DALTON PERSPECTIVE J. Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 1991 C2 Building blocks in the co-ordination sphere of electron-poor transition metals. Aspects of the chemistry of early-transition-metal carbenoide complexes † Rüdiger Beckhaus * Department of Inorganic Chemistry, Technical University Aachen, D-52056 Aachen, Germany Our understanding of the chemistry of the transition-metal–carbon s bond is improved by investigations of the chemistry of alken-1-yl complexes of electron-poor transition metals.There is no other system known in which we can easily switch between the possible reaction pathways, depending on the nature of the metal, the ligands L and the alken-1-yl group. Only reductive elimination, a- and b-H elimination reactions give high selectivity. a-Hydrogen elimination from Cp*2Ti(CH]] CH2)R (Cp* = h-C5Me5) derivatives leads to the versatile titana–allene intermediate [Cp*2Ti]] C]] CH2] 8. A wide range of cycloaddition products of high thermal stability can be prepared using 8.In reactions of 8 with copper and gold complexes, heterodinuclear m-vinylidene compounds, Cp*2Ti(m-C]] CH2)(m-X)M9L, are formed. Additionally the first examples of intermolecular carbene–carbene coupling reaction of a Fischer- and a Schrock-carbene ligand are reported by using the strong nucleophilic vinylidene fragment 8. The possibility of stabilising highly reactive intermediates as well as short-lived molecules is one of the great advantages of organometallic chemistry, located on the borderline between organic and inorganic chemistry.1,2 Well known examples are illustrated in Scheme 1.The formal co-ordination of carbenes to metal fragments, was first achieved in Fischer- and Schrocktype complexes,3 but more recently, stable carbenes in the form of imidazol-2-ylidene ligands,4,5 have been co-ordinated to transition metals to give well characterised complexes.6 Cyclobutadiene, vinylidene and aryne molecules can be well stabilised in the co-ordination sphere of transition metals, leading to compounds which can be handled under acceptable conditions.This knowledge of structure and bonding relationships has developed our understanding of organometallic chemistry and led to useful applications of organometallic reagents in organic synthesis and catalysis. In many cases the actual reactive species, which participates in further reactions, must be generated in the first reaction step from the corresponding organometallic sources.In connection with this, I want to draw the attention of the reader to such complexes of titaniumgroup metals, which are characterised by the primary formation of an intermediate exhibiting a titanium–carbon double bond.7 Especially starting from alken-1-yl transition-metal complexes the general reactivity of a transition-metal–carbon s bond can be well understood.8 The generation of an M]C double bond in a primary reaction step can occur from quite different types of starting molecules. Some schematic drawings are given in Fig. 1. Complexes which act as primary sources of carbene complexes or intermediates, will henceforth be referred to as ‘carbenoide transition metal’ complexes. Different types of titanium-group metal complexes characterised by primary carbene complex formation are known.9 Sometimes, such carbene complexes exist in the form of isolable molecules 10–13 or occasionally in the form of intermediates.7,14–16 Reactants which are characterised by primary titanium carbene complex formation are the Tebbe reagent 1,9 the Takai reagent 2,17,18 metallacyclobutanes 3 19–21 or complexes exhibiting Schrock-type reactivity (4).22 For the * E-Mail: r.beckhaus@ac.rwth-aachen.de † Dedicated to Professor Wolfgang Beck on the occasion of his 65th birthday. generation of Ti]] C intermediates, Me2AlCl must be removed by bases (pyridine) if 1 is used as the starting material,19 additional reducing agents must react with 2,17 olefins must be thermally liberated when 3 is used,19,23 and with 4, a-H elimination of hydrocarbons must occur.24–26 These reactions can be formally defined as 1,2-eliminations.Other sources of Ti]] C species exist, including diazo compounds,27 small cyclic olefins 28 and 1,1-dilithio compounds.29 On the other hand, carbenoide complexes of main-group metals, as investigated by Boche and co-workers (5 and 6), are characterised by the possibility of 1,1-elimination reactions and carbene intermediate formation thereof.30–33 Depending on the nature of the metal, the leaving group (LG), and the reaction conditions, electrophilic as well as nucleohilic properties of the carbon centre are observed.30,31 For several years we have been interested in the chemistry of early-transition-metal complexes, exhibiting M]C double bonds in a cumulative unit.During the experimental work on my habilitation thesis we observed the easy transformation of vinyl complexes 7 via (Cp* = h-C5Me5) vinylidene intermediates 8 and 9 to a metallacyclobutane 10.34,35 This reaction is characterised by the selective transfer of a proton from one vinyl group to the other. In contrast to the well developed chemistry of metalla–allenes of late transition metals,36 the observation of the transformation 7Æ10 was the first discovery in the new research field of vinylidene chemistry of early transition metals.37,38 Petasis and Bzowej 22 reported the successful use of simple Cp2TiMe2 (Cp = h-C5H5) in carbonyl olefination reactions 11Æ12.It was the first instance of use of Ti]] C intermediates in organic synthesis.7 The greatest advantage of this method lies in the application of substituted and functionalised alkylidenes.39–41 A detailed study of the thermolysis of Cp*2- TiMe2 via a [Cp*2Ti]] CH2] intermediate has been published,42 but at that time, attempts to trap the carbene intermediate were unsuccessful.Alken-1-yl Ligands in the Co-ordination Sphere of Early Transition Metals By comparison of the thermal stability of the vinyl, phenyl and alkyl complexes of titanium-group metals, it becomes obvious that the vinyl derivative is the most reactive one.8 As well as a-1992 J. Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 Scheme 1 elimination processes, b- and reductive-elimination reactions are also observed. In each case, only one reaction pathway results in high selectivity.Generally, reductive elimination reactions are the most preferred reactions of organometallic complexes from the thermodynamic point of view.43 However, for alkyl derivatives, H-transformation reactions are kinetically preferred. In the Fig. 1 Organometallic sources of Ti]] C intermediates (carbenoide complexes 1–4) (1 Tebbe reagent; 2 Takai reagent; 3 metallacyclobutanes; 4 complexes of Schrock-type reactivity), and main-group metal carbenoides 5 and 6 bearing leaving groups and metal atoms on the same carbon atom case of alken-1-yl compounds of the titanium-group metals we find that there is a correlation between the rotational barriers of the alken-1-yl ligand around the M]C s bond and the observed reaction pathways (Fig. 3). If free rotation is possible, reductive elimination products become dominant. Owing to the orientation of the acceptor orbitals of the bent metallocene fragment Fig. 2 Schematic drawing of possible subsequent products obtained from vinyl complexes of titanium-group metals Fig. 3 Proposed transition states of reductive elimination A, a-H elimination B and b-H elimination C from Cp2M alkenyl derivatives (M = Ti, Zr or Hf )J. Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 1993 in the equatorial plane between the Cp ligands, electron transfer from the C]C double bond to the transition-metal centre in this 16 electron complex becomes possible if the vinyl group is orientated perpendicular to it (Fig. 3, A). This rotameric orientation leads to a suitable transition state for the reductive elimination, owing to the differences in partial charge of the a-carbon atoms, as calculated by ab initio methods. Generally, for d0 systems, concerted reductive elimination reactions are forbidden by symmetry, although reductive elimination might actually be possible, especially if charge-transfer processes are involved.44,45 On the other hand, if this rotation is hindered by using bulky ligands (Cp* instead of Cp,46 or substituted alken- 1-yl groups instead of the simple CH]] CH2 ligand 47) C]H bond activation reactions become dominant (Fig. 3, B and C). Whereas the a-CH bond is activated in titanium complexes, b-CH activation occurs in zirconium complexes.34 The preferred formation of dienes as the product of reductive elimination is exemplified by reaction of the tetrahalides of titanium, zirconium and hafnium with vinyllithium in a molar ratio of 1 : 4.In all cases, the formation of diene complexes 14 can be proved by using chelating phosphine ligands or by ligand substitution reactions, as in the case of titanium to give [Ti(bipy)3] 13 (bipy = 2,29-bipyridyl).48 Hydrogen elimination from carbon sp2 centres can only occur in a mononuclear manner if the a-CH bond is directed towards the leaving group (C]H inside conformation 16). This name results from the central, as opposed to the lateral (C]H outside) orientation of the acceptor orbitals at the metallocene fragment.However, compounds of type Cp*2Ti(CH]] CH2)X show the C]H outside conformation 15 in solid-state structures and in solution [X = F,49 OC(C6H11)CH2,50 CCPh51 or CH3 52 ]. This means that the C]H inside conformation 16 must be realised in the first step. The energy necessary for the rotation process in Cp*2Ti- (CH]] CH2)2, is found from MMX-force field calculations to be about 52.1 kJ mol21, compared to Cp2Ti(CH]] CH2)2 for which a value of 18.3 kJ mol21 is calculated.43 Therefore we can conclude that the rotation barrier is the main contribution to the activation energy of C]H elimination as determined by kinetic measurements for the process 17Æ8Æ18 [87.9(5) kJ mol21].53 Selective liberation of methane occurs from complex 17 in the temperature range 5–20 8C, forming (via 8) the dark green fulvene complex 18.53 In solution, only the C]H outside rotamer of 17 can be observed by NOE measurements.52 The alternative elimination of ethylene and formation of a [Cp*2Ti]] CH2] intermediate is not observed.The Titana–Allene Building Block [Cp*2Ti]] C]] CH2] The existence of the titana–allene 8 as a real intermediate, generated from 17 by methane elimination (5–20 8C)53 or from 10 by ethylene liberation (70–100 8C),37,38 can be proved by several trapping experiments (see later). Substrates such as ketones, alcohols, cumulenes and heterocumulenes do not react directly with 17 or 10.Generally the formation of 8 is the ratedetermining step in reactions of 10 and 17. Substitution products 21 or ring-opened