A valuable preparative synthetic method is characterized by a broad scope of substrates and high chemo-and regioselectivity in combination with an excellent yield. Many prominent transition-metal-catalyzed reactions, such as the Grubbs olefin metathesis [1] or the Sharpless epoxidation or bishydroxylation [2] fulfill these criteria. Recently we reported the cyclotrimerization of alkynes with simple cobalt-diimine complexes.[3] During this work we found that cobalt-diphosphine complexes, such as [Co-(dppe)Br 2 ], which we successfully applied in the cobaltcatalyzed Diels-Alder reaction of non-activated starting materials, did not efficiently catalyze the cyclotrimerization of terminal alkynes to form the aromatic product 1 (Scheme 1). The transformation of internal alkynes to the corresponding aromatic products could only be realized in acceptable yields at elevated temperatures and prolonged reaction times.Clearly, these results indicate that the starting materials coordinate to the cobalt center and that the desired cyclotrimerization proceeds; however, with internal alkynes the C À C coupling step is relatively slow. Therefore, under milder reaction conditions the coordination of alkynes should still be possible while the reaction rate for the cyclotrimerization should be further reduced. Additional substrates could then be added leading to alternative reaction pathways and new products. Thus, in addition to the internal alkynes, terminal alkenes were added to the catalyst. The outcome of this reaction was a formal intermolecular Alder-ene reaction [4] (Scheme 2) to afford the 1,4-diene 4. This reaction is mechanistically related to the transformations undertaken by Trost with the [CpRu] + (Cp = C 5 H 5 ) fragment.[5] Therefore, the cobalt-catalyzed reaction could also involve the coordination of the two starting materials in the coordination sphere of the cobalt center to form the cobaltacycle 2 (Scheme 2). A b-hydride elimination to 3 and a reductive elimination complete the stepwise formal Alder-ene reaction under very mild reaction conditions.Over the course of the investigation we realized that the transformation with a cobalt-dppm complex was very slow. Good results had already been obtained with a cobalt-dppe [6] complex, whereas the best results to date were with a cobaltdppp [6] catalyst system. When a longer carbon chain was used in the diphosphine ligand, such as dppb, [6] the activity of the cobalt complex was reduced considerably, thus for further investigations the dppp ligand was used.Some questions arise for this atom-economic interconnection of two simple starting materials: a) is the cobalt catalyzed Alder-ene reaction applicable to terminal alkynes? b) can unsymmetrical alkynes be transformed regioselectively? c) can the stereochemistry of the two newly formed double bonds in the 1,4-diene be controlled by the catalyst? and d) are a wide variety of functional groups accepted by the cobalt catalyst?The first question could be answered easily: terminal alkynes prefer the formation of the cycl...

The application of bidentate phosphine ligands in cobalt-catalyzed transformations of cyclic alkenes such as cyclopentene and cycloheptene with internal alkynes led to a chemoselective Alder-ene or a [2 + 2] cycloaddition reaction depending on the electronic nature of the alkyne and the bite angle of the ligand used.

A regio- and chemoselective cobalt(I)-catalysed 1,4-hydrovinylation reaction is the key step in the straightforward and convergent synthesis of moenocinol, the aglycone of moenomycin A.

Dedicated to Professor Reinhard W. Hoffmann on the occasion of his 75th birthday Carbocyclic four-membered rings can be synthesized by cycloaddition reactions through the photochemical conversion of alkenes into cyclobutanes or thermal [2+2] cycloaddition reactions of acceptor-substituted alkynes with alkenes to give cyclobutenes.[1] Transition-metal-catalyzed reactions of non-activated starting materials, which only undergo thermal cycloaddition under harsh conditions, are usually carried out with ruthenium [2] or rhodium [3] complexes, or alternatively with cobalt or nickel complexes. [4] A while ago we reported the first cobalt-catalyzed Alderene reaction between internal alkynes and terminal alkenes to give 1,4-dienes 1 (Scheme 1).[5] It is assumed that the lowvalent cobalt complex coordinates to both unsaturated starting materials, and that the reaction proceeds via a cobaltacycle to form the new C À C bond. Subsequent bhydride elimination and reductive elimination lead to the final product 1. [6] We report herein that the reaction of internal, strained alkenes, such as norbornene, with internal alkynes in the presence of a cobalt-diphosphine complex leads to cyclobutene derivatives in quantitative yield with high chemoselectivity (Scheme 2). The only other reported synthetically useful transition-metal-catalyzed [2+2] cycloaddition reactions with norbornene involve rhodium or ruthenium complexes. [2,3] The use of a very simple cobalt complex, [CoI 2 -(PPh 3 ) 2 ], in the presence of a further equivalent of the free ligand PPh 3 under reductive conditions (Zn powder) in [2+2] cycloaddition reactions was pioneered by Cheng and coworkers.[4] These early cobalt-catalyzed transformations are somewhat limited in that they require harsh conditions (toluene, 90 8C) and a 10-fold excess of the alkyne, and seem to be restricted to 7-oxa-bicyclo[2.2.0]heptenes as substrates. Furthermore, the catalyst system presented herein does not lead to the dimerization of norbornene, as reported by Cheng and co-workers, and the addition of a further equivalent of the phosphine ligand is not necessary.A further advantage of the Co(dppp) catalyst system is that, in contrast to most transition-metal-catalyzed [2+2]-cycloaddition reactions, no large excess of either starting material is required: The substrates can usually be used in a 1:1 ratio, which greatly simplifies the purification of the products.[7] Laborious column-chromatographic purification can be avoided: Simple filtration through a silica-gel plug is sufficient to remove the residual (in)organic catalyst components and provide analytically pure reaction products of type 2. The cobalt catalyst system described herein is also considerably less expensive than alternative rhodium-or ruthenium-based compounds, and the substrate scope is significantly wider than previously reported for cobalt-based systems.The results of the cycloaddition of norbornene derivatives in the presence of the Co(dppp) catalyst (2 or 10 mol %) are summarized in Table 1. The transformations involving...

