reactive intermediates [263] Figure 35. 1H NMR spectrum (400 MHz, CD2Cl2) of the reaction mixture to obtain complex 20. Inset (bottom): HMBC 1H−29Si experiment, where the cross-peaks of the ethyl chains with the Si atom can be observed. As shown above, the reaction is extremely clean and selective towards the formation of C−H and Pt−Si bonds, which is rather surprising considering the possibility of obtaining mixtures of C−Si and C−H products, as reported for Ir(III) species, 70 (Figure 36A) or observing selective C−Si coupling phenomena, as described by Milstein et al. for a different class of Pt(II) complexes in the presence of tertiary alkyl and aryl silanes 71 (Figure 36B). 70 M. Aizenberg, D. Milstein, J. Am. Chem. Soc., 1995, 117, 6456-6464. 71 M. E. van der Boom, J. Ott, D. Milstein, Organometallics, 1998, 17, 4263-4266. Chapter 3. Results and discussion [264] Figure 36. Competitive C−H and C−Si coupling (A) and selective C−Si coupling (B) displayed by Pt(II) complexes, as described by Milstein et al. (adapted from references 70 and 71). It has also been described that the selectivity towards either C−H or C−Si coupling can be modulated depending on the silane employed (e.g. alkylsilanes and arylsilanes behave differently). 72 For this reason, Ph3SiH was used so as to explore a potential different outcome. A similar behavior was found, since the Pt(II) silyl derivative 21 was exclusively formed in two hours by using similar experimental conditions as with Et3SiH (Figure 37). 73 Although the patterns governing the selectivity of this type of reactions are not clear yet, a plausible explanation might be attributed to steric bulk: in 13 the ItBu groups are too large to accommodate any tertiary silane, given that no interaction or reaction is experimentally observed. Perhaps, ItBuiPr provides enough space for the hydride to migrate to the carbon atom, but not enough for the silane, regardless of the orientation of the NHC ligands (Figure 38). 72 M. Aizenberg, D. Milstein, Angew. Chem. Int. Ed., 1994, 33, 317-319. 73 As shown above in section 2.2, complex 21 could be crystallized and analyzed by X-Ray diffraction methods (Figure 27). Cationic Pt(II) σ-SiH complexes as reactive intermediates [265] Figure 37. 1H NMR spectrum (400 MHz, CD2Cl2) of complex 21. Inset (bottom): HMBC 1H−29Si experiment, where the cross-peaks of the phenyl protons with the Si atom can be observed. Figure 38. Plausible explanation based on steric hindrance for the selective formation of the cationic Pt(II) silyl derivatives stabilized by ItBuiPr ligands. Chapter 3. Results and discussion [266] 2.4. Reactivity of the σ-SiH complexes using primary silanes Whereas the use of ItBuiPr-stabilized Pt(II) complexes with tertiary silanes could only lead to the formation of either Pt−silyl derivatives or Pt−hydride compounds with the silane moiety attached to one of the carbenes (not observed), the availability of additional Si−H bonds in primary silanes allowed us to explore new potential patterns of reactivity by using these complexes. Cyclometalated cationic Pt(II) silyl complexes When complex 19 was mixed with either nBuSiH3 or PhSiH3 in CH2Cl2 at room temperature, a quick change of colour was observed from orange to intense yellow along with bubble formation, indicative of gas release. NMR analysis of both reactions revealed the clean and selective formation of an asymmetric species, in which one of the NHC ligands seems to be cyclometalated, according to the signals observed by 1H NMR for the tert-butyl and iso-propyl peaks, the inequivalence of the imidazole protons and the presence of a methylene signal directly attached to Si (example of the reaction between 19 and nBuSiH3 in Figures 39 and 40). The 1H NMR spectrum also shows a peak at 4.60 ppm that has been attributed to H2. All these data suggest that the silane molecule has inserted into the Pt−C bond, releasing hydrogen gas during the process (coming from the two other protons of the silane) to give complexes 22 and 23, as depicted in Scheme 15. The signal for the SiH proton correlates via HMQC 1H−29Si with a silicon peak at -1.2 ppm, and it shows satellite peaks due to coupling with 195Pt (2JPt-H = 120 Hz), similar to those described for complex 16 (Figure 13). 74 Additionally, the SiH proton of complex 23 appears as a triplet (3JH- 74 For NMR data regarding neutral silyl Pt(II) complexes, see: a) D. Chan, S. B. Duckett, S. L. Heath, I. G. Khazal, R. N. Perutz, S. Sabo-Etienne, P. L. Timmins, Organometallics, 2004, 23, 5744-5756; b) J. C. DeMott, W. Gu, B. J. McCulloch, D. E. Herbert, M. D. Goshert, J. R. Cationic Pt(II) σ-SiH complexes as reactive intermediates [267] H = 3 Hz), consistent with the proximity of the CH2 group of the cyclometalated arm. Figure 39. 1H NMR spectrum (400 MHz, CD2Cl2) of complex 22 from the crude reaction mixture. Walensky, J. Zhou, O. V. Ozerov, Organometallics, 2015, 34, 3930-3933; c) C. Mitzenheim, T. Braun, R. Laubenstein, B. Braun, R. Herrmann, Dalton Trans., 2016, 45, 6394-6404. Chapter 3. Results and discussion [268] Figure 40. 13C{1H} NMR spectrum (400 MHz, CD2Cl2) of complex 22 from the crude reaction mixture. (Top left): 13C{1H} NMR spectrum, showing the region of the methylene unit bound to Si, which displays satellite peaks due to coupling with 195Pt. (Top center): HMQC 1H−29Si experiment, where the cross-peak of the SiH proton with the Si atom can be observed (free nBuSiH3 resonates at -60.1 ppm in 29Si NMR). Scheme 15. Reaction of complex 19 with primary silanes to give silyl-cyclometalated complexes 22 and 23. Cationic Pt(II) σ-SiH complexes as reactive intermediates [269] As shown in Figures 39 and 40, there seems to be only one species, with none or negligible trace of C−H coupling resulting derivatives, which highlights the drastic difference in reactivity depending on the type of silane employed, regardless of the nature of the substituents on the Si atom. As it can be seen, this process is rather complex and requires the cleavage of two Si−H bonds and the formation of a C−Si and a Pt−Si bonds with release of a H2 molecule. The structure of phenylsilane derivative 23 could also be confirmed by X-Ray crystallography (Figure 41). It can be seen that a dichloromethane molecule (coming from the crystallization solvent) interacts with the platinum atom (Pt···Cl distance: 2.803 Å), although in a weak manner, since elemental analysis does not reveal the presence of any solvent molecule in the results. The insertion of the silane molecule into the Pt−C bond gives rise to a 6-membered metalacycle that forces the carbene ligands to be almost coplanar (~18°). The Pt−Si bond length (2.269 Å) is comparable to that displayed by similar cationic silyl Pt(II) systems like complex 16 or complex 21 despite the metalacycle constraint. Figure 41. X-Ray structure of 23. BAr4F anion, all hydrogen atoms except the SiH proton have been omitted and some carbon atoms are shown in a wireframe format for clarity. Chapter 3. Results and discussion [276] no CH group in a close position to the metal center to consider the possibility of agostic interactions stabilizing the complex. The Pt−Si bond length possesses the same value (2.270 Å) as that observed for the aforementioned silyl compounds and again, the NHCs arrange themselves with a wide dihedral angle (up to 123°) so as to diminish steric hindrance. Figure 46. X-Ray structure of complex 36. BAr4F anion and all hydrogen atoms except the SiH protons have been omitted for clarity. Selected bond lengths (Å) and angles (°): Pt1−Si1: 2.2703(9); Si1−H1si: 1.39(3); Si1-H2si: 1.44(3); Ccarbene−Pt−Ccarbene: 174.78(12); Dihedral angle (NHC planes): 123.4. 2.5. Mechanistic studies on the C−Si and/or Pt−Si bond formation mediated by σ-SiH complexes Low temperature NMR experiments In order to get insight about the mechanism by which the C−Si and/or Pt−Si coupling events take place, several cationic Pt(II) complexes were Cationic Pt(II) σ-SiH complexes as reactive intermediates [277] evaluated by means of low temperature NMR experiments. Among them, the results obtained with cyclometalated species [Pt(IMes*’)(IMes*)][BAr4F], 25, and nBuSiH3 are very informative: first, addition of nBuSiH3 to a solution of 25 in CD2Cl2 at -80 °C leads to the formation of the σ-SiH complex 25·nBuSiH3. As shown in Figure 47, the bridging hydride resonates around -8 ppm in the 1H NMR spectrum, 78 and it shows satellite peaks due to coupling with 29Si (1JSi-H = 67 Hz) and 195Pt (1JPt-H = 500 Hz). The latter coupling constant is somewhat higher than that observed for 19·Et3SiH (1JPt-H = 396 Hz), which suggests a stronger interaction with the metal center. Figure 47 also shows that the σ complex is stable from -80 °C up to -15 °C, where an intermediate species (Intermediate A) appears. 78 The terminal hydrogen atoms resonate at 3 and 3.84 ppm. The three SiH protons correlate with a peak at -53.5 ppm in the 29Si spectrum (HMQC 1H−29Si). Additionally, NOESY experiments reveal exchange between bridging and terminal protons, suggesting a fluxional process. Chapter 3. Results and discussion [278] Figure 47. 1H NMR spectra (400 MHz, CD2Cl2) of the mixture of 25 and nBuSiH3 from -80 °C to - 15 °C, showing the formation of the σ-SiH complex 25·BuSiH3 and the appearance of Intermediate A at -15 °C. The sample was kept to this temperature (ca. 1h) so as to observe the transformation of 25·BuSiH3 into A (Figure 48), which shows a broad signal with Pt satellites integrating for 2 protons at -5.9 ppm (1JPt-H = 990 Hz). A NOESY experiment reveals that this signal exchanges with a multiplet (which integrates for 1 proton) at ~4 ppm. Unfortunately, additional information of the new appearing intermediate A was not possible since HMQC 1H−29Si did not show any cross-peak at that temperature (probably due to signal broadening) and the rest of the spectrum was too complex to extract any information from it. Nevertheless, DFT calculations (see below in this section) indicated that intermediate A seems to be the agostic SiH platinum hydride shown in Scheme 18. Cationic Pt(II) σ-SiH complexes as reactive intermediates [279] Scheme 18. Intermediates observed during the low temperature NMR experiment in the reaction between 25 and nBuSiH3. Figure 48. 1H NMR spectra (400 MHz, CD2Cl2) of the mixture of 25 and nBuSiH3 from -30 °C to - 15 °C, showing the growth of intermediate A and the disappearance of 25·nBuSiH3 at -15 °C. In fact, cooling back the sample down to -80 °C makes the signal of intermediate A split into two different broad peaks, each one of them with satellite peaks to platinum (Figure 49). The first one, at -4.74 ppm, possesses a 1JPt-H ≈ 390 Hz and it shows (along with the signal a 3.97 ppm) a cross-peak in the Chapter 3. Results and discussion [280] HMQC 1H−29Si spectrum with a signal at -21.0 ppm. The values of the chemical shift and the coupling constant are similar to those shown by the bridging hydrides of the σ-SiH complexes displayed in Table 4. Thus, it seems reasonable to assign the signal at 3.97 ppm to proton Ha and the signal at -4.74 ppm to proton Hb (Figure 49). Regarding the peak resonating at -6.88 ppm, it also integrates for one proton, yet it does not show any appreciable correlation on the HMQC 1H−29Si experiment. Besides, its coupling constant to platinum is much higher (1JPt-H ≈ 1500 Hz), typical from a cationic Pt(II) hydride with the coordination site trans to the hydride occupied by another ligand,35a,36 as shown in the examples of Figure 50. Therefore, the signal at -6.88 ppm was assigned to proton Hc. Figure 49. Portion of the 1H NMR spectrum (400 MHz, CD2Cl2) of the mixture of 25 and nBuSiH3 at -80 °C after consuming most of 25·BuSiH3 into intermediate A at -15 °C. Inset: zoom-in of the signals assigned to protons Hb and Hc, showing satellite peaks due to coupling with 195Pt. Cationic Pt(II) σ-SiH complexes as reactive intermediates [281] Figure 50. Cationic Pt(II) hydrides previously synthesized in our group with the coordination site trans to the hydride involved in π or σ interactions. Nevertheless, an additional experiment was performed so as to get more NMR data supporting the assignment of intermediate A. Hydride 34 was mixed with nBuSiH3 in order to form the corresponding σ-SiH complex 34·BuSiH3 (which resembles the proposed structure for intermediate A). Three different signals were observed when the temperature reached -90 °C. The first one, a broad singlet at 3.71 ppm, integrates for 2 protons and it shows a cross-peak with a signal at -49.80 ppm in the HMQC 1H−29Si. This signal was attributed to terminal protons Hte (Figure 51). The second one, a doublet (integrates for 1 H) at -5.55 ppm, correlating with the same Si signal in the HMQC experiment, is coupled to 195Pt (1JPt-H = 423 Hz) and to another proton (2JH-H = 17 Hz). The similarity in the chemical shift and