Liganden und Rezeptoren der TNF-Familie regulieren eine Vielzahl zellulärer Prozesse, darunter Apoptose und Immunprozesse. TNF-Liganden kommen in Form löslicher und membranständiger trimerer Moleküle vor, wobei die trimere Organisation durch die konservierte THD vermittelt wird. Im Gegensatz zu den membranständigen Molekülen können lösliche TNF-Liganden nicht immer an ihren TNF-Rezeptor binden oder ihn effektiv aktivieren. Für zwei solcher inaktiven TNF-Liganden, nämlich TRAIL und CD95L, konnte gezeigt werden, dass durch sekundäre Oligomerisierung oder durch artifizielle Herstellung einer Membranständigkeit mittels Antikörperdomänen gegen zelloberflächenexprimierte Proteine hochaktive Ligandenvarianten generiert werden können. Inwieweit sich diese Verfahren auf die T-Zell-kostimulatorischen TNF-Liganden OX40L, 41BBL und CD27L übertragen lassen, wurde in dieser Arbeit untersucht. Lösliche Flag- und Flag-TNC-Varianten von OX40L und 41BBL zeigten eine gute Bindung an die Rezeptoren OX40 und 41BB. Die lösliche Variante Flag-CD27L konnte nicht an ihren Rezeptor CD27 binden. Dies war aber nach Einführung der trimerstabilisierenden TNC-Domäne möglich. Eine effektive Aktivierung ihres Rezeptors, nachgewiesen durch Analyse der IL8-Induktion, bewirkten die löslichen TNF-Ligandenvarianten nur nach sekundärer Oligomerisierung mittels des Flag-spezifischen Antikörpers M2. Eine ähnlich gute TNFR-Aktivierung ließ sich durch Einführung der hexamerisierenden Fc-Domäne erzielen. Fc-Flag-OX40L und Fc-Flag-41BBL induzierten bereits ohne sekundäre Quervernetzung effektiv IL8. Die Hexamerisierung alleine reichte für die lösliche CD27L-Variante nicht aus, hier war zusätzlich zur Fc- wiederum auch die TNC-Domäne erforderlich, um die Bindung an CD27 und eine schwache IL8-Induktion zu erzielen. Für die FAP-bindenden Fusionsproteine antiFAP-Flag- OX40L, antiFAP-Flag-41BBL und antiFAP-Flag-TNC-CD27L war die Bindung an OX40, 41BB und CD27 sowie an FAP nachweisbar. Erst durch die artifizielle Membranständigkeit nach Bindung an FAP konnten diese Fusionsproteine über ihren Rezeptor effektiv IL8 induzieren. Zusammenfassend ließ sich somit zeigen, dass sich schwach oder nicht aktive lösliche Ligandenvarianten von OX40L und 41BBL durch sekundäre Oligomerisierung, durch die Fc-Hexamerisierungsdomäne und durch artifizielle Membranständigkeit in hochaktive Liganden verwandeln lassen. Lösliche CD27L-Varianten benötigen zusätzlich die trimerstabilisierende TNC-Domäne, um CD27 binden und aktivieren zu können. Für das bessere Verständnis der Ligand-Rezeptor-Interaktionen wurden zusätzlich OX40L-, 41BBL- und CD27L-Fusionsproteine mit der hochaktiven Gaussia princeps Luziferase (GpL) generiert, um Gleichgewichtsbindungs-, Dissoziationsstudien und homologe Kompetitionsassays durchführen zu können. Für die Fusionsproteine GpL-Flag-TNC-OX40L, GpL-Flag-TNC-41BBL und GpL-Flag-TNC-CD27L konnte gezeigt werden, dass die IL8-Induktion nicht von der Rezeptorbelegung abhängt, sondern von der sekundären Oligomerisierung, da bei gleicher Rezeptorbelegung durch sekundär quervernetzte TNF-Liganden mehr IL8 induziert wird, die Rezeptoraktivierung also qualitativ besser sein muss. / Ligands and receptors of the TNF family regulate diverse cellular processes such as apoptosis and immune processes. TNF ligands are soluble or membrane-bound molecules with a trimeric organization mediated by the conservative THD. Unlike the membrane-bound molecules soluble TNF ligands may fail in binding or in activating their receptor efficiently. It was shown for two such inactive TNF ligands, namely TRAIL and CD95L that highly active ligand variants could be generated with secondary oligomerization or artificial cell surface immobilization through antibody domains recognizing cell surface expressed proteins. In this work it was examined if these procedures are also applicable to the T cell costimulating TNF ligands OX40L, 41BBL and CD27L. Soluble Flag- and Flag-TNC-variants of OX40L and 41BBL showed good binding to their receptors OX40 and 41BB. The soluble variant Flag-CD27L did not bind its receptor CD27 while that was possible after introduction of the trimer stabilizing TNC-domain. An effective receptor-activation proved by analysis of the induction of IL8 was gained by soluble TNF ligand variants only after secondary oligomerization with the Flag-specific antibody M2. A similar efficient TNFR-activation was achieved by introduction of a hexamerizing Fc-domain. Fc-Flag-OX40L and Fc-Flag-41BBL induced already without secondary cross-linking efficiently IL8. For the soluble CD27L-variant hexamerization alone was not sufficient, but the Fc-domain was necessary in addition to the TNC-domain to enable binding to CD27 and weak induction of IL8. For the FAP-binding fusion proteins antiFAP-Flag-OX40L, antiFAP-Flag-41BBL and antiFAP-Flag-TNC-CD27L binding to OX40, 41BB and CD27 as well as to FAP was detectable. These fusion proteins were able to induce IL8 efficiently via their receptor only as artificial membrane-bound molecules after binding to FAP. In summary, it has been shown that poorly or not active soluble ligand-variants of OX40L and 41BBL can be turned into highly active ligands by secondary oligomerization, by the hexamerizing Fc-domain and by artificial immobilization on the cell surface. Soluble CD27L variants additionally need the trimer-stabilizing TNC-domain to bind and activate CD27. For better understanding of ligand-receptor-interactions additionally OX40L-, 41BBL- and CD27L-fusion proteins with the highly active Gaussia princeps luciferase (GpL) were generated to enable equilibrium binding and dissociation studies as well as homologous competition assays. It was shown here with the fusion proteins GpL-Flag-TNC-OX40L, GpLFlag-TNC-41BBL and GpL-Flag-TNC-CD27L that induction of IL8 is independent of receptor occupancy, but depends on secondary oligomerization. Secondary cross-linked TNF ligands induced more IL8 with equal receptor occupancy. Therefore a qualitative better receptor activation has to be assumed.