Eine präparativ wertvolle Synthesemethode zeichnet sich durch ein breites Substratspektrum, eine hohe Chemo-und Regioselektivität sowie eine hohe Ausbeute aus. Viele prominente übergangsmetallkatalysierte Reaktionen, wie die Grubbs-Olefin-Metathese [1] oder die Sharpless-Epoxidierung oder -Bishydroxylierung, [2] erfüllen diese Kriterien. Vor einiger Zeit berichteten wir über die Cobalt-katalysierte Cyclotrimerisierung von Alkinen durch Cobalt-DiiminKomplexe.[3] Im Rahmen dieser Arbeiten stellten wir fest, dass Cobalt-Diphosphan-Komplexe wie [Co(dppe)Br 2 ], das wir für die Diels-Alder-Reaktion nichtaktivierter Reaktanten einsetzten, die Cyclotrimerisierung von terminalen Alkinen zu 1 nur unzureichend katalysieren (Schema 1). Die Umsetzungen interner Alkine zu den entsprechenden Arenen ergaben selbst bei erhöhten Temperaturen erst nach langen Reaktionszeiten akzeptable Ausbeuten.Die Reaktanten koordinieren offensichtlich an das Cobaltzentrum, und die gewünschte Cyclotrimerisierung läuft auch ab, aber die C-C-Verknüpfungsschritte sind mit internen Alkinen relativ langsam. Unter milderen Bedingungen sollten die Alkine zwar koordinieren, die Reaktionsgeschwindigkeit der Cyclotrimerisierung sollte aber so gering sein, dass mit anderen Substraten neue Reaktionen zu beobachten wären. Daher wurden zusätzlich zu den internen Alkinen auch terminale Alkene zum Katalysator hinzugegeben. Das Ergebnis dieser Umsetzungen war eine formale intermolekulare AlderEn-Reaktion [4] (Schema 2) zum 1,4-Dien 4, die den Umsetzungen nach Trost mit dem [CpRu] + -Fragment mechanistisch wohl sehr ähnlich ist.[5] Es ist daher anzunehmen, dass die beiden Reaktanten am Cobalt koordinieren und einen Cobaltacyclus 2 bilden (Schema 2). Unter b-Hydrid-Eliminierung zu 3 und reduktiver Eliminierung wird die schrittweise formale Alder-En-Reaktion dann unter milden Reaktionsbedingungen zum Abschluss geführt.Im Verlauf der Untersuchungen stellte sich heraus, dass die Umsetzungen mit dppm als Ligand nur schleppend ablaufen, doch mit dppe werden bereits recht gute Ergebnisse erzielt, die von dppp noch übertroffen werden. Bei Liganden mit einer längeren Kohlenstoffkette (z. B. dppb) sinkt die Aktivität der Cobaltkomplexe deutlich, sodass für die weiteren Umsetzungen der dppp-Ligand gewählt wurde. [6] Für diese atomökonomische Verknüpfung zweier einfacher Ausgangsverbindungen ergeben sich nun eine Reihe von Fragen: a) Gelingt die Cobalt-katalysierte Alder-En-Reaktion auch mit terminalen Alkinen? b) Lassen sich unsymmetrische Alkine regioselektiv umsetzen? c) Kann der Katalysator die Konfigurationen der beiden gebildeten Doppelbindungen im 1,4-Dien steuern? d) Akzeptiert der Cobaltkatalysator funktionelle Gruppen?Die erste Frage ließ sich schnell beantworten: Terminale Alkine reagierten bevorzugt zu den Cyclotrimerisierungsprodukten 1 (R 2 = H) und waren somit für die Cobalt-katalysierte Alder-En-Reaktion nicht geeignet. Andererseits führten die Umsetzungen von internen Alkinen naturgemäß zu höher substituierten 1,4-Dienen und zu interessanteren Produkten. D...