the Pt−H coupling constant bear resemblance to Hb in intermediate A. Based on this, this signal was attributed to the bridging proton Hbr. Finally, a doublet integrating for 1 H at -6.08 ppm does not show any correlation in the HMQC experiment, it shows the same H−H coupling constant than proton Hbr (17 Hz) and it is coupled to the metal with a greater coupling constant (1JPt-H = 1544 Hz). These data suggest that this signal corresponds to the hydride directly bound to the platinum atom, analogous to proton Hc in intermediate A. Additionally, all three signals show exchange in the NOESY Chapter 3. Results and discussion [282] experiment, which is in accordance with the proposed structure depicted in Figure 51 for the σ-SiH complex 34·BuSiH3. As the temperature increases, the two signals in the hydride region collapse into a single broad signal at -60 °C (with coupling to 195Pt of 1544 Hz) until it coalesces at about -40 °C. This behaviour, and all NMR data discussed above, provides additional support for the previous assignment of intermediate A as a platinum hydride agostic SiH complex. Complex 34·BuSiH3 was transformed into silyl derivative 37 upon warming to room temperature, with concomitant loss of H2. Figure 51. Hydride region of the 1H NMR spectra (400 MHz, CD2Cl2) of the mixture of 34 and nBuSiH3 from -90 °C to -40 °C, where the formation of the σ-SiH complex 34·BuSiH3 can be observed. Inset: HMQC 1H−29Si experiment performed at -90 °C. Cationic Pt(II) σ-SiH complexes as reactive intermediates [283] DFT calculations The mechanism of the formation of C−Si/Pt−Si complexes 26-29, 31-32 and 35-38 was investigated by DFT methods 79 in order to gather additional information that can complement the experimental observations so as to propose a pathway by which these phenomena might be taking place, and to explain the remarkable selectivity experimentally obtained. This work was made in collaboration with Prof. Agustí Lledós from Universitat Autónoma de Barcelona (UAB), who ran the calculations. Cyclometalated Pt(II) complexes with primary silanes In this study, IMes derivative 24 and nBuSiH3 were chosen as models, since the experimental outcome is known (i.e. formation of complex 26). Four different alternatives were considered, which consist of either C−H bond formation with concomitant generation of silyl complex 35 or C−Si coupling to yield the cyclometalated C−Si derivative 26. 80 Both of them were calculated by considering pathways in which the NHC ligands initially adopt a trans or a cis arrangement in the course of the reaction. The four alternative pathways are summarized in Scheme 19, and might involve σ-bond metathesis processes or Pt(IV) intermediates: 79 Calculations were carried out at the M06/6-31g(d,p)/SDD//M06/6-311++g(d,p)/SDD level of theory. SMD (Dichloromethane) continuum was used to model the solvent. 80 Mechanistic pathways involving silylene species were also taken into account, yet their energy barriers were too high to be considered as experimentally feasible. For the sake of simplicity, these preliminary results are not included in this thesis. Chapter 3. Results and discussion [284] Scheme 19. Summary of the 4 different pathways calculated in this work for the reaction between 24 and nBuSiH3.  a) C−H coupling (NHCs in trans) Considering cyclometalated complex 24 and butylsilane as the energy reference (Figure 52), formation of the σ-SiH complex 24·BuSiH3 (Pt−Hbridging distance = 1.859 Å; Si−H = 1.549 Å; Pt−H−Si angle = 121.2°) only requires 0.4 kcal/mol (ΔG273 K), 81 after which oxidative addition of one of the Si−H bonds of nBuSiH3 can take place through transition state TS1, demanding 29.7 kcal/mol for the process to occur. No Pt(IV) species was found after the transition state as a local energy minimum, but the system directly evolved towards the formation of silyl complex 35, which is 22.2 kcal/mol more stable than the reagents.  b) C−H coupling (NHCs in cis) Starting from the same energy reference, this pathway is slightly more complex due to the initial isomerization step required to achieve a cis orientation of the carbene ligands (Figure 53). In this process, isomerization of 24 (C−Pt−C angle of 175.1°) to 24cis (C−Pt−C = 118.6°) requires 13.9 kcal/mol (TSisom1, 81 It is worth remembering that the formation of the σ-SiH complexes is slightly endergonic due to entropic factors (see section 2.1) but exothermic enough to observe their formation at low temperature. Cationic Pt(II) σ-SiH complexes as reactive intermediates [285] C−Pt−C angle = 135.5°) and formation of the σ-SiH complex 24cis·BuSiH3 (Pt−Hbridging distance = 1.633 Å; Si−H = 1.824 Å; Pt−H−Si angle = 90.6°) leaves the system 12.1 kcal/mol above the origin. It can be observed from both systems that the complex with a cis arrangement gives rise to a σ-SiH complex with much more η2 character and greater weakening and elongation of the Si−H bond, probably due to the bigger size of the cavity available for the metal−silane interaction. From here, one of the protons of the silane is close to the methylene group of the cyclometalated carbene (C−Hsilane distance = 1.795 Å) and it can bounce to form a methyl fragment at the same time that the silane establishes a new bond with the metal (Pt−Si = 2.403 Å). This step (TS2) has an energy cost of 26.4 kcal/mol, after which the complex rearranges itself to the trans isomer back again to yield silyl species 35. Chapter 3. Results and discussion [292] Figure 56. Energy profile for the formation of complex 26 from 24 considering a trans arrangement of the carbene ligands. Cationic Pt(II) σ-SiH complexes as reactive intermediates [293] Figure 57. Energy profile for the formation of complex 26 from 24 considering an initial cis arrangement of the carbene ligands. Chapter 3. Results and discussion [294] Pt(II) hydride complexes with primary silanes In this study, ItBuiPr derivative [Pt(H)(ItBuiPr)2][BAr4F], 30, and nBuSiH3 were used in calculations where the formation of silyl complex 31 was analysed by considering two mechanistic alternatives: the reaction with the NHC ligands in a trans arrangement, or a scenario where both carbenes are cis to each other.  a) Pt−Si coupling (NHCs in trans) Starting with complex [Pt(H)(ItBuiPr)2][BAr4F] (30) and nBuSiH3 as the energy reference, formation of the σ-SiH complex 30·BuSiH3 is once again slightly endergonic (2.7 kcal/mol) due to entropic factors (Figure 58). Then, this species evolves towards the oxidative addition of one of the Si−H bonds of the silane to give transition state TS8, which is 35.2 kcal/mol above the origin. As described above, an orbiting-like motion of the bridging hydride places the SiH proton close to the hydride ligand bound to the metal center. The outcome of this transition state is intermediate I8, which shows the formation of the Pt−Si bond as well as the dihydrogen σ-complex. This geometry is almost isoenergetic to the reagents, with 0.2 kcal/mol. Stabilization of the system (-6.4 kcal/mol) comes from the liberation of the H2 molecule, releasing silyl complex [Pt(SiH2nBu)(ItBuiPr)2][BAr4F] (31) in the process. As can be deduced, the high energy observed for transition state TS8 makes this pathway unlikely to happen.  b) Pt−Si coupling (NHCs in cis) In a similar manner to the previously described mechanisms, isomerization of the platinum precursor is necessary for the reaction to be conducted. This comes about through transition state TSisom2, which requires 16.8 kcal/mol (Figure 59). Carbene rearrangement gives isomer 30cis, 8.8 kcal/mol less stable than its trans counterpart. Approach of one molecule of butylsilane produces 30cis·BuSiH3, whose formation is again somewhat endergonic to almost the same extent as the trans isomer, given that this complex is 2.6 kcal/mol above the Cationic Pt(II) σ-SiH complexes as reactive intermediates [295] cyclometalated species. The cis arrangement of the carbene ligands places the bridging SiH hydrogen close enough to bounce to the position where the platinum hydride is located so as to form the σ-dihydrogen complex. This migration takes place through transition state TS9, requiring 20.0 kcal/mol, making this mechanism a feasible possibility by which this reaction takes place. Final formation of the Pt−Si bond yields the dihydrogen intermediate I9. Extrusion of H2 gives silyl derivative [Pt(SiH2nBu)(ItBuiPr)2][BAr4F] (31) 6.4 kcal/mol below the reagents. In the light of the energy data obtained for both pathways, it seems reasonable to think that this transformation takes place through the mechanism where both carbene ligands are located cis to each other. Chapter 3. Results and discussion [296] Figure 58. Energy profile for the formation of complex 31 from 30 considering a trans arrangement of the carbene ligands. Cationic Pt(II) σ-SiH complexes as reactive intermediates [297] Figure 59. Energy profile for the formation of complex 31 from 30 considering an initial cis arrangement of the carbene ligands. Chapter 3. Results and discussion [298] 2.6. Crystallization of Pt(II) hydride and/or deuteride complexes. Molecular rotors Numerous attempts of crystallizing the σ-SiH complexes with ItBu ligands were performed yet all of them proved unfruitful as no crystals were obtained, or precipitate or oils were formed instead. This is not surprising, given the broad signals displayed in the NMR spectra, which indicate that the interaction is not strong enough to hold both chemical entities together. Nonetheless, a crystal obtained in one of the few presumed successful experiments was measured by X-Ray diffraction techniques and turned out to be hydride complex 15 (Figure 60). The solid-state structure of this species could not be elucidated in the past since this molecule is unstable towards cyclometalation to give 13 when exposed to vacuum or mild heating, and attempts to crystallize it under an atmosphere of H2 failed.55 However, the unintentional hydrolysis of the silane during crystallization of the mixture of complex 13 and nBuSiH3 provided a reservoir of H2 leading to hydride derivative 15. Figure 60. Crystal structure of hydride complex 15. BAr4F anion and most of the hydrogen atoms have been omitted for clarity. Cationic Pt(II) σ-SiH complexes as reactive intermediates [299] In the same way as in complex 16, the crystal structure shows a trans arrangement of the carbene groups, and the hydride has no ligand in the opposite direction to it, although there is an agostic interaction stabilizing the Pt(II) center. The shorter distance between Pt and H5c (1.971(1) Å) and Pt and C5 (2.639(5) Å) with respect to silyl complex 16 can be explained by the weaker trans influence of hydride ligands in comparison to silyl groups.47 These structural parameters perfectly fit those previously reported in our group in collaboration with Prof. Agustí Lledós (UAB) by DFT calculations.36, 82 Moreover, these calculations revealed an unexpected fluxional behaviour of the hydride ligands, which are able to move around the metal center in an orbital fashion while the NHCs are static (Scheme 20), similar to that described in the previous section. These results explain the reasons for the unobserved agostic interaction of the platinum center with the tert-butyl groups by 1H NMR spectroscopy. Unfortunately, solution NMR spectroscopy is not a convenient technique to demonstrate this fluxional process. However, soon after publishing our results, the groups of Ulrich Fekl and Robert W. Schurko at the Universities of Toronto and Windsor, respectively, were able to determine the capability of deuteride ligands in trans-D(L)-Pt(PtBu3)2 (L = deuteride or chloride) complexes to behave as rotors by combining 2H and 195Pt solid-state NMR studies with theoretical calculations. 83 82 M. A. Ortuño, P. Vidossich, S. Conejero, A. Lledós, Angew. Chem. Int. Ed., 2014, 53, 14158- 14161. 83 E. Prack, C. A. O’Keefe, J. K. Moore, A. Lai, A. J. Lough, P. M. Macdonald, M. S. Conradi, R. W. Schurko, U. Fekl, J. Am. Chem. Soc., 2015, 137, 13464-13467. Chapter 3. Results and discussion [300] Scheme 20. Orbital-like motion of hydride ligands around the metal center (adapted from reference 82). The fact that it is possible to isolate complex 15 in the solid state raised the question of whether it is possible to determine experimentally such a fluxional process. For this reason, a collaboration has been stablished with these researchers, and several trans-D(L)-Pt(NHC)2 compounds containing ItBu and IPr 84 carbenes (Figure 61) have been prepared through modified procedures. Figure 61. Platinum deuteride complexes employed in the molecular rotors investigation. The reason for the selection of these compounds is the availability of their corresponding crystal structures,55, 85 so they can be employed in the molecular dynamics calculations to evaluate their rotor behaviour. Complex 39- D was previously synthesized in our group by H/D exchange using the 84 IPr stands for 1,3-bis-(2,6-diisopropylphenyl)-imidazol-2-ylidene. 