Thesis advisor: Shih-Yuan Liu / Described herein are three projects that derive from in-depth studies of the distinct electronic structure of monocyclic 1,2-dihydro-1,2-azaborine (heretofore referred to as simply 1,2-azaborine). In the first chapter, the first comprehensive review of the late-stage functionalization methods available for 1,2-azaborines as well as their bicyclic and polycyclic (BN-PAH) counterparts is presented. In the second chapter, the development of a general method for both C4 and C5 functionalization based on the building block approach is described. The distinct electronic structure of 1,2-azaborine enables the chemical separation and further functionalization of C4 and C5 borylated isomers. In the second part, the C4, C5, and C6 isomers of BN-styrene analogues were prepared using the newly developed azaborine building blocks. The corresponding polymers were synthesized and extensively characterized in order to compare the effects of the BN-bond positioning relative to the polymer chain. In the fourth and final chapter, 1,2-azaborine-containing phosphine ligands featuring a P-B bond are synthesized. A comparative electronic structure analysis is performed between the BN-phosphine ligands and their direct all-carbon counterparts. / Thesis (PhD) — Boston College, 2019. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Chemistry.

The inhibitory glycine receptors are one of the major mediators of rapid synaptic inhibition in the mammalian brainstem, spinal cord and higher brain centres. They are ligand-gated ion channels that are mainly involved in the regulation of motor functions. Dysfunction of the receptor is associated with motor disorders such as hypereklepxia or some forms of spasticity. GlyR is composed of two glycosylated integral membrane proteins α and β and a peripheral membrane protein of gephyrin. Moreover, there are four known isoforms of the α-subunit (α1-4) of GlyR while there is a single β-subunit. Glycine receptors can be homomeric including α subunits only or heteromeric containing both α and β subunits. To date, strychnine is the ligand that has the highest affinity as glycine receptor ligand. It acts as a competitive antagonist of glycine that results in the inhibition of Cl- ions permeation and consequently reducing GlyR-mediated inhibition. For a long time, the details of the molecular mechanism of GlyRs inactivation by strychnine were insufficient due to the lack of high-resolution structures of the receptor. Only homology models based on structures of other cys-loop receptors have been available. Recently, 3.0 Å X-ray structure of the human glycine receptor- α3 homopentamer in complex with strychnine, as well as electro cryo-microscopy structures of the zebra fish α1 GlyR in complex with strychnine and glycine were published. Such information provided detailed insight into the molecular recognition of agonists and antagonists and mechanisms of GlyR activation and inactivation. Very recently, a series of dimeric strychnine analogs obtained by diamide formation of two molecules of 2-aminostrychnine with diacids of different chain length was pharmacologically evaluated at human α1 and α1β glycine receptors. None of the dimeric analogs was superior to strychnine. The present work focused on the extension of the structure-activity relationships of strychnine derivatives at glycine receptors All the synthesized compounds were pharmacologically evaluated at human α1 and α1β glycine receptors in a functional FLIPRTM assay and the most potent analogs were pharmacologically evaluated in a whole cell patch-clamp assay and in [3H]strychnine binding studies. It was reported that 11-(E)-isonitrosostrychnine displayed a 2-times increased binding to both α1 and α1β glycine receptors which prompted us to choose the hydroxyl group as a suitable attachment point to connect two 11-(E)-isonitrosostrychnine molecules using a spacer. In order to explore the GlyR pocket tolerance for oxime extension, a series of oxime ethers with different spacer lengths and sterical/lipophilic properties were synthesized biologically evaluated. Among all the oxime ethers, methyl, allyl and propagyl oxime ethers were the most potent antagonists displaying IC50 values similar to that of strychnine. These findings indicated that strychnine binding site at GlyRs comprises an additional small lipophilic pocket located in close proximity to C11 of strychnine and the groups best accommodated in this pocket are (E)-allyl and (E)-propagyl oxime ethers. Moreover, 11-aminostrychnine, and the corresponding propionamide were prepared and pharmacologically evaluated to examine the amide function at C11 as potential linker. A series of dimeric strychnine analogs designed by linking two strychnine molecules through amino groups in position 11 with diacids were synthesized and tested in binding studies and functional assays at human α1 and α1β glycine receptors. The synthesized bivalent ligands were designed to bind simultaneously to two α-subunits of the pentameric glycine receptors causing a possibly stronger inhibition than the monomeric strychnine. However, all the bivalent derivatives showed no significant difference in potency compared to strychnine. When comparing the reference monomeric propionamide containing ethylene spacer to the dimeric ligand containing butylene spacer, a 3-fold increase in potency was observed. Since the dimer containing (CH2)10 spacer length was found to be equipotent to strychnine, it is assumed that one molecule of strychnine binds to the receptor and the ‘additional’ strychnine molecule in the dimer probably protrudes from the orthosteric binding sites of the receptor. / Die inhibitorischen Glycin-Rezeptoren (GlyR) gehören zu den wichtigsten Mediatoren der schnellen synaptischen Hemmung im Säugetierhirnstamm, Rückenmark und in höheren Gehirnzentren. Sie sind ligandgesteuerte Ionenkanäle, die hauptsächlich an der Regulation der motorischen Funktionen beteiligt sind. Dysfunktion des Rezeptors ist assoziiert mit motorischen Störungen wie Hyperekplexie und einigen Formen von Spastizität. GlyR sind Proteinkomplexe, die aus zwei glykosylierten integralen Membranproteinen α und β und dem peripheren Membranprotein Gephyrin bestehen. Von der α-Untereinheit sind vier Isoformen bekannt (α1-4), von der β-Untereinheit nur eine. GlyR können homomer (nur α-Untereinheiten) oder heteromer (α und ß-Untereinheiten) sein. Das Alkaloid Strychnin weist eine sehr hohe Affinität zu den GlyR auf. Es wirkt als kompetitiver Antagonist von Glycin und führt nach Bindung zu einer Hemmung des Chlorid-Ionen-Einstroms und folglich zu einer Verringerung der GlyR-vermittelten Inhibition. Lange Zeit waren die genauen Details des molekularen Mechanismus der GlyR-Inaktivierung durch Strychnin aufgrund des Fehlens von hochauflösenden Röntgenstrukturen des Rezeptors nicht bekannt; es standen nur Homologie-Modelle basierend auf Strukturen anderer cys-Loop-Rezeptoren zur Verfügung. Vor kurzem wurden eine 3.0-Å-Röntgenstruktur des humanen GlyR (α3-Homopentamer) im Komplex mit Strychnin sowie eine Kryoelektronenmikroskopie-Struktur des Zebrafisches (α1-GlyR im Komplex mit Strychnin und Glycin) veröffentlicht. Dadurch erhielt man detailliertere Informationen über die molekulare Erkennung von Agonisten und Antagonisten sowie den Mechanismen der Aktivierung und Inaktivierung von GlyR. Kürzlich wurde eine Reihe von dimeren Strychnin-Analoga, bei denen jeweils zwei Moleküle 2-Aminostrychnin durch Reaktion mit Disäuren unterschiedlicher Kettenlänge zu den entsprechenden Diamiden miteinander verknüpft wurden, pharmakologisch an humanen α1- und α1β-GlyR untersucht. Keines der dimeren Analoga war Strychnin überlegen. Die vorliegende Arbeit konzentriert sich auf der Erweiterung der Struktur-Wirkungs-Beziehungen von Strychnin-Derivaten bzgl. der Aktivität an Glycin-Rezeptoren. Die strukturellen Änderungen, die an Strychnin durchgeführt wurden, sind in Abbildung 27 dargestellt. ...