85 The crystal structure of complex 39 contained hexafluoroantimonate(V) as a noncoordinating anion. Cationic Pt(II) σ-SiH complexes as reactive intermediates [301] corresponding hydride derivatives.36 With respect to complex 15-D, as mentioned before, it is unstable towards cyclometalation under vacuum or mild heating, but we have found that it can be prepared in the solid state (without solvent) by pressurizing the cyclometalated derivative 13 under D2. This compound is stable under a D2 atmosphere, but undergoes cyclometalation even in the solid state at r.t. after a few hours under air. However, it is stable enough for studying it by solid state NMR spectroscopy at low temperatures. Complex 40 was also prepared in our group, but only in the dihydride form.51 The synthetic procedure for the dideuteride complex 40-D, that entails the reaction of complex 39-H with NaH, has now been modified and the yield improved: when searching for commercially available anionic sources of deuterium that could transform 39-D into 40-D, alkali salts such as LiD or LiAlD4 were unsuccessful to achieve such transformation. 86 However, NaBD4 was reactive enough to transform 39-D into adduct 41 in 1 hour in excellent yield (Scheme 21). Subsequent removal of the BD3 unit was possible by using a large excess (85 eq) of Et3N at 70 °C for 19 hours, affording the corresponding amino-borane Et3N·BD3 and clean 40-D in good yield (Figure 62). Scheme 21. Synthetic route for platinum dideuteride 40−D. 86 NaD was not commercially available by the time the synthesis of 40-D was performed. Chapter 3. References [308] 62 Formation of silyl complex 21 is a consequence of the reactivity patterns displayed by the σ-silane complexes studied in this thesis. This behaviour will be described in detail in the next section of the chapter. 63 A BCP is observed between Pt and Si in the calculated structure of 19·Ph3SiH upon further closing the Pt−H−Si angle to around 80°, which suggests that the situation observed above is close to the threshold for a BCP to appear. 64 Some studies have shown that the presence of a RCP close to a BCP is indicative of the instability of the latter. See: P. Meixner, K. Batke, A. Fischer, D. Schmitz, G. Eickerling, M. Kalter, K. Ruhland, K. Eichele, J. E. Barquera-Lozada, N. P. M. Casati, F. Montisci, P. Macchi, W. Scherer, J.Phys. Chem. A, 2017, 121, 7219. 65 Calculations were carried out at the PBE/DZVP/GTH level of theory. 66 P. Vidossich, A. Lledós, Dalton Trans., 2014, 43, 11145-11151. 67 M. A. Ortuño, P. Vidossich, G. Ujaque, S. Conejero, A. Lledós, Dalton Trans., 2013, 42, 12165-12172. 68 This rearrangement can be observed in the video file included in the CD attached to this thesis. 69 P. Vidossich, G. Ujaque, A. Lledós. Chem. Commun., 2012, 48, 1979-1981. 70 M. Aizenberg, D. Milstein, J. Am. Chem. Soc., 1995, 117, 6456-6464. 71 M. E. van der Boom, J. Ott, D. Milstein, Organometallics, 1998, 17, 4263-4266. 72 M. Aizenberg, D. Milstein, Angew. Chem. Int. Ed., 1994, 33, 317-319. 73 As shown above in section 2.2, complex 21 could be crystallized and analyzed by X-Ray diffraction methods (Figure 27). 74 For NMR data regarding neutral silyl Pt(II) complexes, see: a) D. Chan, S. B. Duckett, S. L. Heath, I. G. Khazal, R. N. Perutz, S. Sabo-Etienne, P. L. Timmins, Organometallics, 2004, 23, 5744-5756; b) J. C. DeMott, W. Gu, B. J. McCulloch, D. E. Herbert, M. D. Goshert, J. R. Walensky, J. Zhou, O. V. Ozerov, Organometallics, 2015, 34, 3930-3933; c) C. Mitzenheim, T. Braun, R. Laubenstein, B. Braun, R. Herrmann, Dalton Trans., 2016, 45, 6394-6404. Cationic Pt(II) σ-SiH complexes as reactive intermediates [309] 75 A small amount (~5%) of complex 37 ([Pt(IMes*)2SiH2nBu][BAr4F], see below) is observed in the synthesis of species 28. See Experimental part for more details. 76 a) Z. Mo, Y. Liu, L. Deng, Angew. Chem. Int. Ed., 2013, 52, 10845-10849; b) J. Sun, C. Ou, C. Wang, M. Uchiyama, L. Deng, Organometallics, 2015, 34, 1546-1551; c) J. Sun, L. Luo, Y. Luo, L. Deng, Angew. Chem. Int. Ed., 2017, 56, 2720-2724. 77 Z. Ouyang, L. Deng, Organometallics, 2013, 32, 7268-7271. 78 The terminal hydrogen atoms resonate at 3 and 3.84 ppm. The three SiH protons correlate with a peak at -53.5 ppm in the 29Si spectrum (HMQC 1H−29Si). Additionally, NOESY experiments reveal exchange between bridging and terminal protons, suggesting a fluxional process. 79 Calculations were carried out at the M06/6-31g(d,p)/SDD//M06/6- 311++g(d,p)/SDD level of theory. SMD (Dichloromethane) continuum was used to model the solvent. 80 Mechanistic pathways involving silylene species were also taken into account, yet their energy barriers were too high to be considered as experimentally feasible. For the sake of simplicity, these preliminary results are not included in this thesis. 81 It is worth remembering that the formation of the σ-SiH complexes is slightly endergonic due to entropic factors (see section 2.1) but exothermic enough to observe their formation at low temperature. 82 M. A. Ortuño, P. Vidossich, S. Conejero, A. Lledós, Angew. Chem. Int. Ed., 2014, 53, 14158-14161. 83 E. Prack, C. A. O’Keefe, J. K. Moore, A. Lai, A. J. Lough, P. M. Macdonald, M. S. Conradi, R. W. Schurko, U. Fekl, J. Am. Chem. Soc., 2015, 137, 13464-13467. 84 IPr stands for 1,3-bis-(2,6-diisopropylphenyl)-imidazol-2-ylidene. 85 The crystal structure of complex 39 contained hexafluoroantimonate(V) as a non-coordinating anion. 86 NaD was not commercially available by the time the synthesis of 40-D was performed. Chapter 3. References [310] 87 The other Pt−D distance is 1.961(9) Å, which is oddly long. This might be due to the difficulties found when locating the electron density of the deuteride ligand during refinement of the structure. DFT calculations on 40 (Level of theory: M06/6-31g(d,p/SDD) gave a length of 1.68 Å, in agreement with the experimental findings (1.703(5) Å). 88 a) P. G. Owston, J. M. Partridge, J. M. Rowe, Acta Crystallogr., 1960, 13, 246-252; b) G. B. Robertson, P. A. Tucker, W. A. Wickramasinghe, Aust. J. Chem., 1986, 39, 1495-1507; c) A. Albinati, G. Bracher, D. Carmona, J. H. P. Jans, W. T. Klooster, T. F. Koetzle, A. Macchioni, J. S. Ricci, R. Thouvenot, L. M. Venanzi, Inorg. Chim. Acta, 1997, 265, 255-265. Chapter 4 Catalytic processes assisted by cationic Pt(II) σ-SiH complexes Chapter 4. Introduction [312] 1. Introduction 1.1. Reactivity with electrophilic sources of Silicon As described throughout this thesis, activation of hydrosilanes by Lewis acids can lead to an increase in the electrophilicity of the silicon atom. This process can be mediated by main-group Lewis acids or electrophilic transitionmetal complexes, in which silylene species may be involved. 1 Depending on the degree of activation, partial or total exposure of the silicon atom (and silylium character) 2 might be observed. Silylium cations R3Si+ species are extremely strong Lewis acids due to their empty porbital and their unfilled valence shell with only six electrons in it. In spite of having properties that would suggest a higher stabilization over their carbon analogues (i.e. carbenium ions) such as bigger size, lower electronegativity (2.55 for C vs 1.90 for Si) and greater polarizability, they are indeed much more unstable. This is because of kinetic reasons, which make silylium cations exceptionally reactive and susceptible to nucleophiles in comparison to their carbon counterparts. 3 For this reason, the presence of even weak Lewis bases (like coordinating solvents) must be avoided in the synthesis of this type of species. One of the most common methods for preparing them takes advantage of the polarization of the Si−H bond to abstract the hydrogen atom (with hydride character) by using strong electrophiles containing weak-coordinating anions 1 M. C. Lipke, A. L. Liberman-Martin, T. D. Tilley, Angew. Chem. Int. Ed., 2017, 56, 2260-2294. 2 D. H. Binh, M. Milovanović, J. Puertes-Mico, M. Hamdaoui, S. D. Zarić, J-P. Djukic, Chem. Eur. J., 2017, 23, 17058-17069. 3 V. Y. Lee, A. Sekiguchi in Organosilicon Compounds, 2017, Academic Press. Chapter 5: Silicon-centered cations, 197-230. Catalytic processes assisted by Pt(II) σ-SiH complexes [313] such as the trityl salt [Ph3C][B(C6F5)4]. This procedure is frequently referred to as Bartlett-Condon-Schneider hydride-transfer reaction (Scheme 1). 4 Scheme 1. Hydride-transfer reaction to give silylium cations (adapted from reference 3). In fact, this methodology was adopted by the group of Oestreich several years ago, where they synthesized the ferrocene-stabilized silylium ion displayed in Scheme 2 (top), which was able to effectively carry out Diels-Alder reactions at low temperature 5 as well as hydrosilation of carbonyl groups (Scheme 2, bottom). 6 4 P. D. Bartlett, F. E. Condon, A. Schneider, J. Am. Chem. Soc., 1944, 66, 1531-1539. 5 H. F. T. Klare, K. Bergander, M. Oestreich, Angew. Chem. Int. Ed., 2009, 48, 9077-9079. 6 K. Müther, M. Oestreich, Chem. Commun., 2011, 47, 334-336. Chapter 4. Introduction [314] Scheme 2. Top: hydride-transfer reaction to give the ferrocenium stabilized silylium ion described by Oestreich et al. Bottom: proposed catalytic cycle for the silylium-assisted hydrosilation of carbonyl groups (adapted from reference 6). The short reaction times and low temperatures required for performing such reactions reflects the high reactivity of these derivatives. Moreover, the powerful Lewis acidity of silylium cations allows them to activate inert entities such as carbon dioxide. Examples of CO2 activation by silylium species include those in which these cations are forming part of silylium/phosphane, 7 7 M. Reissmann, A. Schafer, S. Jung, T. Müller, Organometallics, 2013, 32, 6736-6744. Catalytic processes assisted by Pt(II) σ-SiH complexes [315] silylium/NHCs 8 or silylium/guanidine 9 Frustrated Lewis Pairs, to name a few. The formation of a FLP is not unavoidable, given that “naked” silylium ions can transform carbon dioxide by themselves, as reported by Müller et al., since they can convert CO2 into different products by changing the conditions of the stoichiometric reactions they carry out (Scheme 3, top). 10 By using the same strategy as Oestreich (i.e. hydrogen abstraction assisted by the trityl cation), the resulting silylium ion can react with hydrosilanes in chlorobenzene to give disilylated formate A, which can undergo hydrolysis to formic acid. In the specific case of Et3SiH, more equivalents of hydrosilane can react with A to afford disilyl methyl oxonium ion B, which can in turn be hydrolyzed to methanol. On the other hand, reactions carried out in benzene lead to the formation of cation C. In this situation, two alternative pathways were described by the authors: deprotonation with 2,4,6-collidine to give silyl ester D or hydrolysis to benzoic acid E. The group demonstrated the role of the solvents and their participation in the different mechanisms originating such a variety of species from the same reagents. Additionally, a silylium-assisted version of the Gatterman-Koch reaction to produce aromatic silyl carboxonium compounds from silyl cations, benzene and CO was published very recently, where a similarity between protons and silylium ions (“fat” protons, according to the article) was made, establishing a new chemical point of view for this reactive species (Scheme 3, bottom). 11 8 M. F. S. Valverde, E. Theuergarten, T. Bannenberg, M. Freytag, P. G. Jones, M. Tamm, Dalton Trans., 2015, 44, 9400-9408. 9 N. von Wolff, G. Lefèvre, J.-C. Berthet, P. Thuéry, T. Cantat, ACS Catal., 2016, 6, 4526-4535. 10 A. Schäfer, W. Saak, D. Haase, T. Müller, Angew. Chem. Int. Ed., 2012, 51, 2981-2984. 11 L. Omann, Z-W. Qu, E. Irran, H. F. T. Klare, S. Grimme, M. Oestreich, Angew. Chem. Int. Ed., 2018, 57, 8301-8305. Chapter 4. Introduction [316] Scheme 3. Silylium-assisted carbon dioxide (top) and carbon monoxide reduction (bottom). Adapted from references 10 and 11. In some of the previous cases, CO2 activation takes place through the oxygen atom which attacks the electrophilic silylium species, opposed to the carbon dioxide insertion mechanisms described in chapter 1, where a hydride ligand usually attacked the electrophilic carbon atom. This activation mechanism has been described for other processes where extremely Lewis acidic species are involved, like the protonation of CO2 by the carborane superacid H(CHB11F11). 12 In this case, the resulting product is the proton di-solvate H(CO2)2+ (Scheme 4) and not the monoprotonated salt [HCO2][CHB11F11]. This is not surprising, given 12 S. Cummings, H. P. Hratchian, C. A. Reed, Angew. Chem. Int. Ed., 2016, 55, 1382-1386. Catalytic processes assisted by Pt(II) σ-SiH complexes [317] the prevalence of proton disolvates 13 like in the case of Brookhart’s acid, [H(Et2O)2][BAr4F]. 