The present studies which have been performed in the work-group C-2 (Prof. W. Kiefer) within the program of the Sonderforschungsbereichs 347, deal with the FT-Raman and –IR spectroscopy on new organometallic complexes, synthesized in the work-groups B-2 (Prof. W. Malisch), B-3 (Prof. W. A. Schenk), D-1 (Prof. H. Werner) and D-4 (Prof. D. Stalke). The FT-Raman spectra recorded at 1064 nm led to very useful and interesting information. Furthermore, the DFT calculations which are known to offer promise of obtaining accurate vibrational wavenumbers, were successfully used for the assignment of the vibrational spectra. For the first time it has been possible to ascribe exactly the n(RhC) stretching mode in the vinylidene rhodium(I) complex trans-[RhF(=C=CH2)(PiPr3)2] by using isotopic substitution, in conjunction with theoretical calculations. This is also true for the complexes trans-[RhF(CO)(PiPr3)2], trans-[RhF(C2H4)(PiPr3)2], trans-[RhX(=C=CHPh)(PiPr3)2] (X = F, Cl, Br, I, Me, PhCºC) and trans-[RhX(CN-2,6-xylyl)(PiPr3)2] (X = F, Cl, Br, I, CºCPh). In addition, the comparison between the n(RhC) wavenumbers of the complexes trans-[RhF(=13C=13CH2)(PiPr3)2] and trans-[RhF(CO)(PiPr3)2], containing the isoelectronic ligands 13C=13CH2 and CO, which have the same reduced mass, indicated that the Rh-C bond is stronger in the carbonyl than in the vinylidene complex. Besides, the n(RhF) stretching mode, which has been observed at higher wavenumbers in the FT-Raman and -IR spectra of trans-[RhF(CO)(PiPr3)2], showed that the carbonyl ligand is a better p-acceptor and a less effective s-donor than the vinylidene one. Moreover, the comparison of the n(CºC) and n(Rh-C) modes from the FT-Raman spectrum of the complexes trans-[Rh(CºCPh)(L)(PiPr3)2] (L = C=CHPh, CO, CN-2,6-xylyl) point out that the p-acceptor ability of the ligand trans to CºCPh should rise in the order C=CH2 < CO < CN-2,6-xylyl £ C=CHPh. The investigated sensitivity of the n(RhC), n(CC), n(CO) and n(CN) vibrational modes to the electronic modifications occuring in the vinylidene, carbonyl, ethylene and isonitrile complexes, should allow in the future the examination of the p-acceptor or p-donor properties of further ligands. Likewise, we were able to characterize the influence of various X ligands on the RhC bond by using the n(RhC) stretching mode as a probe for the weakening of this. The calculated wavenumbers of the n(RhC) for the vinylidene complexes trans-[RhX(=C=CHR)(PiPr3)2], where R = H or Ph, suggested that the strength of the Rh=C bond increases along the sequence X = CºCPh < CH3 < I < Br < Cl < F. For the series of carbonyl compounds trans-[RhX(CO)(PiPr3)2], where X = F, Cl, Br and I, analogous results have been obtained and confirmed from the model compounds trans-[RhX(CO)(PMe3)2]. Since, the calculated vibrational modes for the ethylene complex trans-[RhF(C2H4)(PiPr3)2] were in good agreement with the experimental results and supported the description of this complex as a metallacyclopropane, we were interested in getting more information upon this class of compounds. In this context, we have recorded the FT-Raman and -IR spectra of the thioaldehyde complexes mer-[W(CO)3(dmpe)(h2-S=CH2)] and mer-[W(CO)3(dmpe)(h2-S=CD2)] which have been synthezised by B-3. The positions of the different WL vibrational modes anticipated by the DFT calculations, were consistent with the experimental results. Indeed, the analysis of the band shifts in the FT-Raman and –IR spectra of the isotopomer mer-[W(CO)3(dmpe)(h2-S=CD2)] confirmed our assignment. The different stereoisomers of complex mer-[W(CO)3(dmpe)(h2-S=CH2)] were investigated too, since RMN and IR-data have shown that complex mer-[W(CO)3(dmpe)(h2-S=CH2)] lead in solution to an equilibrium. Since the information on the vibrational spectra of the molybdenum and tungsten complexes Cp(CO)2M-PR2-X (M = Mo, W; R = Me, tBu, Ph; X = S, Se) is very scarce, we extended our research work to this class of compounds. We have tried to elucidate the bonding properties in these chalcogenoheterocycle complexes by taking advantage of the mass effect on the different metal atoms (W vs. Mo). Thus, the observed band shifts allowed to assign most of the ML fundamental modes of these complexes. This project and the following one were a cooperation within the work-group B-2. The Raman and IR spectra of the matrix isolated photoproducts expected by the UV irradiation of the iron silyl complex Cp(CO)2FeSiH2CH3 have been already reported by Claudia Fickert and Volker Nagel in their PhD-thesis. Since no exact assignment was feasible for these spectra, we were interested in the study of the reaction products created by irradiation of the carbonyl iron silyl complex Cp(CO)2FeCH2SiH3. Although the calculated characteristic vibrational modes of the metal ligand unit for the various photoproducts are significantly different in constitution, they are very similar in wavenumbers, which did not simplify their identification. However, the theoretical results have been found to be consistent with the earlier experimental results. Finally, the last part of this thesis has been devoted to the (2-Py)2E- anions which exhibit a high selectivity toward metal-coordination. All di(2-pyridyl) amides and -phosphides which were synthesized by D-4, coordinate the R2Al+ fragment via both ring nitrogen atoms. This already suggests that the charge density in the anions is coupled into the rings and accumulated at the ring nitrogen atoms, but the Lewis basicity of the central nitrogen atom in Et2Al(2-Py)2N is still high enough to coordinate a second equivalent AlEt3 to form the Lewis acid base adduct Et2Al(2-Py)2NAlEt3. Due to the higher electronegativity of the central nitrogen atom in Me2Al(2-Py)2N, Et2Al(2-Py)2N and Et2Al(2-Py)2NAlEt3, compared to the bridging two coordinated phosphorus atom in Me2Al(2-Py)2P and Et2Al(2-Py)2P, the di(2-pyridyl)amide is the hardest Lewis base. In the phosphides merely all charge density couples into the rings leaving the central phosphorus atom only attractive for soft metals. These results were confirmed by using DFT and MP2 calculations. Moreover, a similar behaviour has been observed and described for the benzothiazolyl complex [Me2Al{Py(Bth)P}], where complementary investigations are to be continued. The DFT calculations carried out on the model compounds analysed in these studies supply very accurate wavenumbers and molecular geometries, these being in excellent agreement with the experimental results obtained from the corresponding isolated complexes. / Die vorliegende Arbeit wurde im Rahmen des Sonderforschungsbereichs 347 „Selektive Reaktionen Metall-aktivierter Moleküle“ im Teilprojekt C-2 (Prof. W. Kiefer) „Laserspektroskopie zur Charakterisierung der Struktur und Dynamik Metall-gebundener Moleküle“ durchgeführt. Diese befaßt sich mit den Infrarot- und Raman-spektroskopischen Untersuchungen an Übergangsmetallverbindungen, die in den Teilprojekten B-2 (Prof. W. Malisch), B-3 (Prof. W. A. Schenk), D-1 (Prof. H. Werner), D-4 (Prof. D. Stalke) synthetisiert wurden. Durch den Einsatz der FT-Raman-Spektroskopie mit langwelliger Laseranregung im NIR-Bereich und zum Teil von isotopenmarkierten Molekülen konnten aussagekräftige Spektren erhalten werden. Die Dichtefunktionnaltheorie stellte sich als geeignetes Mittel zur Vorhersage und Interpretation der Schwingungsspektren heraus. Abhängig von der Größe der betrachteten Komplexe waren jeweils Rechnungen nötig, die auf sehr unterschiedlichen theoretischen Niveaus basierten. Zum ersten Mal wurde mit Hilfe der Isotopenmarkierung und der Dichtefunktionaltheorie die Valenzschwingung n(Rh=C) in trans-[RhF(L)(PiPr3)2] (L = C=CH2, 13C=13CH2) charakterisiert. Diese zeigte sich als eine starke Raman-Bande und konnte ebenfalls im trans-[RhF(CO)(PiPr3)2] identifiziert werden. Darüber hinhaus erkannte man beim Vergleich von trans-[RhF(13C=13CH2)(PiPr3)2] und trans-[RhF(CO)(PiPr3)2] eine Verschiebung nach höheren Wellenzahlen der Valenzschwingung n(RhC) für den Carbonyl-Komplex. Einerseits haben beide Liganden 13C=13CH2 und CO die gleiche reduzierte Masse, was die elektronische Natur der n(RhC)-Verschiebung zeigt, welche eine Verstärkung der RhC-Valenzkraftkonstanten im Fall des Carbonyls belegt. Anderseits weist die Verschiebung der n(RhF)-Streckschwingung nach höheren Wellenzahlen im Carbonyl-Komplex für deren Ligand bessere p-Akzeptor- und schlechtere s-Donor-Eigenschaften gegenüber dem Vinyliden auf. Durch die aus solchen Untersuchungen gewonnenen n(RhC)- und n(CºC)-Verschiebungen in den verschiedenen untersuchten Komplexen ergibt sich die folgende Reihe abnehmender p-Akzeptorstärke: C=CHPh ³ CN-2,6-xylyl > CO > C=CH2. Die Empfindlichkeit der Valenzschwingungen n(RhC), n(CC), n(CO) und n(CN) gegenüber Veränderungen der elektronischen Verhältnisse in Vinylidene-, Carbonyl-, Ethylene- und Isonitrile-Komplexen läßt sich ihrerseits als „Sonde“ zur Untersuchung der p-Akzeptor-bzw. p-Donor-Eigenschaften anderer Liganden nutzen. Die Beeinflussung, vor allem die Schwächung der RhC-Bindung durch einen trans-ständigen Liganden konnte dadurch an den Komplexen trans-[RhX(13C=13CH2)(PiPr3)2] (X = F, Cl, Br, I), trans-[RhX(C=CHPh)(PiPr3)2] (X = F, Cl, Br, I, Me, PhCºC), trans-[RhX(CO)(PiPr3)2] (X = F, Cl, Br, I, PhCºC) und trans-[RhX(CN-2,6-xylyl)(PiPr3)2] (X = F, Cl, Br, I, PhCºC) untersucht werden. Die FT-Raman Spektroskopie zeigte sich als eine nützliche Methode zur Untersuchung des Trans-Einflusses. MO- und NBO-Berechnungen waren dabei sehr hilfreich, um diesen Effekt zu charakterisieren. Eine weitere Substanzklasse der hier untersuchten Übergangsmetallverbindungen stellen die verschiedenen Molybden- und Wolframkomplexe dar, die in den Teilprojekten B-3 und B-2 synthetisiert wurden. In diesem Zusammenhang wurden die FT-Raman- und –IR-Spektren von den polykristallinen Thioaldehyd-Komplexen mer-[W(CO)3(dmpe)(h2-S=CR2)] (R = H, D) aufgenommen und mit Dichtefunktionalrechnungen verglichen. Die Isotopenmarkierung lieferte eine klare Zuordnung der n(WC) und n(CS) Valenzstreckschwingungen, welche den partialen CS-Doppelbindungscharakter in diesen Verbindungen zeigte. Zudem konnte eine vollständige Analyse dieser Komplexe mit Hilfe der DFT-Rechnungen erlangt werden. NMR- und IR-Daten zeigten, daß bei einer Lösung von mer-[W(CO)3(dmpe)(h2-S=CH2)] ein Gleichgewicht stattfindet. Infolgedessen wurden die Energien der unterschiedlichen Stereo-Isomere von mer-[W(CO)3(dmpe)(h2-S=CH2)] untersucht, welche in sehr guter Übereinstimmung mit dem experimentellen Befund standen. Die Umsetzung der Phosphenium-Komplexe Cp(CO)2M=PR2 (M = W, Mo; R = tBu, Ph) mit Schwefel oder Selen lieferte entsprechende stabile [2+1]-Cycloaddukte in guten bis sehr guten Ausbeuten. FT-Ramanspektren von solchen Verbindungen wurden auf der Basis von Dichtefunktionalrechnungen aufgenommen und diskutiert. Diese Untersuchungen fanden in Zusammenarbeit mit Teilprojekt B-2 statt und hatten die Aufklärung der Bindungseigenschaften des Dreirings in diesen Komplexen zum Ziel. Die wichtigsten n(M-L) Valenzschwingungen konnten ebenso charakterisiert werden. Bei der UV-Bestrahlung von Cp(CO)2FeSiH2Me und Cp(CO)2FeCH2SiH3 sind verschiedene Photoprodukte bzw. Intermediate zu erwarten. In den Dissertationen von Claudia Fickert und Volker Nagel sind Veränderungen an den Raman- bzw. IR-Spektren der UV-Bestrahlungexperimente der matrixisolierten Substanzen vorgestellt und diskutiert worden. Dabei wurde die a-H-Umlagerung nach photochemischer Decarbonylierung als stabilstes Intermediat postuliert. Jedoch konnten keine eindeutigen Aussagen getroffen werden. Aufgrund dessen wurde die theoretische photochemisch induzierte Decarbonylierung und anschließenden Umlagerungen von Cp(CO)2FeSiH2Me und Cp(CO)2FeCH2SiH3 mit Hilfe der Dichtefunktionaltheorie behandelt und in einem Kapitel der vorliegenden Dissertation dargestellt. Im letzten Teil dieser Arbeit wurden Raman-Spektren und quantenchemische Rechnungen an Di(2-Pyridyl)systemen durchgeführt, die im Teilprojekt D-4 synthetisiert wurden. Die Py2E--Anionen weisen eine außergewöhnliche Selektivität bezüglich der Metallkoordination auf. Um geeignete Vorläufermoleküle zur Darstellung dünner III/V-Schichten mittels MOCVD-Experimente darzustellen, wurde Py2NH und Py2PH mit Et3Al bzw. Me3Al umgesetzt. Ein deutlicher Unterschied zwischen Amid und Phosphid ist in der Reaktivität gegenüber einem weiteren Lewis-säuren Äquivalent Et3Al bzw. Me3Al zu erkennen. Das bivalente amidische Stickstoffatom ist im Gegensatz zum Phosphoratom zu einer weiteren Koordination befähigt, was mittels DFT- und MP2-Rechnungen belegt wurde. Der Py(Bth)P--Ligand in [Me2Al{Py(Bth)P}] kann als doppelter Hart/Weich-Chelatligand bezeichnet werden. Das Me2Al+-Fragment koordiniert über die „harte Seite“ des Liganden (den Pyridylstickstoffatomen), während die „weiche Seite“ als P-S-Chelatligand weiterhin in der Lage sein sollte, weiche Übergangsmetallkomplexfragmente [M] zu stabilisieren. Diese Verbindung wurde zum Teil mit den obengenannten Methoden charakterisiert und sollte in weiter untersucht werden. Die durchgeführten DFT-Rechnungen lieferten nicht nur eine Möglichkeit der Interpretation von Schwingungsspektren, sondern erlaubten auch den Vergleich berechneter Molekülgeometrien mit Daten von Kristallstrukturanalysen und lieferten wichtige Antworten zu verschiedenen Problemstellungen.