14 Scheme 4. Protonation of CO2 to give the proton disolvate [H(CO2)2][CHB11F11] (adapted from reference 12). σ-SiH complexes It has been shown throughout this thesis that sigma-silane complexes can also carry out catalytic processes in which silyl group transfers are involved. Even though they do not give rise to fully “naked” silylium ions, they render the silicon atom with enough electrophilic character to perform this type of reactions. The degree of electrophilicity is related to the coordination mode of the silane (i.e. η1 or η2 fashion), as previously seen in chapter 3. σ-SiH complexes have been proposed as intermediates in silane alcoholysis reactions (Scheme 5), where a Si−O bond is forged at the same time dihydrogen is released. 15 Scheme 5. Silane alcoholysis reaction. Crabtree et al. observed that the kinetic and mechanistic data obtained with Ir(III) complexes of the type [Ir(H)2(PPh3)2S2][SbF6] (where S = THF, H2O, MeOH or Me2CO) suggested an alternative mechanism to the oxidative addition pathway, in which the silane was bound to the metal through the Si−H bond in a 13 C. A. Reed, Acc. Chem. Res., 2013, 46, 2567-2575. 14 M. Brookhart, B. Grant, A. F. Volpe Jr., Organometallics, 1992, 11, 3920-3922. 15 E. Lukevics, M. Dzintara, J. Organomet. Chem., 1985, 295, 265-315. Chapter 4. Introduction [324] Figure 1. Some selected recent examples of transition metal complexes able to selectively transform CO2 into silyl formates by using hydrosilanes. Also in 2013, the group of Mizuno achieved successful chemical manipulation of the silyl formate products after its catalytic synthesis. After testing a set of commercially available metal salts, bases and solvents, the authors concluded that the best catalytic results were those obtained by using Rh2(OAc)4, inorganic O-donor bases like K2CO3 and acetonitrile (Figure 1D). 27 Then, the catalysis products were subjected to hydrolysis (to give formic acid) or they were exposed to Grignard reagents (to give benzhydrol derivatives) or 27 S. Itagaki, K. Yamaguchi, N. Mizuno, J. Mol. Catal. A: Chem., 2013, 366, 347-352. Catalytic processes assisted by Pt(II) σ-SiH complexes [325] amines (to give formylamines). Other groups also make use of formoxysilanes as formylating agents 28 or as surrogates of silanes. 29 More recent examples of selective carbon dioxide hydrosilation include the bimetallic Pd-based system developed in the group of Iwasawa (Figure 1E). The Al−metalloligand complex achieved the highest TOF value (19300) reported so far, working at room temperature in DMF under 1 atmosphere of CO2 (balloon) and in the presence of cesium pivalate as an additive. 30 On the other hand, Oro and co-workers have studied the carbon dioxide hydrosilation reaction by using Ir(III) complexes stabilized by NSiN ligands (Figure 1F). 31 As a result, selective reduction to the formic acid stage was observed, and experimental and theoretical (DFT) studies led to the conclusion that different mechanisms (e.g. outer or inner-sphere pathways) were taking place depending on the iridium species employed. In addition, the trifluoroacetate ligand proved to be non-innocent from a certain temperature (55 °C), since hydrosilation to the silyl ether CF3CH2OSiR3 was observed. Moreover, the authors noticed a great dependence on the hydrosilane in terms of reaction rates by means of kinetic analysis. It is worth mentioning that all the cases described above employed tertiary silanes, in which only one Si−H bond is chemically available. If secondary or primary silanes were to be used, the selectivity challenge found in carbon dioxide reduction (i.e. selective transformation towards formic acid, formaldehyde, methanol or methane) would find an additional issue, since mono-, di- or trisubstituted species could be formed (Scheme 9). Indeed, 28 L. González-Sebastián, M. Flores-Alamo, J. J. García, Organometallics, 2013, 32, 7186-7194. 29 C. Chauvier, P. Thuéry, T. Cantat, Angew. Chem. Int. Ed., 2016, 55, 14096-14100. 30 J. Takaya, N. Iwasawa, J. Am. Chem. Soc., 2017, 139, 6074-6077. 31 A. Julián, J. Guzmán, E. A. Jaseer, F. J. Fernández-Álvarez, R. Royo, V. Polo, P. García-Orduña, F. J. Lahoz, L. A. Oro, Chem. Eur. J., 2017, 23, 11898-11907. Chapter 4. Introduction [326] insertion of CO2 into multiple Si−H bonds has been experimentally observed in some cases. 32 Scheme 9. Mixture of products that can be obtained during catalytic CO2 hydrosilation when using primary silanes. 1.3. Synthesis of silazanes via dehydrogenative coupling reactions Silazanes and polysilazanes Contrary to their organic counterparts, inorganic polymers have remained elusive for a long time until recently, since the synthesis of inorganic molecular chains is much more challenging due to several reasons. For instance, addition mechanisms (e.g. polymerization of alkenes, alkynes, nitriles, etc) are not easy to achieve in inorganic entities, given that isolating the hypothetical monomers with multiple bonds between inorganic elements frequently involves the use of bulky substituents so as to kinetically overcome the “double-bond rule” 33 and prevent the uncontrolled oligomerization. 34 In this regard, some of the first stable dinuclear species like disilenes or iminoboranes date back to the 1980s, which evidences that this chemistry is still very young (Figure 2A). 35 32 a) A. Jansen, H. Görls, S. Pitter, Organometallics, 2000, 19, 135-138; b) M. L. Scheuermann, S. P. Semproni, I. Pappas, P. J. Chirik, Inorg. Chem., 2014, 53, 9463-9465. 33 L. E. Gusel’nikov, N. S. Nametkin, Chem. Rev., 1979, 79, 529-577. 34 I. Manners, Angew. Chem. Int. Ed., 1996, 35, 1602-1621. 35 a) Information about disilenes: R. West, Angew. Chem. Int. Ed., 1987, 26, 1201-1211; b) Information about iminoboranes: P. Paetzold, Adv. Inorg. Chem., 1987, 31, 123-170. Catalytic processes assisted by Pt(II) σ-SiH complexes [327] Figure 2. A) Examples of disilane and iminoborane as kinetically stable unsaturated inorganic molecules (adapted from references 35a and 35b). B) General formula of the first inorganic polymers synthesized (adapted from reference 34). The drawbacks in the synthesis of inorganic polymers narrowed the scope of this type of materials, limiting their variety to polysiloxanes, polyphosphazenes, polysilanes and polyiminoboranes (Figure 2B). Thus, the formation of different polymers such as those with Si−N bonds as the main repeating unit (polysilazanes) has been underdeveloped. In addition, the chemical instability of the primitive structures of silazanes and polysilazanes towards water, oxygen or protic species made their investigation even more difficult. Nonetheless, their relevance as precursors in the synthesis of silicon nitride (Si3N4) or silicon carbonitride (SixNyCz) boosted the research in this field, given that these materials are extensively employed in the production of ceramics 36 due to their exceptional stability properties under extreme temperatures and oxidation environments, making them ideal candidates for 36 A. Soum in Silicon-containing Polymers. The Science and Technology of their Synthesis and Applications, 2000, Springer Netherlands. Chapter 11: Polysilazanes, 323-349. Chapter 4. Introduction [328] abrasives or highly resistant components of many devices. 37 As a matter of fact, recent developments on this area have led to the creation of a new family of polymers which serve as extremely resistant coatings for a number of surfaces at the same time they provide protection against stains, inks and paints due to their hydrophobic nature. 38 Moreover, silazanes are useful compounds that can serve as ligands, protecting groups or bases.18c Regarding their structure, silazanes or amino-silanes are entities in which a silane molecule is bound to one, two or three amine groups, forming mono-, di- or trisilazanes respectively (Scheme 10). The binding and cross-linking of these species give rise to the corresponding polysilazanes. These polymers are often referred to as organopolysilazanes when hydrocarbon chains are present as substituents across the Si−N backbone.36 Scheme 10. General formulae of silazanes and polysilazanes. 37 M. Birot, J-P. Pillot, J. Dunoguès, Chem. Rev., 1995, 95, 1443-1477. 38 This family of polysilazanes has been commercialized by Merck KGaA under the name Durazane® . For more information, see: https://www.merckgroup.com/en/brands/pm/durazane.html. Catalytic processes assisted by Pt(II) σ-SiH complexes [329] The most common method for the synthesis of silazanes and oligosilazanes is the aminolysis of chlorosilanes. In this process, hydrochloric acid is formed as a by-product and it is usually neutralized by an excess of the amine involved in the reaction, forming an equilibrium between the involved species (Scheme 11A).18d, 39 In addition, the presence of a strong acid in the reaction medium limits the functional group tolerance. These disadvantages, along with the toxic, irritant and corrosive nature of HCl, 40 demand for alternative methods of synthesizing Si−N derivatives. In the light of this situation, dehydrogenative coupling of hydrosilanes and amines arises as a clean alternative (Scheme 11B) given that the only by-product is dihydrogen, which can be leveraged for several other purposes. 41 Scheme 11. Synthetic methods for the preparation of silazanes. A) Aminolysis of chlorosilanes. B) Dehydrogenative coupling of hydrosilanes and amines. Metal-assisted dehydrogenative Si−N formation The development of homogeneous catalytic methods to forge Si−N bonds from hydrosilanes and amines has gained interest during the last couple of decades. 42 This transformation is challenging because of the wide variety of 39 R. Fessenden, J. S. Fessenden, Chem. Rev., 1961, 61, 361-388. 40 S. Austin, A. Glowacki. Hydrochloric acid. Ullmann’s Encyclopedia of Industrial Chemistry, 2000, doi: 10.1002/14356007.a13_283. 41 a) R. Noyori, T. Ohkuma, Angew. Chem. Int. Ed., 2001, 40, 40-73; b) S. Werkmeister, K. Junge, M. Beller, Org. Process Res. Dev., 2014, 18, 289-302; c) G. A. Filonenko, R. V. Putten, E. J. M. Hensen, E. A. Pidko, Chem. Soc. Rev., 2018, 47, 1459-1483. 42 Metal-free catalytic systems have also been described in the literature (see reference 18d). Chapter 4. Introduction [330] species that might arise from the reaction, since mixtures of products with diverse Si / N ratios, oligo- or polysilazanes may be found. Thus, aside from high activities, selective systems capable of leading to just one type of amino-silane are desirable. Alkaline earth based systems proved to be efficient for this type of transformation, as shown by the work of Harder et al. in 2007, where the calcium azacyclopropane depicted in Scheme 12 (left) is able to give monosilazanes out of the bulky silane Ph3SiH and aliphatic and aromatic amines at room temperature and catalyst loadings of around 5-10 mol %. 43 A few years later, the group of Sadow employed magnesium complexes stabilized by tris(oxazolinyl)borato ligands (ToM) to selectively transform primary and secondary hydrosilanes into amino-silanes (Scheme 12, right). 44 The degree of substitution on the Si atom (i.e. the amount of amine groups bound to silicon) depends on the type and stoichiometry of the reagents. Interestingly, the Mg complex is effective even in the transformation of NH3, which affords monosilazanes upon reaction with tertiary silanes. This phenomenon is remarkable, as incorporation of silyl groups in ammonia makes the rest of the N−H bonds more acidic, turning the monosilation reaction difficult to achieve. 43 F. Buch, S. Harder, Organometallics, 2007, 26, 5132-5135. 44 J. F. Dunne, S. R. Neal, J. Engelkemier, A. Ellern, A. D. Sadow, J. Am. Chem. Soc., 2011, 133, 16782-16785. Catalytic processes assisted by Pt(II) σ-SiH complexes [331] Scheme 12. Examples of alkaline earth-based catalysts and their proposed catalytic cycles for the formation of monosilazanes. (adapted from references 43 and 44). Although the mechanisms by which these processes take place are not completely clear, the authors propose analogous catalytic cycles that involve nucleophilic attack of the amide group to the silane followed by hydrogen transfer from Si to the metal, in a process that resembles β-elimination. Elimination of the silazane and protonation of the resulting hydride finish the cycle by releasing H2 and regenerating the metal−amide bond.43,44 This transformation is not limited to group 2 elements, given that transition metal-based catalysts can also be found in the literature, like the rhodium system described by Eisenberg et al., which consisted of binuclear complexes of the type Rh2(H)2(CO)2(dppm)2 or Rh2(μ-SiRH)2(CO)2(dppm)2 45 that are capable of producing small oligosilazanes (up to Si−N−Si−N−Si−N products). The nature of the resulting chain depends on the size of the substituents: bulky groups like iBu give linear products, whereas small alkyl fragments like CH3 45 R = Ph, Et or C6H13. dppm stands for bis(diphenylphosphino)methane. Chapter 4. Introduction [332] produce branched oligomers. 46 Binuclear rhodium derivatives have also been used in the silylation of indoles, pyrroles and carbazoles, by means of Rh2(OAc)4 accompanied by the base TBA2WO4 (TBA = tetrabutylammonium), as described by the group of Mizuno. 47 These heterocyclic substrates were also used by Oestreich and coworkers for the same purpose, extending the scope to indolines and anilines, which undergo Si−N coupling by means of catalytic loadings of 1 mol % of the same Ru complex described in chapter 1 for CO2 reduction to the bis(silyl)acetal stage (Figure 3). 48 Figure 3. Ruthenium system described by the group of Oestreich for Si−N dehydrogenative coupling and CO2 hydrosilation (adapted from reference 48). Curiously, many of the precatalysts reported in the literature contain the bis(trimethylsilyl)amide or hexamethyldisilazide (HMDS) ligand, probably due to the solubility and steric stability that they provide (Figure 4). This makes possible the stabilization of lanthanide-based compounds such as the ytterbium species reported by Cui (Figure 4, top left), which also possess an NHC ligand that modulates the reactivity of the complex (preliminary studies in the absence of NHCs led to the conclusion that an ancillary ligand was necessary). For instance, coupling PhSiH3 to Et2NH by using 5 mol % of the IiPr derivative gives a 1:1 mixture of the mono- and disilazane products, yet it yields quantitative 46 W-D. Wang, R. Eisenberg, Organometallics, 1991, 10, 2222-2227. 47 S. Itagaki, K. Kamata, K. Yamaguchi, N. Mizuno, Chem. Commun., 2012, 48, 9269-9271. 48 C. D. F. Königs, M. F. Müller, N. Aiguabella, H. F. T. Klare, M. Oestreich, Chem. Commun., 2013, 49, 1506-1508. Catalytic processes assisted by Pt(II) σ-SiH complexes [333] formation of the disilazane in the presence of the IMes complex. 49 Similar enhancements due to auxiliary ligands were observed in the group of Crimmin for the yttrium precatalysts depicted in Figure 4 (top right). Whereas both complexes similarly catalyze the dehydrocoupling reaction of primary or secondary phenylsilanes with primary or sterically accessible secondary amines (e.g. dibenzylamine, piperidine), there is a dramatic difference in the reaction with bulky amines such as diisopropylamine: the complex bearing a cyclometalated phosphonium methylide ligand is able to achieve 95% yield of the desired monosilazane after 44 h, but [Y{N(SiMe3)2}3] gives no product after 262 hours of reaction under the same conditions (10 mol % of catalyst at room temperature). 50 On the contrary, Waterman et al. showed that no auxiliary ligands were needed when using the lanthanum precatalyst shown in Figure 4 (bottom right), that turned out to be the most active lanthanide-based one for this type of reaction, given that even trisilazanes can be obtained with catalyst loadings as low as 0.8 mol % by slightly increasing the amine/silane ratio (up to 4 equivalents) and the temperature up to 90 °C. 51 49 W. Xie, H. Hu, C. Cui, Angew. Chem. Int. Ed., 2012, 51, 11141-11144. 50 A. E. Nako, W. Chen, A. J. P. White, M. R. Crimmin, Organometallics, 2015, 34, 4369-4375. 51 M. P. Cibuzar, R. Waterman, Organometallics, 2018, 37, 4395-4401. Chapter 4. Results and discussion [340] Figure 5. Typical 1H NMR spectrum (400 MHz, CD2Cl2) of a catalytic experiment with 13 (0.5 mol %), nBuSiH3 and CO2 (5 bar) after consumption of the silane. Catalytic processes assisted by Pt(II) σ-SiH complexes [341] Figure 6. Comparison of the different plots when using platinum complexes 13, 15 and 16 (0.5 mol %) and nBuSiH3 in the presence of CO2 (5 bar) for the formation of HCOOSiH2nBu. The use of secondary silanes such as Et2SiH2 also gives monoformates as products, but in this occasion longer reaction times are needed in all cases (Figures 7 and 8), despite the fact that higher catalytic loadings of 13 are necessary (3 mol %) due to partial hydrolysis (and concomitant formation of hydride 15, which is less efficient) due to the presence of adventitious water in the reaction media. Tertiary silanes such as Et3SiH seemed to be too bulky to interact with the platinum center, since no evolution was observed after prolonged reaction times and mild heating (45 °C). 58 58 Aromatic silanes like PhSiH3 were also studied, although several unidentified species were observed rather than silyl formate derivatives. 0 20 40 60 80 100 120 0 200 400 600 800 Conversion (%) t / min Catalysis at 0.5 mol % [Pt] with n-BuSiH3 13 (Cyclometalated) 15 (Hydride) 16 (Silyl) Chapter 4. Results and discussion [342] Figure 7. Comparison of the different plots when using 13 and different hydrosilanes in the presence of CO2 (5 bar) for the formation of the corresponding silyl formate. Figure 8. Monitoring of the amount of HCOOSiHEt2 produced when using 16 (0.5 mol %) and Et2SiH2 in the presence of 5 bar of CO2. 0 20 40 60 80 100 120 0 100 200 300 400 500 Conversion (%) t / min CO2hydrosilation with 13 nBuSiH3 (0.5 mol %) Et2SiH2 (3 mol %) 0 20 40 60 80 100 120 0 2000 4000 6000 8000 10000 Conversion (%) t / min Catalysis with 16 (0.5 mol %) and Et2SiH2 Catalytic processes assisted by Pt(II) σ-SiH complexes [343] Nonetheless, the mild conditions (room temperature, 5 bar of CO2) at which these reactions are carried out and the resulting catalytic activities obtained (Table 1) make these Pt(II) complexes competitive catalysts for carbon dioxide hydrosilation to the formic acid stage. Moreover, the excellent chemoselectivity observed towards monoformates in spite of using primary or secondary silanes is remarkable and uncommon, given that the usual strategy is to employ tertiary silanes so as to avoid mixtures of products (Scheme 9),32 as stated in the introduction of this chapter. Although steric hindrance may account for the kinetic differences experimentally observed (primary silanes like butylsilane rapidly react to yield the monoformate, yet secondary silanes undergo between 6 and 15 times longer reactions to get the same conversion values), several reasons might be behind the observation of silyl monoformate derivatives (Figure 9): a) formation of a second formate unit on the product might require longer reaction times; b) the electron-withdrawing nature of the formate group on the silane might favour a η2-coordination mode (less electrophilic than η1 SiH complexes) when interacting with the platinum atom, similar to the results obtained with Cl3SiH in chapter 3; c) the silyl monoformate may interact with the metal through one of the oxygen atoms, competing with the hydrosilane. As a consequence, the corresponding silyl monoformate builds up in the reaction mixture, becoming the only product of the catalytic process. Chapter 4. Results and discussion [344] [Pt] Silane Loading (mol %) time (h) T (°C) TON TOF (h-1) 13 nBuSiH3 0.5 0.28 25 200 714 13 Et2SiH2 3 4.2 25 33.33 8 15 nBuSiH3 0.5 10.8 25 200 18.5 15 Et2SiH2 0.5 60.3 25 177a 3 16 nBuSiH3 0.5 10.8 25 200 18.5 16 Et2SiH2 0.5 158.5 25 196b 3 Table 1. Reaction conditions and results for the Pt-catalyzed CO2 hydrosilation to silyl formate derivatives. Every catalytic experiment has been repeated at least twice. TON = (μmoles of silane/μmoles of catalyst)*formate ratio. a Value determined at a conversion of 88.4%. b Value determined at a conversion of 98%. Figure 9. Plausible explanations for the observed chemoselectivity during carbon dioxide hydrosilation by using complexes 13, 15 and 16. On the other hand, the superior activity of 13 in comparison to 15 and 16 may be explained in terms of stereoelectronic properties, as described in chapter 3: the bigger cavity of 13 with respect to 15 and 16, its preorganization towards interaction with hydrosilanes and the weaker trans influence of the methylene moiety against hydride or silyl ligands 59 may contribute to the stronger σ-silane species observed for complex 13 by experimental and theoretical methods, thus leading to a higher activation of the silane and consequently a faster catalytic transformation when CO2 is present. This is in 59 J. Zhu, Z. Lin, T. B. Marder, Inorg. Chem., 2005, 44, 9384-9390. Catalytic processes assisted by Pt(II) σ-SiH complexes [345] agreement with the observed experimental trend when the catalysts were subjected to an atmosphere of carbon dioxide and diethylsilane, since the required reaction times follow the sequence 13 (4 hours) > 15 (60 hours) > 16 (158 hours). 2.2. Dehydrogenative coupling of silanes and amines The mechanistic evidence shown by Crabtree et al. for the alcoholysis of silanes16 (assisted by σ-SiH complexes, see section 1.1 and Scheme 5) and by our group for dehydrocoupling reactions leading to amino- 60 and diaminoboranes 61 point to an initial step in which the Lewis base (i.e. silane or amine-borane) approaches the metal atom to undergo electrophilic activation so as to start the catalytic process. Thus, we rationalized that amine groups could in principle react with the Pt(II) σ-silane complexes observed in the laboratory to give rise to silazane derivatives. Dehydrocoupling of primary silanes Initial studies were performed by using PhSiH3 and tBuNH2 (1:1 ratio) in the presence of 0.5 mol % of complex 13 in dichloromethane (Scheme 16). This system gave rise to an extremely intense bubbling and the formation of the monosilazane PhSiH2(NHtBu) after a few seconds (Figure 10). At the end of the process, hydride 15 was observed as the only detected species, produced from the hydrogenation of 13 due to the H2 produced during the dehydrocoupling event. Whereas the selectivity is identical to that reported by the group of Sadow for the same substrates, their Mg system needs 24 h and a catalyst loading of 5 mol %.44 60 M. Roselló-Merino, J. López-Serrano, S. Conejero, J. Am. Chem. Soc., 2013, 135, 10910- 10913. 61 M. Roselló-Merino, R. J. Rama, J. Díez, S. Conejero, Chem. Commun., 2016, 52, 8389-8392. Chapter 4. Results and discussion [346] Scheme 16. Si−N dehydrogenative coupling of primary silanes with amines catalyzed by 13 to give monosilazanes. Figure 10. 1H NMR spectrum (400 MHz, CD2Cl2) of the crude reaction between PhSiH3 and tBuNH2 catalyzed by 13 (0.5 mol %). In the light of these encouraging results, the Pt loading was decreased to 0.005 mol % (50 ppm, Table 2, entry 1) and 0.003 mol % (30 ppm, Table 2, entry 2). In these reactions, 13 consumed the reactants in 12 and 35 min, respectively. The latter experiment afforded TON and TOF values of 3.3x104 and 5.7x104 h-1 respectively, which are much higher than those reported for the uranium system Catalytic processes assisted by Pt(II) σ-SiH complexes [347] [(Et2N)3U][BPh4] for the same product (TON = 167, TOF = 24 h-1). 62 The employment of Et2NH produced even faster reactions, since 50 ppm of 13 completely converted the silane into PhSiH2(NEt2) in 3.6 minutes (TON = 2x104, TOF = 3.3x105 h-1, entry 3 of Table 2), giving the highest activity values reported to date so far (as a comparison, previous works reported TON values of 20 and TOF values of 0.83-1.6 h-1).44, 63 The extremely fast catalysis observed allowed us to decrease the loading even more, down to 10 ppm (Table 2, entry 4), achieving 98% conversion in 3 hours. Several other amines also react with PhSiH3 in the presence of 13, such as the selectivity challenging substrate pyrrolidine, giving the corresponding monosilazane PhSiH2[N(CH2)4] in 35 min under 50 ppm of catalyst (Table 2, entry 6). However, bigger amines such as iPr2NH or less basic ones like MesNH2 need higher loadings in order to give similar conversions and reaction times (Table 2, entries 7-10). Nonetheless, the activities observed are still much higher than those reported for other methods (e.g. TON and TOF values of 20 and 20 h- 1 for a 5 mol % loading of the NHC-stabilized Yb system described in Figure 4).49 Alkyl silanes such as nBuSiH3 can also successfully undergo dehydrocoupling with amines with similar reaction times and catalyst loadings (50 ppm, entries 12-15 of Table 2). Nevertheless, the employment of iPr2NH gave rise to a rate increase of an order of magnitude compared to the reaction with PhSiH3 (entries 8 vs. 15). A plausible explanation for this phenomenon is explained below. 62 J. X. Wang, A. K. Dash, J. C. Berthet, M. Ephritikhine, M. S. Eisen, J. Organomet. Chem., 2000, 610, 49-57. 