Abstract not available

"September 2002" Includes as appendix: a list of publications by the author arising from this work; and, copies of some published journal articles Includes bibliographical references. Describes the synthesis, properties and reactions of transition metal complexes containing poly-ynyl ligands

This thesis reports on the synthesis, characterisation, enantiomeric resolution, 1H NMR structural study and physical evaluation of a series of certain bidentate ligand metal complexes, where ‘L-L’ denotes the ancillary bidentate ligand and ‘intercalator’ indicates the intercalating bidentate ligand. The L-L series varies in size and shape. Results of many tests and projects conducted are explained in detail. / Master of Science (Hons)

The ortho metallation of some aromatic ring systems has been investigated in regard to the influence of several types of phosphorus-centred directing groups upon the reactivity, regioselectivity, and utility in later synthetic elaboration. The metallation step allows derivatisation in several useful ways, offering several routes to the synthesis of novel chiral ditertiary phosphines. Thus, an ortho lithiation of N,N,N',N'-tetramethyl-P-phenylphosphonic diamide (10) led to the interesting primary phosphine, 2-(diphenylphosphino)phenylphosphine (14), after elaboration of the phosphonic diamide directing group. This primary phosphine undergoes an unprecedented facile phenyl group exchange process between its two phosphorus atoms, upon di-lithiation of the primary phosphorus centre. The primary phosphorus centre of (14) has been elaborated in several ways to yield new ditertiary phosphines. The alkylation of this centre in the copper(I) chelate complex has been investigated in several directions. In another direction, (14) has been chemically elaborated to give a new hybrid chiral ditertiary phosphine ligand, &quotSemiPHOS&quot, containing both a chiral phospholane ring and an adjacent diphenylphosphino group. SemiPHOS has been obtained in optically pure forms by a stereoselective synthesis and, independently, by a resolution procedure on its racemate. The molecular design of SemiPHOS was devised such that, when chelated to a metal atom, a subtle steric interaction appears to allow the chirality of the phospholane ring to influence the neighbouring diphenylphosphino group to adopt a complementary chiral conformation. This idea was tested and evaluated by applying SemiPHOS in catalytic asymmetric hydrogenations of (Z)-a-(Nacylamino) acrylate substrates to produce the R-amino acid precursors. Aryl species lithiated ortho to phosphorus-centred directing groups were coupled oxidatively by a convenient in situ method, to yield biaryl species that could then be elaborated to give biaryl ditertiary phosphine ligands. This method was used to make several atropisomeric chiral ditertiary phosphines.

Six novel platinum(II) intercalators of the form [Pt(AL)(IL)]Cl2, where AL = ethylenediamine (en), 1R,2R-diaminocyclohexane (R,R-dach), or 1S,2S-diaminocyclohexane (S,S-dach) and IL = 4,7-dihydroxy-1,10-phenanthroline (4,7-dhp) or 4,7-dicarboxy-1,10-phenanthroline (4,7-dcp), were synthesised. All complexes were prepared by the addition of the intercalating ligand followed by the addition of the diamine ancillary ligand. The complexes with 4,7-dhp were soluble in DMSO and were characterised by 1H, 13C, and 195Pt NMR, elemental analysis, UV-vis, ESI-MS, and CD. The complexes with 4,7-dcp were only soluble in a highly acidic solution and, therefore, were characterised only by 1H NMR and elemental analysis. The cytotoxicity of the 4,7-dhp complexes was tested in the L1210 murine leukaemia cell line. [Pt(S,S-dach)(4,7-dhp)]Cl2 showed an IC50 value of > 80 μM. The antitumour and antibacterial activities of all six complexes were tested in vitro using the Kirby-Bauer disc diffusion method with Staphylococcus aureus and Agrobacterium tumefaciens. The 4,7-dhp complexes showed no activity to these bacteria strains. The activities of the 4,7-dcp complexes were not able to be tested due to their solubility only in acidic solutions, which itself inhibits cell growth. The diffusion coefficients of the Pt(II) intercalators of the form [Pt(AL)(IL)]Cl2, where AL = en, R,R-dach, or S,S-dach and IL = phen, 4-mp, 4,7-dmp, 4,7-dhp, 4,7-dcp or 3,4,7,8-tmp and various starting materials used during the synthesis of these complexes were measured using pulsed gradient spin-echo (PGSE) NMR. The diffusion coefficients of both 4,7-dcp and [Pt(4,7-dcp)Cl2] were observed to be lower than other compounds with similar molecular weights indicating dimerisation of the compounds. The binding studies of the systems, [Pt(en)(phen)]Cl2 to (i) BSA, (ii) delipidated BSA, and (iii) d(GTCGAC)2 were studied using a simple two-site binding model with diffusion NMR. The binding of [Pt(en)(phen)]Cl2 – BSA was well described by the model giving the values Kd = 0.0021 ± 0.0002 M and n = 5.85 ± 0.31. On the