63 S. Anga, Y. Sarazin, J-F. Carpentier, T. K. Panda, ChemCatChem, 2016, 8, 1373-1378. Chapter 4. Results and discussion [348] Entry Silane Amine (equiv.) mol % (ppm) cat.a Product % yieldb (isol.) TON TOF (h-1) 1 PhSiH3 tBuNH2 (1) 0.005 (50) PhSiH2(tBuNH) 99(90) 2x104 9.6x104 2 PhSiH3 tBuNH2 (1) 0.003 (30) PhSiH2(tBuNH) 99 3.3x104 5.7x104 3 PhSiH3 Et2NH (1) 0.005 (50) PhSiH2(NEt2) 99(88) 2x104 3.3x105 4 PhSiH3 Et2NH (1) 0.001 (10) PhSiH2(NEt2) 98 9.8x104 2.9x104 5 PhSiH3 (CH2)4NH (1) 0.01 (100) PhSiH2[N(CH2)4] 99 1x104 6.4x104 6 PhSiH3 (CH2)4NH (1) 0.005 (50) PhSiH2[N(CH2)4] 99 2x104 3.4x104 7 PhSiH3 iPr2NH (1) 0.3 (3000) PhSiH2(NiPr2) 99(93) 333 1600 8 PhSiH3 iPr2NH (1) 0.1 (1000) PhSiH2(NiPr2) 99 1000 667 9 PhSiH3 iPr2NH (4) 0.1 (1000) PhSiH2(NiPr2) 99 1000 1.7x104 10 PhSiH3 MesNH2 (2) 0.1 (1000) PhSiH2(NHMes) 95 950 840 11 PhSiH3 MesNH2 (4) 0.1 (1000) PhSiH2(NHMes) 99 1000 2300 12 nBuSiH3 tBuNH2 (1) 0.005 (50) nBuSiH2(tBuNH) 99(90) 2x104 1.3x105 13 nBuSiH3 Et2NH (1) 0.005 (50) nBuSiH2(NEt2) 99(88) 2x104 1.2x105 14c nBuSiH3 (CH2)4NH (1) 0.005 (50) nBuSiH2[N(CH2)4] 99(91) 2x104 3x104 15 nBuSiH3 iPr2NH (1) 0.1 (1000) nBuSiH2(NiPr2) 96(86) 1000 3.5x104 Table 2. Summary of the catalytic dehydrocoupling Si−N reactions performed by 13 with primary silanes to give monosilazanes. a Catalyst loadings were determined with respect to silane. b Determined by 1H NMR. c Around 5% of disilazane nBuSiH[N(CH2)4]2 has been detected by NMR. Although the majority of these products are cleanly formed, their volatility makes possible their purification and isolation by vacuum distillation techniques or extraction with pentane, removing the catalyst in this way (Figure 11): Catalytic processes assisted by Pt(II) σ-SiH complexes [349] Figure 11. 1H NMR spectrum (400 MHz, CD2Cl2) of nBuSiH2(NEt2). Inset: Monodimensional and bidimensional 29Si NMR (400 MHz, CD2Cl2) experiments. Dehydrocoupling of secondary silanes Regarding secondary silanes, arylsilanes like Ph2SiH2 need higher catalyst loadings (0.1 mol %) in order to react with relatively small amines (Scheme 17) like tBuNH2, Et2NH or (CH2)4NH to yield the corresponding Ph2SiH(NR2) at short reaction times (entries 1-3 of Table 3). Bulkier amines such as iPr2NH show no reaction at room temperature, and heating the mixture to 60 °C for 4 days yields a mixture of multiple species. Alkylsilanes like Et2SiH2 react faster than their aromatic counterparts under the same conditions, giving activities up to two orders of magnitude higher (Table 3, entries 4-6 vs. 1-3). Chapter 4. Results and discussion [356] Figure 15. Plausible explanation for the observed experimental differences depending on the amine used in the dehydrocoupling reaction. The obtainment of reliable kinetic plots made possible the kinetic analysis so as to determine partial orders in the species involved in the reaction (see Experimental Part for more details). The reaction between PhSiH3 and Et2NH was chosen for this purpose. Unfortunately, partial order in amine could not be determined because of the reactivity between 13 and silanes, as described in chapter 3. At any rate, a negative partial order is expected, given the deceleration of the dehydrocoupling process with increasing amounts of Et2NH. Notwithstanding, the reaction proved to be zero-order in silane (similar to the calcium and magnesium systems described by Hill et al.)52 and second order (2.2) in catalyst (similar to the strontium systems described by the same group). This latter result will be discussed later on this chapter. Finally, it is interesting to mention that secondary silanes cause a different behaviour than primary ones. For instance, reactions carried out with Ph2SiH2 give exponential profiles in all cases, as indicated in Figure 16. This behavior is observed even when an excess of amine is present (Figure 17). Catalytic processes assisted by Pt(II) σ-SiH complexes [357] Figure 16. Kinetic profile (H2 evolution) of the reaction of Ph2SiH2 with 1 equivalent of: (a) tBuNH2, (b) Et2NH and (c) (CH2)4NH (catalyst loadings of 0.1, 0.2 and 0.3 mol % respectively). Figure 17. Effect of the concentration of Et2NH on the rate of H2 evolution in the reaction of Ph2SiH2 and Et2NH with 0.3 % of catalyst 13. Chapter 4. Results and discussion [358] In order to find an explanation to this disparity, a collaboration was established with Prof. Joaquín López-Serrano from the University of Seville, who performed DFT calculations on the interaction between PhSiH3 and Ph2SiH2 with 13. 68 Scheme 19 shows the energy breakdown of the bonding between the silanes and the metal complex. As it can be seen, the overall bonding electronic energy (ΔEb) is composed of two terms: the deformation energy (ΔEdef) that the fragments undergo upon their approach and the interaction energy (ΔEint) between the entities with the geometry that they adopt in the adducts. Scheme 19. Electronic energy breakdown of the overall bonding energy in the formation of 13·PhSiH3 and 13·Ph2SiH2 Table 3 displays the energies extracted from these calculations, and it clearly shows how diphenylsilane interacts more strongly than phenylsilane by 4.8 kcal/mol. This accounts for the stronger interaction of Ph2SiH2 with complex 13, and it may explain why there is no induction period when using these substrate regardless of the amine, giving rise to exponential (and not sigmoidal) kinetic plots. 68 Calculations were carried out at the M06/6-31g(d,p)/SDD level of theory. SMD (Dichloromethane) continuum was used to model the solvent. Catalytic processes assisted by Pt(II) σ-SiH complexes [359] PhSiH3 Ph2SiH2 ΔΔE ΔEdef (13) 4.2 5.3 1.1 ΔEdef (silane) 0.8 1.2 0.4 ΔEint -18.3 -23.1 -4.8 ΔEb -13.3 -16.6 -3.3 ΔEbDCM -7.9 -8.8 -0.9 Table 3. Electronic energies (kcal/mol) obtained from the DFT analysis of the interaction between 13 and PhSiH3 and Ph2SiH2. The interaction and bonding energies include basis set superposition error (BSSE, Counterpoise method) corrections. Origin of the selectivity Table 4 showed that catalyst loadings of at least 0.2 mol % are able to catalyse the formation of disilazanes when a silane:amine ratio of 1:2 is employed for non-coordinating amines such as tBuNH2, Et2NH or (CH2)4NH, being the process faster for nBuSiH3 than for PhSiH3. However, mixtures are observed in some cases depending on the reaction conditions. For instance, 0.5 mol % of 13 catalyses the reaction between phenylsilane and pyrrolidine in a 1:1 proportion, giving the mono- and disilazane in a 1:5.7 ratio along with unreacted silane. Addition of a second equivalent of amine selectively affords the product PhSiH[N(CH2)4]2. Nonetheless, entry 6 of Table 2 reveals that the selective formation of the monosilazane PhSiH2N[(CH2)4] is possible if the catalyst loading is decreased down to 50 ppm. The explanation behind this divergent behavior might be found in the nature of the active catalyst. It was mentioned at the beginning of Section 2.2 (see Dehydrocoupling of primary silanes) that the only Pt species remaining after the catalytic process was hydride 15. 69 Previously, it has been discussed in chapter 3 that 13 forms stronger σ-SiH complexes than 15, and indeed, this phenomenon 69 No decomposition was observed in any case, discarding the formation of solid platinum or any type of colloid or nanoparticles. Chapter 4. Results and discussion [360] has consequences that influence the outcome of some catalytic events such as carbon dioxide hydrosilation, as shown in section 2.1 (i.e. 15 is a less active catalyst than 13, since the degree of electrophilic activation on the silane is lower). In order to compare these facts with the Pt-assisted Si−N dehydrogenative coupling, a series of experiments were performed so as to obtain more information about the system: Hindered amines  Entries 7 and 8 of Table 2 demonstrate that 13 requires 12 min and 1.5 h (catalyst loadings of 0.3 and 0.1 mol %, respectively) to transform a 1:1 mixture of PhSiH3 and iPr2NH into the monosilazane PhSiH2(NiPr2). If this reaction is carried out with 0.5 mol % of 15 (generated in situ by reaction of 13 with H2), 65% conversion is observed after 5 hours, and 10 hours are needed to achieve full conversion.  Two consecutive additions of PhSiH3 and iPr2NH were carried out in a reaction flask initially containing 13. It was observed that the second run was much slower than the first one, indicative of a change in the nature of the catalyst. Non-hindered amines  The reaction between PhSiH3 and Et2NH catalyzed by 50 ppm of 13 (entry 3 of Table 2) was carried out in 3 consecutive runs. In this case, the rates of the first, second and third runs are practically the same, exhibiting similar kinetic plots (Figure 18). This suggests that the whole process is catalysed by the same species (presumably complex 15, given that 13 is readily hydrogenated to hydride 15 in the presence of H2). Catalytic processes assisted by Pt(II) σ-SiH complexes [361] Figure 18. Kinetic profile (H2 evolution) of the reaction (3 consecutive additions) between PhSiH3 and Et2NH (1 equivalent) under 50 ppm of catalyst 13.  An interesting result was observed when monitoring the released H2 during the reaction between PhSiH3 and pyrrolidine (2 equivalents) under 0.2 mol % of 13 (Entry 3 of Table 4). As it can be seen in Figure 19, there are two clear kinetic profiles. The first one corresponds to the first dehydrogenation event, and the second one, starting at ~50-60% of the expected amount of H2, corresponds to the dehydrocoupling of the newly formed monosilazane PhSiH2[N(CH2)4]. As expected for the employed substrates, the first kinetic profile is sigmoidal (pyrrolidine is a coordinating amine) and the second kinetic profile is exponential (as expected for a secondary silane, according to the previously mentioned experimental and theoretical results). Chapter 4. Results and discussion [362] Figure 19. Kinetic profile (H2 evolution) of the reaction of PhSiH3 with 2 equivalents of (CH2)4NH under 0.2 mol % of catalyst 13. These experiments seem to suggest that at low catalytic loadings (ppm), cyclometalated catalyst 13 is easily transformed into hydride complex 15 because of the formed H2, and 15 is responsible for the observed selectivity towards monosilazanes, given that the second Si−N dehydrogenative coupling is so extremely slow that no disilazane is detected. At higher catalytic loadings (from 0.2 mol % upwards), the amount of catalyst present in the reaction (either 13 or 15) is enough to carry out the mono and/or disubstitution at an appreciable reaction time. As shown above, these processes occur much faster when 13 is present, and the reaction rates will vary depending on the nature and amount of amine added. Mechanistic investigations: Low temperature NMR studies With the purpose of obtaining more experimental evidence about the mechanism by which these reactions occur, the reaction between Ph2SiH2 and Catalytic processes assisted by Pt(II) σ-SiH complexes [363] Et2NH was analysed at low temperature through 1H NMR spectroscopy. Initial studies in CD2Cl2 showed no trace of any intermediate from -80 °C to 25 °C. Thus, the solvent was switched to d8-THF due to its ability to coordinate to the metal and to lead to slower reaction rates. At -30 °C, a stoichiometric mixture of 13, Ph2SiH2 and Et2NH showed the presence of some unreacted complex 13 and neutral hydride 14 (Figure 20), which is presumably formed after hydride transfer from the η1 σ-SiH complex right after the nucleophilic attack of the amine, in the same way as the amine-borane dehydrocoupling mechanism described by our group (see Scheme 6 of chapter 3).60 Figure 20. Portion of the 1H NMR spectrum (400 MHz, d8-THF) of the reaction between 13, Ph2SiH2 and Et2NH (1 equivalent) acquired at -30 °C. Apart from 13 and 14, monosilazane Ph2SiH(NEt2) is also observed, and another unknown species (A) which seems to be present in a 1:1 ratio with Chapter 4. Results and discussion [364] complex 14. Regarding A, it shows signals that point to a NEt2 fragment, and it possesses no Si atom in close proximity according to 1H-29Si HMBC experiments, given that the only compounds with silicon are the some unreacted diphenylsilane and the previously mentioned monosilazane. The ethyl chains of A correlate (COSY) with a signal at around 7.6 ppm at -40 °C. Upon warming up the sample, the peaks corresponding to 14 and A simultaneously decrease until they disappear at 25 °C. Although our first tempting thought was to identify A as the silylium intermediate Ph2SiH(NHEt2)+ (in analogy to the boronium cation BH2(NHMe2)2+ observed during the dehydrocoupling of NMe2H·BH3),60 the lack of correlating signals of any proton of this intermediate with a silicon atom in the 1H−29Si HMBC experiments led us to propose a different type of ammonium cation. One reasonable possibility would be that species A corresponds to the cationic derivative [Et2NH2][BAr4F], generated from the deprotonation reaction of the short-lived cation Ph2SiH(NEt2H)+ by free diethylamine. In order to confirm this hypothesis, the ammonium salt [Et2NH2][BAr4F] was synthesized by reaction of Et2NH and [H(Et2O)2][BAr4F], and to our delight, the low temperature 1H NMR spectrum 70 of pure [Et2NH2][BAr4F] was nearly identical to that observed for intermediate A: 70 The NH2 protons of the ammonium cation are only visible below -30 °C. Catalytic processes assisted by Pt(II) σ-SiH complexes [365] Figure 21. 1H NMR of [Et2NH2][BAr4F] (400 MHz, d8-THF) at -50 °C. It can be seen that the chemical shifts of the alkyl chains perfectly match those highlighted in Figure 20. Moreover, the chemical shifts of the 15N NMR spectrum (obtained from a 1H-15N HMBC experiment) of both [Et2NH2][BAr4F] and intermediate A are nearly the same (49.4 and 49.1 ppm, respectively) thus providing further evidence of intermediate A being the ammonium salt [Et2NH2][BAr4F]. Therefore, all these data indicate that two NEt2H molecules are needed to generate the aminosilane Ph2SiH(NEt2) in the experiment carried out at low temperature, explaining the reasons why we observed some unreacted free Ph2SiH2 and complex 13 (that are only consumed when reaching room temperature). In fact, another identical experiment with 2 equivalents of amine instead of 1 led to the observation of species 14, A and Ph2SiH(NEt2), but neither free Ph2SiH2 nor 13 were detected, corroborating the need of 2 equivalents of amine per catalytic cycle. At the end of the experiment (room temperature), cationic hydride 15 is observed. This is in agreement with the formation of H2 Chapter 4. Results and discussion [372] As a final comment, it is necessary to mention that the catalytic process studied in this section is highly complex due to the multiple phenomena that can take place simultaneously, since complex 13 transforms into 15 during the reaction as a consequence of the newly formed H2. Moreover, both complexes can efficiently catalyse the dehydrocoupling reaction, making the analysis even more difficult. For these reasons, kinetic data such as the KIE values or the partial orders must be cautiously taken. For instance, the partial order of 2.2 in catalyst is usually related to the involvement of 2 metal atoms in the reaction. However, neither experimental nor computational studies seem to lead in this direction. Instead, this value must be associated with a different phenomenon, such as the progressive change in the nature of the active catalyst from 13 to 15 (Scheme 21A). 76 In light of these results, and taking into consideration the complexity of this process, a plausible catalytic cycle can be proposed (Scheme 21B): 76 T. Rosner, J. L. Bars, A. Pfaltz, D. G. Blackmond, J. Am. Chem. Soc., 2001, 123, 1848-1855. In this example, the role of an inactive Pd species in the partial order determination is described. Catalytic processes assisted by Pt(II) σ-SiH complexes [373] Scheme 21. Proposed overall catalytic cycle of the Si−N dehydrogenative coupling catalyzed by complex 13. Chapter 4. References [374] References 1 M. C. Lipke, A. L. Liberman-Martin, T. D. Tilley, Angew. Chem. Int. Ed., 2017, 56, 2260-2294. 2 D. H. Binh, M. Milovanović, J. Puertes-Mico, M. Hamdaoui, S. D. Zarić, J-P. Djukic, Chem. Eur. J., 2017, 23, 17058-17069. 3 V. Y. Lee, A. Sekiguchi in Organosilicon Compounds, 2017, Academic Press. Chapter 5: Silicon-centered cations, 197-230. 4 P. D. Bartlett, F. E. Condon, A. Schneider, J. Am. Chem. Soc., 1944, 66, 1531-1539. 5 H. F. T. Klare, K. Bergander, M. Oestreich, Angew. Chem. Int. Ed., 2009, 48, 9077- 9079. 6 K. Müther, M. Oestreich, Chem. Commun., 2011, 47, 334-336. 7 M. Reissmann, A. Schafer, S. Jung, T. Müller, Organometallics, 2013, 32, 6736- 6744. 8 M. F. S. Valverde, E. Theuergarten, T. Bannenberg, M. Freytag, P. G. Jones, M. Tamm, Dalton Trans., 2015, 44, 9400-9408. 9 N. von Wolff, G. Lefèvre, J.-C. Berthet, P. Thuéry, T. Cantat, ACS Catal., 2016, 6, 4526-4535. 10 A. Schäfer, W. Saak, D. Haase, T. Müller, Angew. Chem. Int. Ed., 2012, 51, 2981- 2984. 11 L. Omann, Z-W. Qu, E. Irran, H. F. T. Klare, S. Grimme, M. Oestreich, Angew. Chem. Int. Ed., 2018, 57, 8301-8305. 12 S. Cummings, H. P. Hratchian, C. A. Reed, Angew. Chem. Int. Ed., 2016, 55, 1382- 1386. 13 C. A. Reed, Acc. Chem. Res., 2013, 46, 2567-2575. 14 M. Brookhart, B. Grant, A. F. Volpe Jr., Organometallics, 1992, 11, 3920-3922. 15 E. Lukevics, M. Dzintara, J. Organomet. Chem., 1985, 295, 265-315. 16 X-L. Luo, R. H. Crabtree, J. Am. Chem. Soc., 1989, 111, 2527-2535. Catalytic processes assisted by Pt(II) σ-SiH complexes [375] 17 J. M. Blackwell, K. L. Foster, V. H. Beck, W. E. Piers, J. Org. Chem., 1999, 64, 4887- 4892. 18 a) B-H. Kim, M-S. Cho, H-G. Woo, Synlett, 2004, 5, 761-772; b) D. J. Harrison, D. R. Edwards, R. McDonald, L. Rosenberg, Dalton Trans., 2008, 3401-3411; c) E. M. Leitao, T. Jurca, I. Manners, Nat. Chem., 2013, 5, 817-829; d) R. L. Melen, Chem. Soc. Rev., 2016, 45, 775-788. 19 a) V. Gevorgyan, J-X. Liu, M. Rubin, S. Benson, Y. Yamamoto, Tetrahedron Lett., 1999, 40, 8919-8922; b) V. Gevorgyan, M. Rubin, S. Benson, J-X. Liu, Y. Yamamoto, J. Org. Chem., 2000, 65, 6179-6186. 20 J. Yang, P. S. White, M. Brookhart, J. Am. Chem. Soc., 2008, 130, 17509-17518. 21 T. Robert, M. Oestreich, Angew. Chem. Int. Ed., 2013, 52, 5216-5218. 22 a) H. Koinuma, F. Kawakami, H. Kato, H. Hirai, J. Chem. Soc. Chem. Commun., 1981, 213-214; b) G. Süss-Fink, J. Reiner, J. Organomet. Chem., 1981, 221, C36- C38. 23 Some recent reports include the following: a) F. J. Fernández-Álvarez, A. M. Aitani, L. A. Oro, Catal. Sci. Technol., 2014, 4, 611-624; b) C. Chauvier, T. Cantat, ACS Catal., 2017, 7, 2107-2115; c) F. J. Fernández-Álvarez, L. A. Oro, ChemCatChem, 2018, 10, 4783-4796 24 P. Deglmann, E. Ember, P. Hofmann, S. Pitter, O. Walter, Chem. Eur. J., 2007, 13, 2864-2879. 25 K. Motokura, D. Kashiwame, N. Takahashi, A. Miyaji, T. Baba, Chem. Eur. J., 2013, 19, 10030-10037. 26 L. Zhang, J. Cheng, Z. Hou, Chem. Commun., 2013, 49, 4782-4784. 27 S. Itagaki, K. Yamaguchi, N. Mizuno, J. Mol. Catal. A: Chem., 2013, 366, 347-352. 28 L. González-Sebastián, M. Flores-Alamo, J. J. García, Organometallics, 2013, 32, 7186-7194. 29 C. Chauvier, P. Thuéry, T. Cantat, Angew. Chem. Int. Ed., 2016, 55, 14096-14100. 30 J. Takaya, N. Iwasawa, J. Am. Chem. Soc., 2017, 139, 6074-6077. Chapter 4. References [376] 31 A. Julián, J. Guzmán, E. A. Jaseer, F. J. Fernández-Álvarez, R. Royo, V. Polo, P. García-Orduña, F. J. Lahoz, L. A. Oro, Chem. Eur. J., 2017, 23, 11898-11907. 32 a) A. Jansen, H. Görls, S. Pitter, Organometallics, 2000, 19, 135-138; b) M. L. Scheuermann, S. P. Semproni, I. Pappas, P. J. Chirik, Inorg. Chem., 2014, 53, 9463- 9465. 33 L. E. Gusel’nikov, N. S. Nametkin, Chem. Rev., 1979, 79, 529-577. 34 I. Manners, Angew. Chem. Int. Ed., 1996, 35, 1602-1621. 35 a) Information about disilenes: R. West, Angew. Chem. Int. Ed., 1987, 26, 1201- 1211; b) Information about iminoboranes: P. Paetzold, Adv. Inorg. Chem., 1987, 31, 123-170. 36 A. Soum in Silicon-containing Polymers. The Science and Technology of their Synthesis and Applications, 2000, Springer Netherlands. Chapter 11: Polysilazanes, 323-349. 37 M. Birot, J-P. Pillot, J. Dunoguès, Chem. Rev., 1995, 95, 1443-1477. 38 This family of polysilazanes has been commercialized by Merck KGaA under the name Durazane® . For more information, see: https://www.merckgroup.com/en/brands/pm/durazane.html. 39 R. Fessenden, J. S. Fessenden, Chem. Rev., 1961, 61, 361-388. 40 S. Austin, A. Glowacki. Hydrochloric acid. Ullmann’s Encyclopedia of Industrial Chemistry, 2000, doi: 10.1002/14356007.a13_283. 41 a) R. Noyori, T. Ohkuma, Angew. Chem. Int. Ed., 2001, 40, 40-73; b) S. Werkmeister, K. Junge, M. Beller, Org. Process Res. Dev., 2014, 18, 289-302; c) G. A. Filonenko, R. V. Putten, E. J. M. Hensen, E. A. Pidko, Chem. Soc. Rev., 2018, 47, 1459-1483. 42 Metal-free catalytic systems have also been described in the literature (see reference 18d). 43 F. Buch, S. Harder, Organometallics, 2007, 26, 5132-5135. 44 J. F. Dunne, S. R. Neal, J. Engelkemier, A. Ellern, A. D. Sadow, J. Am. Chem. Soc., 2011, 133, 16782-16785. Catalytic processes assisted by Pt(II) σ-SiH complexes [377] 45 R = Ph, Et or C6H13. dppm stands for bis(diphenylphosphino)methane. 46 W-D. Wang, R. Eisenberg, Organometallics, 1991, 10, 2222-2227. 47 S. Itagaki, K. Kamata, K. Yamaguchi, N. Mizuno, Chem. Commun., 2012, 48, 9269-9271. 48 C. D. F. Königs, M. F. Müller, N. Aiguabella, H. F. T. Klare, M. Oestreich, Chem. Commun., 2013, 49, 1506-1508. 49 W. Xie, H. Hu, C. Cui, Angew. Chem. Int. Ed., 2012, 51, 11141-11144. 50 A. E. Nako, W. Chen, A. J. P. White, M. R. Crimmin, Organometallics, 2015, 34, 4369-4375. 51 M. P. Cibuzar, R. Waterman, Organometallics, 2018, 37, 4395-4401. 52 M. S. Hill, D. J. Liptrot, D. J. MacDougall, M. F. Mahon, T. P. Robinson, Chem. Sci., 2013, 4, 4212-4222. 53 C. Bellini, J-F. Carpentier, S. Tobisch, Y. Sarazin, Angew. Chem. Int. Ed., 2015, 54, 7679-7683. 54 a) C. Bellini, V. Dorcet, J-F. Carpentier, S. Tobisch, Y. Sarazin, Chem. Eur. J., 2016, 22, 4564-4583; b) C. Bellini, T. Roisnel, J-F. Carpentier, S. Tobisch, Y. Sarazin, Chem. Eur. J., 2016, 22, 15733-15744. 55 J. Yang, P. S. White, C. K. Schauer, M. Brookhart, Angew. Chem. Int. Ed., 2008, 47, 4141-4143. 56 S. Park, D. Bézier, M. Brookhart, J. Am. Chem. Soc., 2012, 134, 11404-11407. 57 The conversion was measured by 1H NMR integration of the silane and silyl formate peaks. 58 Aromatic silanes like PhSiH3 were also studied, although several unidentified species were observed rather than silyl formate derivatives. 59 J. Zhu, Z. Lin, T. B. Marder, Inorg. Chem., 2005, 44, 9384-9390. 60 M. Roselló-Merino, J. López-Serrano, S. Conejero, J. Am. Chem. Soc., 2013, 135, 10910-10913. 61 M. Roselló-Merino, R. J. Rama, J. Díez, S. Conejero, Chem. Commun., 2016, 52, 8389-8392. Chapter 4. References [378] 62 J. X. Wang, A. K. Dash, J. C. Berthet, M. Ephritikhine, M. S. Eisen, J. Organomet. Chem., 2000, 610, 49-57. 63 S. Anga, Y. Sarazin, J-F. Carpentier, T. K. Panda, ChemCatChem, 2016, 8, 1373- 1378. 64 http://www.manonthemoontech.com/x102-gas-evolution.html. 65 “Síntesis de complejos coordinativamente insaturados de Pt(II) estabilizados por ligandos carbeno N-heterocíclicos. Activación y funcionalización de enlaces C−H”. O. Rivada-Wheelaghan. PhD Thesis, 2013. 66 M. A. Ortuño, S. Conejero, A. Lledós, Beilstein J. Org. Chem., 2013, 9, 1352-1382. 67 The initial increase of pressure above 100% in the plot on the left is attributed to the partial evaporation of the low-boiling point solvent (CH2Cl2) due to the exothermic nature of the reaction. When the process is slower (plot on the right) the heat is gradually released, avoiding solvent evaporation. 68 Calculations were carried out at the M06/6-31g(d,p)/SDD level of theory. SMD (Dichloromethane) continuum was used to model the solvent. 69 No decomposition was observed in any case, discarding the formation of solid platinum or any type of colloid or nanoparticles. 70 The NH2 protons of the ammonium cation are only visible below -30 °C. 71 Calculations were carried out at the PBE0-D3/6-31g(d,p)/SDD level of theory. SMD (Dichloromethane) continuum was used to model the solvent. 72 Calculations discard mechanisms involving Si−H oxidative addition. 73 M. Besora, A. Lledós, F. Maseras, Chem. Soc. Rev., 2009, 38, 957-966. 74 This KIE was obtained in the reaction Ph2SiH2 + Et2ND (due to the availability of deuterated reagents). No KIE was detected when using Ph2SiD2 (kH/kD = 0.88). 75 O. Rivada-Wheelaghan, M. Roselló-Merino, M. A. Ortuño, P. Vidossich, E. Gutiérrez-Puebla, A. Lledós, S. Conejero, Inorg. Chem., 2014, 53, 4257-4268. 76 T. Rosner, J. L. Bars, A. Pfaltz, D. G. Blackmond, J. Am. Chem. Soc., 2001, 123, 1848-1855. In this example, the role of an inactive Pd species in the partial order determination is described. Experimental Part Experimental Part [380] Experimental Part General Methods All manipulations were carried out using standard Schlenk and glovebox techniques, under an atmosphere of argon and of high purity nitrogen, respectively. All solvents were dried and degassed prior to use. n-pentane, toluene and tetrahydrofuran were distilled under nitrogen over sodium and stored under Na/K alloy. Dichloromethane was distilled under nitrogen over calcium hydride and stored under 3 Å activated molecular sieves. Hexamethyldisiloxane was degassed (freeze-pump-thaw) and stored under 3 Å activated molecular sieves. Benzene-d6, toluene-d8 and tetrahydrofuran-d8 were distilled under argon over sodium/benzophenone and stored under 3 Å activated molecular sieves. Dichloromethane-d2 was heated under reflux over calcium hydride for several hours, distilled under argon and stored under 3 Å activated molecular sieves. Solution NMR spectra were acquired on Bruker AMX-300, DRX-400 and DRX-500 spectrometers at 298 K unless otherwise specified. 1H and 13C chemical shifts were referenced to the residual solvent peak. 31P NMR spectra were referenced to an external 85% solution of H3PO4. 19F NMR spectra were referenced to an external standard of CFCl3. 