contrary, the binding of [Pt(en)(phen)]Cl2 – delipidated BSA showed a poor fit to the model. From the poor fit of the data, it was speculated that the transverse relaxation of BSA largely affected the system. The binding of [Pt(en)(phen)]Cl2 – d(GTCGAC)2 showed results where the diffusion coefficient decreases as the concentration of the drug increases but an opposite effect was observed from the point where the drug reached equimolar concentrations to d(GTCGAC)2. It was speculated that the drug undergoes allosteric binding to the biomolecule or that a conformational change occurred as the drug concentration increases in the system. A further study of [Pt(en)(phen)]Cl2 and K2PtCl4 using 195Pt diffusion NMR was conducted giving a diffusion coefficient of 3.08 ± 0.04 × 10-10 m2 s-1 for K2PtCl4. The diffusion coefficient of [Pt(en)(phen)]Cl2, however, were unobtainable due to the short transverse relaxation of the Pt complex. / Master of Science (M.Sc.) (Hons)

The dinuclear cyclometallated gold(I) complex [Au2(μ-2-C6F4PPh2)2] was prepared in high yield from the reaction of 2-LiC6F4PPh2 with either [AuBr(AsPh3)] or [AuCl(tht)], and from the reaction of 2-Me3SnC6F4PPh2 with [AuCl(tht)]. The digold(I) complex undergoes oxidative addition reactions with halogens to give the metal-metal bonded dihalodigold(II) complexes [Au2IIX2(μ-2-C6F4PPh2)2] (X = Cl, Br, I), which on warming or exposure to light, isomerise to give the heterovalent gold(I)-gold(III) species [XAu(µ-2-C6F4PPh2)(κ2-2-C6F4PPh2)AuX] containing a four-membered cyclometallated ring on a gold(III) centre. Unlike its protio analogue, [Au2(μ-2-C6F4PPh2)2] did not undergo oxidative addition of methyl iodide or dibenzoyl peroxide. The dihalodigold(II) [Au2IIX2(μ-2-C6F4PPh2)2] and gold(I)-gold(III) compounds [XAu(µ-2-C6F4PPh2)(κ2-2-C6F4PPh2)AuX] (X = Cl, Br) are further oxidised by halogens to give the digold(III) species [Au2X4(μ-2-C6F4PPh2)2] and [X3Au(μ-2-C6F4PPh2)(κ2-2-C6F4PPh2)AuX], respectively. The complexes [Au2X4(μ-2-C6F4PPh2)2] are reduced to the dihalodigold(II) complexes in the presence of one equivalent of zinc powder; further addition of zinc gave the parent digold(I) dimer. Treatment of [Au2IICl2(μ-2-C6F4PPh2)2] and [ClAu(µ-2-C6F4PPh2)(κ2-2-C6F4PPh2)AuCl] with an excess of silver nitrate, benzoate, acetate, trifluoroacetate or triflate gave the corresponding oxyanion complexes. Slow crystallisation of the di(benzoato)digold(II) complex from dichloromethane and methanol gave the parent digold(I) complex derived by reductive elimination. The di(triflato)digold(II) complex behaved similarly, although in this case the novel gold(I) tetramer [Au4(μ-2-C6F4PPh2)4] was formed together with the dimer. Two closely related gold complexes containing the chelating κ2(C,O) phosphine oxide ligand, 2-C6F4P(O)PPh2, were isolated from the reaction of [ClAu(µ-2-C6F4PPh2)(κ2-2-C6F4PPh2)AuCl] with an excess of silver nitrate. The reaction of [Au2IICl2(μ-2-C6F4PPh2)2] with two equivalents of potassium trifluoroethoxide failed to give the corresponding digold(II) bis(alkoxo) complex; instead, reduction took place to form the digold(I) dimer [Au2(μ-2-C6F4PPh2)2]. Treatment of a solution of the di(benzoato)digold(II) complex with C6F5Li gave the pentafluorophenyl complex [Au2(C6F5)2(μ-2-C6F4PPh2)2] which, when heated in toluene, rearranged to the gold(I)-gold(III) complex [(C6F5)Au(µ-2-C6F4PPh2)(κ2-2-C6F4PPh2)Au(C6F5)], analogous to the behaviour of the dihalodigold(II) complexes. The heterovalent, gold(I)-gold(III) dimethyl compound [Au2I,III(CH3)2(μ-2-C6F4PPh2)2] was obtained from the reaction of the di(benzoato)digold(II) complex with dimethylzinc. This compound is structurally similar to its tetraprotio analogue. The cycloaurated dinuclear gold complexes [Au2(μ-C6H3-n-F-2-PPh2)2] (n = 5, 6) were made similarly to the 2-C6F4PPh2 analogue from the appropriate lithium or tin reagents, though in some cases the dimers were formed in admixture with the corresponding gold(I) tetramers. Like their tetrafluoro analogues, the 6-fluoro complexes [Au2X2(μ-C6H3-6-F-2-PPh2)2] (X = Cl, Br, I) rearrange on heating to give the heterovalent gold(I)-gold(III) species [XAu(µ-C6H3-6-F-2-PPh2)(κ2-C6H3-6-F-2-PPh2)AuX]. Thus, the presence of a fluorine atom in place of hydrogen in the 6-position of the bridging aryl group is sufficient to stop the isomerisation of the digold(II) complexes [Au2X2(μ-2-C6H4PPh2)2] at the gold(I)-gold(III) stage and to prevent subsequent C-C coupling of the aryl groups at the gold(III) centre. In contrast, the dihalodigold(II) complexes containing the 5-fluoro substituted ligand undergo reductive elimination and coupling of the metallated aryl groups to give the digold(I) biphenyldiyl complexes [Au2X2(2,2'-Ph2P-5-FC6H3C6H3-5-F-PPh2)] (X = Cl, Br, I). The described complexes were characterised using 1H NMR, 31P NMR, 19F NMR spectroscopy, elemental analysis, mass spectroscopy, IR spectroscopy, X-ray diffraction and 197Au Mössbauer spectroscopy.