11B NMR spectra were referenced to an external standard of BF3·Et2O. 29Si chemical shifts were referenced to an external standard of Si(CH3)4. Spectral assignments were made by means of 1D and 2D NMR experiments (1H, 13C, 13C{1H}, COSY, NOESY, HSQC and HMBC). Elemental analysis was carried out with a LECO TruSpec CHN elementary analyser. Infra-red spectra (Nujol on NaCl plates) were measured on a Bruker Vector 22 instrument. Experimental Part [381] Mass spectra were recorded on a Bruker Esquire 6000 instrument with an ESI ion source and ion trap analyser. HR-MS spectra were recorded on a QExactive mass spectrometer with an APCI (Atmospheric Pressure Chemical Ionization) source in positive mode at the CITIUS central services of the University of Seville. Flash column chromatography was carried out on Merck Silica Gel 60 (0.063- 0.200 mm), and the process was monitored by using TLC pre-coated silica gel plates (Merck Silica Gel 60 F254). Spots were visualized by means of UV light (254 nm) or using a solution of potassium permanganate or cerium molybdate (Hanessian’s stain). Tris(pentaflorophenyl)borane was purchased from TCI and sublimed at 100 °C under static vacuum. Bis(catecholato)diboron and bis(pinacolato)diboron were purchased from Sigma-Aldrich and sublimed at 100 °C (former) and 70 °C (latter) under dynamic vacuum. Triethylsilane, dimethylphenylsilane, diphenylmethylsilane, diethylsilane and n-butylsilane were distilled from CaH2 and stored under 3 Å activated molecular sieves. All other commercially available reagents were purchased from Sigma-Aldrich, TCI or Alfa Aesar and used as received. Carbon dioxide, carbon monoxide and 13C-labeled carbon dioxide (99.05% 13C) were supplied by Air Liquide (12CO2 and CO) and Euriso-top (13C), and used as received. Experimental Part [388] Experimental Part for Chapter 2 Synthesis of PBP-stabilized Ni (II) complexes (tBuPBP)Ni−BCat (11) Synthesis In the glovebox, a solution of complex 4 (31 mg, 0.061 mmol) in dry tetrahydrofuran (0.5 mL) was added to a vial that contains B2cat2 (16 mg, 0.067 mmol, 1.1 eq). The resulting suspension was manually stirred until a clear yellow solution was observed (i.e. until both reagents are completely dissolved). At this point, hexamethyldisiloxane was slowly added with manual stirring until a cloudy solution/precipitate in suspension was formed. Then, the minimum amount of tetrahydrofuran (a few drops) was added so as to redissolve the sample. The vial was closed and placed into the freezer of the glovebox. After 2- 3 days, yellow crystals of complex 11 were formed. By the time this thesis is being written, isolation of the crystals to determine a yield and complete the characterization is ongoing. Spectroscopic and analytical data Yellow crystals. Molecular weight: 610.99 g/mol. 1H-NMR (400 MHz, d8-THF, 298 K) δ 1.30 (vt, 36H, 3JH-P = 6.5 Hz, tBu), 4.08 (s, 4H, CH2), 6.80, 6.92, 7.03 and 7.18 (m, 8 H, arom CH) ppm. Experimental Part [389] 13C{1H}-NMR (100 MHz, d8-THF, 298 K) δ 30.2 ((CH)3C), 36.4 (t, 1JC-P = 7.5 Hz, (CH)3C), 43.7 (t, 1JC-P = 17.3 Hz, CH2), 109.7, 111.1, 118.6, 120.8 (aromatic CH), 140.2 (t, 3JC-P = 7.3 Hz, ligand Cq), 151.6 (BCat Cq) ppm. 31P{1H}-NMR (161 MHz, d8-THF, 298 K) δ 114.9 (s, PtBu2) ppm. 11B-NMR (128 MHz d8-THF, 298 K) δ 47.24 and 56.2 (bs, boryl) ppm. (tBuPBP)Ni−Ph (12) In the glovebox, a previously flame-dried J. Young flask was charged with complex 2 (100 mg, 0.175 mmol) and a stirrer bar. Toluene (12 mL) was added under argon, after which a yellow solution was obtained. Then, a solution of PhMgBr (0.23 mL, 3 M in Et2O, 0.7 mmol) was added, and the resulting orange solution was stirred at 80 ⁰C for 22 h. The solvent was removed under reduced pressure, and the dark solid obtained was extracted with pentane (20 mL) and filtered via cannula. The resulting yellow solution was evaporated under vacuum to obtain a yellow solid which was washed with hexamethyldisiloxane (3 mL) at -40 ⁰C, affording a bright yellow solid (51 mg, 0.090 mmol, 51% yield). Yellow crystals suitable for X-Ray diffraction analysis were obtained by cooling at -25 ⁰C a concentrated solution in toluene. Spectroscopic and analytical data Yellow crystals or solid. Molecular weight: 569.19 g/mol. Experimental Part [390] 1H-NMR (400 MHz, C6D6, 298 K) δ 1.13 (vt, 36 H, 3JHP= 6.2 Hz, tBu), 3.70 (m, 4 H, CH2), 6.98 (dd, 2 H, 3JHH = 5.6 Hz, 4JHH = 3.4 Hz, ligand aromatic CH), 7.08 (t, 1H, 3JHH = 7.3 Hz, CH-Phpara) 7.16 (dd, 2 H, 3JHH = 5.4 Hz, 4JHH = 3.4 Hz, ligand aromatic CH), 7.33 (t, 2H, 3JHH = 7.3 Hz, CH-Phmeta), 8.09 (d, 2H, 3JHH = 7.1 Hz, CH-Phortho). 13C{1H}-NMR (100 MHz, C6D6, 298 K) δ 30.1 (t, 2JCP = 2.5 Hz, (CH)3C), 36.5 (t, 1JCP = 5.5 Hz, (CH)3C), 42.3 (t, 1JCP = 17.2 Hz, CH2), 109.0 (ligand aromatic CH), 118.6 (ligand aromatic CH), 121.0 (CH-Phpara), 125.4 (CH-Phmeta), 139.6 (t, 3JCP = 7.3 Hz, ligand aromatic Cq), 143.4 (CH-Phortho), 176.2 (t, 2JCP = 14 Hz, Ph-Cipso). 31P{1H}-NMR (161 MHz, C6D6, 298 K) δ 93.8 (s, PtBu2) ppm. 11B{1H}-NMR (128 MHz, C6D6, 298 K) δ 44.9 (bs, boryl) ppm. Anal. Calcd. for C30H49BN2NiP2: C, 63.31; H, 8.68; N, 4.92. Found: C, 63.22; H, 8.52; N, 4.79. General procedure for low temperature NMR experiments In the glovebox, a solution of 4 (15 mg, 0.030 mmol) was dissolved in d8-THF (0.4 mL) and placed into a J. Young NMR tube. The same was repeated for a solution of B2cat2 (8.4 mg, 0.035 mmol, 1.1 eq) in d8-THF (0.2 mL) in a different NMR tube. In the Schlenk line, both tubes were cooled down to -60 °C and the solution of diborane was transferred under argon via cannula to the NMR tube containing the nickel complex. The NMR tube was closed and shaken in the cold bath, after which it was introduced into the pre-cooled (-60 °C) NMR spectrometer. Once the mixture of 4 + Intermediate was observed by NMR, PMe3 (up to 4 equivalents) or CO (1 bar) was added to the sample under inert atmosphere at low temperature (-60 °C, cold bath). The NMR tube was brought back to the NMR spectrometer to monitor the evolution of the sample at different temperatures. Experimental Part [391] General procedure for the borylation of styrenes and isolation of the products. Example of styrene and B2pin2 (1 equivalent) using DMF as additive (entry 2 of Table 5) In the glovebox, a vial was charged with complex 4 (4 mg, 0.008 mmol, 10 mol %), B2pin2 (20.3 mg, 0.08 mmol), styrene (9.2 μL, 0.08 mmol), and dry DMF (12.4 μL, 0.16 mmol). The resulting yellow solution was transferred to a J. Young NMR tube, which was heated to 70 °C for 14 hours. Then, the sample was placed into an HPLC vial, and the solvent was removed under vacuum. Heptane (2 mL) was added to the brown residue and the solvents were removed again under vacuum. This process was repeated 2 more times. The resulting brown solid was purified by column chromatography (ø = 1.2 cm, 11 cm of silica), eluting with Et2O- Hexane 1:9, affording the pure (E)-monoborylated styrene. Dehydrogenative borylation products (E)-4,4,5,5-tetramethyl-2-styryl-1,3,2-dioxaborolane 1H-NMR (400 MHz, CDCl3, 298 K): δ 1.32 (s, 12H, CH3), 6.18 (d, 3JH-H = 18.6 Hz, 1H, =CH), 7.32 (m, 3H, arom CH), 7.40 (d, 3JH-H = 18.7 Hz, 1H, =CH), 7.50 (m, 2H, arom CH) ppm. 11B-NMR (128 MHz CDCl3, 298 K) δ 30.2 (Bpin) ppm. (E)-4,4,5,5-tetramethyl-2-(4-(trifluoromethyl)styryl)-1,3,2-dioxaborolane Experimental Part [392] 1H-NMR (400 MHz, d8-Toluene, 298 K): δ 1.13 (s, 12H, CH3), 6.24 (d, 3JH-H = 18.4 Hz, 1H, =CH), 7.07 (d, 3JH-H = 8.1 Hz, 2H, arom CH), 7.23 (d, 3JH-H = 8.1 Hz, 2H, arom CH), 7.40 (d, 3JH-H = 18.2 Hz, 1H, =CH) ppm. (E)-4,4,5,5-tetramethyl-2-(2-phenylprop-1-en-1-yl)-1,3,2-dioxaborolane 9 1H-NMR (400 MHz, CDCl3, 298 K): δ 1.32 (s, 12H, OCCH3), 2.41 (d, 4JH-H = 0.9 Hz, 3H, C=CCH3), 5.76 (broad quartet, 4JH-H = 0.9 Hz, 1H, =CH), 7.31 (m, 3H, arom CH), 7.50 (m, 2H, arom CH) ppm. (E)-2-(4-fluorostyryl)-4,4,5,5-tetramethyl-1,3,2-dioxaborolane 1H-NMR (400 MHz, d8-Toluene, 298 K): δ 1.15 (s, 12H, CH3), 6.13 (d, 3JH-H = 18.4 Hz, 1H, =CH), 6.68 (m, 2H, arom CH), 7.06 (m, 2H, arom CH), 7.45 (d, 3JH-H = 18.5 Hz, 1H, =CH) ppm. (E)-2-(4-methoxystyryl)-4,4,5,5-tetramethyl-1,3,2-dioxaborolane 9 The assignation of the E-isomer was made based on the coupling constants and chemical shifts observed by 1H NMR, which match those described in the literature. See: S. Tanaka, Y. Saito, T. Yamamoto, T. Hattori, Org. Lett., 2018, 20, 1828-1831 (Supporting information). Experimental Part [393] 1H-NMR (400 MHz, d8-Toluene, 298 K): δ 1.13 (s, 12H, CH3), 3.31 (s, 3H, OCH3) 6.14 (d, 3JH-H = 18.4 Hz, 1H, =CH), 6.60 (d, 3JH-H = 8.5 Hz, 2H, arom CH), 7.22 (d, 3JH- H = 8.8 Hz, 2H, arom CH), 7.54 (d, 3JH-H = 18.5 Hz, 1H, =CH) ppm. 2,2'-(2-phenylethene-1,1-diyl)bis(4,4,5,5-tetramethyl-1,3,2- dioxaborolane) 1H-NMR (500 MHz, CDCl3, 298 K): δ 1.27 (s, 12H, CH3), 1.31 (s, 12H, CH3), 7.33 (m, 3H, arom CH), 7.47 (m, 2H, arom CH), 7.70 (s, 1H, =CH) ppm. 2,2'-(2-(4-fluorophenyl)ethene-1,1-diyl)bis(4,4,5,5-tetramethyl-1,3,2- dioxaborolane) 1H-NMR (400 MHz, d8-Toluene, 298 K): δ 1.07 (s, 12H, CH3), 1.15 (s, 12H, CH3), 6.99 (m, 2H, arom CH), 7.46 (m, 2H, arom CH), 7.88 (s, 1H, =CH) ppm. Experimental Part [394] 2,2'-(2-(4-methoxyphenyl)ethene-1,1-diyl)bis(4,4,5,5-tetramethyl-1,3,2- dioxaborolane) 1H-NMR (400 MHz, d8-Toluene, 298 K): δ 1.15 (s, 24H, CH3), 3.31 (s, 3H, OCH3), 6.67 (d, 3JH-H = 8.6 Hz, 2H, arom CH), 7.47 (d, 3JH-H = 8.7 Hz, 2H, arom CH), 7.98 (s, 1H, =CH) ppm. Experimental Part [395] Experimental Part for Chapter 3 Synthesis of Pt ItBu complexes Complex [Pt(SiHEt2)(ItBu)2][BAr4F] (16) Synthesis A J. Young flask was charged with a stirrer bar, complex [Pt(ItBu’)(ItBu)][BAr4F], 13, (190 mg, 134 μmol) and CH2Cl2 (3 mL). Et2SiH2 (0.52 mL, 4.02 mmol, 30 eq) was added to the solution, and the mixture was stirred at room temperature for 15 min, after which a change of color from orange to yellow was observed. A layer of pentane was then added (30 mL), and slow mixing of both solvents yielded yellow crystals suitable for X-Ray analysis. The solution was removed via cannula, the crystals were dried under vacuum at -40 °C and stored at -20 °C in the glovebox (170 mg, 113 μmol, 84% yield). Complex 16 is thermally unstable towards cyclometalation, but it can be stored for months at -20 °C without decomposition. Spectroscopic and analytical data Yellow crystals. Molecular weight: 1506.11 g/mol. 1H-NMR (400 MHz, CD2Cl2, 298 K): δ 0.19 (m, 2H, CH2CH3), 0.63 (m, 2H, CH2CH3), 0.69 (m, 6H, CH2CH3), 1.96 (s, 36H, (CH3)3C), 3.64 (m + d + d, 2JPt-H = 115 Hz, 1JH-Si = 188 Hz, 1H, SiH), 7.26 (s, 4H, =CH) ppm. Experimental Part [396] 13C{1H}-NMR (100 MHz, CD2Cl2, 298 K): δ 10.1 (CH3CH2), 10.5 (CH3CH2), 33.4 ((CH3)3C), 60.7 ((CH3)3C), 120.1 (=CH), 173.8 (Ccarbene-Pt) ppm. 29Si{1H}-NMR (79 MHz, CD2Cl2, 298 K): δ 18.12 (SiHEt2) ppm. LRMS (ESI): m/z calculated for C26H51N4PtSi [M+]: 642.35. Found: 642.4 Elem. Anal. Calcd. for C58H63BF24N4PtSi: C, 46.25; H, 4.22; N, 3.72. Found: C, 46.45; H, 3.97; N, 3.47. Variable Temperature NMR experiments using Pt ItBu complexes (13, 15 and 16) A J. Young NMR tube was charged with [Pt] (around 40 mg) and dissolved in CD2Cl2 (0.4 mL). The system was brought to the NMR spectrometer, where it was analysed (1H NMR) at room temperature before lowering the temperature to -80 °C. At this point, the tube was taken out of the spectrometer and introduced in a cold (-78 °C) bath. 1 equivalent of silane was added under argon, and the tube was then introduced again in the precooled NMR spectrometer. The behavior of the system was monitored by 1H NMR spectroscopy at different temperatures. In the case of [Pt(H)(ItBu)2][BAr4F], 15, the J. Young NMR tube was initially charged with [Pt(ItBu’)(ItBu)][BAr4F], 13, CD2Cl2 and H2 (2 bar). Once the hydride complex was formed, the solution was degassed with 4 freeze-pump- thaw cycles and brought to the NMR spectrometer as in the other cases. Experimental Part [397] Synthesis of Pt ItBuiPr complexes 1-(tert-butyl)-3-(iso-propyl)-1H-imidazolium iodide, ItBuiPr·HI (17) Synthesis In the Schlenk line, a J. Young flask was charged with a stirrer bar, 1-(tertbutyl)-1H-imidazol (3.2 g, 25.8 mmol) and 2-iodopropane (5.44 g, 32 mmol). Toluene (30 mL) was added, and the mixture was heated up to 100 °C for 24 hours, after which the resulting white solid was filtered off and washed with diethyl ether (3 x 30 mL). 7.1 g (24.1 mmol, 94 % yield) of an analytically pure white solid was obtained. Spectroscopic and analytical data White solid. Molecular weight: 294.18 g/mol. 1H-NMR (400 MHz, CD2Cl2, 298 K): δ 1.62 (d, 3JH-H = 6.8 Hz, 6H, CH(CH3)2), 1.73 (s, 9H, C(CH3)3), 5.11 (sept, 3JH-H = 6.8 Hz , 1H, CH(CH3)2), 7.48 (br, 2H, =CH), 10.06 (s, 1H, =CH) ppm. 13C{1H}-NMR (100 MHz, CD2Cl2, 298 K): δ 23.2 (CH(CH3)2), 30.3 (C(CH3)3), 53.6 (CH(CH3)2), 61.0 (C(CH3)3), 119.9 and 120.0 (=CH), 134.5 (=CH) ppm. Elem. Anal. Calcd. for C10H19IN2: C, 40.83; H, 6.51; N, 9.52. Found: C, 40.9; H, 6.3; N, 9.7.