derivatives 20 could not be detected in reactions of 10 and 17 with acidic substrates. However, strong acidic substrates, e.g. thiophenol lead to formation of products of type 20.49 Several attempts were made to stabilise the vinylidene intermediate 8 itself by using electron-donating ligands like phosphines, pyridines or by using the Jutzi ligand C5(CH3)4CH2CH2NMe2 instead of one Cp*.54 In all of these experiments C]H bond activation, forming the fulvene complex 18, is dominant.Owing to its electronic structure, the vinylidene intermediate 8 is very useful in cycloaddition reactions.37,43 By using isocyanides an azabutatriene complex 22 is formed in the first reaction step in a [2 1 1] cycloaddition.55,56 Owing to the high reactivity of the intermediate 22 subsequent reactions occur, forming the five-membered metallacycle 23, which exhibits a heteroradialene substructure.56 The metallacyclic four-membered ring compounds 24 are synthesised in high yields by [2 1 2] cycloaddition reactions.All1994 J. Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 these complexes were isolated as crystalline solids of high thermal stability, which allowed an extensive investigation of the reactivity of these molecules.57–63 Oxatitanacyclobutanes 64 and azatitanacyclobutenes 65–67 are discussed as intermediates in reactions of carbenoide titanium complexes with carbonyl compounds or nitriles, but could not be isolated because of the generally high electrophilicity of the metal centre.Spontaneous ring-opening reactions afford carbonyl olefination,19 or products of vinylimido intermediates.66,68 Metallaoxetanes, such as Cp2TiCH2CR2O, have been proposed as intermediates in various transition-metal catalysed oxygen-transfer reactions.64,69,70 Only a few metallaoxetanes, formed in the reaction of transition-metal carbenes with carbonyl compounds‡,71 or by the reaction of a terminal metal oxide fragment (Cp*2Ti]] O) Fig. 4 Comparison of the energies of products derived from Cl2Ti]] CH2 or Cl2Ti]] C]] CH2 and O]] CH2, results of ab initio calculations ‡ Known structurally characterised metallaoxetanes comprise Ta,71 Mo,72 Ti 63,57 and Cr.73 with an allene,74 have been characterised by X-ray diffraction methods. The striking characteristic of the metallacycles 24 is their high thermal stability compared to products derived from a titanium methylene intermediate.From ab initio calculations on Cl2Ti]] C]] CH2 model complexes and derivatives thereof it can be shown that the formation of the carbonyl olefination products from Cl2Ti]] C]] CH2 and H2C]] O is 101 kJ mol21 less exothermic than the model system Cl2Ti]] CH2 1 O]] CH2 (Fig. 4).43 Our attempts to obtain titanaoxetanes from 8 and ketones have so far failed.In all cases, the formation of enolates 25 is observed. Due to the preferred six-membered transition state 26 compared to the side-on geometry 27, the reaction leads in a stereo- and regio-selective manner to D1 and E-configured products.50 The crystal structure (R1 = H, R2 = C6H11) and reactivity of the enolates 25 show typical alkoxide character instead of nucleophilic properties on the b-C atom.50 A similar behaviour is observed for enolates derived from a Cp*2Ti]] CH2 intermediate.75 However, under less sterically crowded conditions, the formation of oxetanes 29 by reaction of 28 with ketones is proposed from the formation of the carbonyl olefination products 30.76 This reaction is very useful for the formation of substituted allenes.The higher electrophilicity of the Cp2Ti fragment compared to the permethylated Cp*2Ti in 8 prohibited the isolation or even the spectroscopic characterisation of 29. Remarkably, depending on the nature of the heteroatom, different subsequent reactions of the titanacycles 24 are observed.In the case of the titanacyclobutane 10, cycloreversion reactions are dominant, forming the vinylidene intermediate 8. In the case of the oxetanes 24a, metathesis reactions are observed in the mass spectrometer, whereas the bimetallic oxetanes 24c fragment to give the starting materials. These characteristics lead into the classification of classical and non-classical reaction behaviour of titanaoxetanes.Classical behaviour means that the formation of Ti]] O metathesis products dominates, whereasJ. Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 1995 the inverse cycloreversion from the oxetane to Ti]] C is observed in complexes exhibiting non-classical behaviour, see Scheme 2. Fast ring-opening reactions occur in the case of the azatitanacyclobutenes, leading to products of insertion into the Ti]C bond, which exhibit the exo-methylene group.60 This behaviour is explained by the orientation of the lone pair on the heteroatom in 10 towards the acceptor orbitals in the equatorial plane of the metallocene Cp*2Ti fragment.77 The best orbital overlap can be expected in the case of the azatitanacyclobutenes 24d, because the lone pair at the nitrogen atom and the lateral acceptor orbital of the titanium centre are orientated in the same plane, leading to fast ring opening. The imido intermediate 31 is formed and when a further nitrile molecule is added, the formation of 32 is observed.60 On the other hand, if the hybridisation of the nitrogen atom is changed, as in the azatitanacyclobutanes 24g, no electrocyclic ring-opening reaction is observed.In the molecule 24g the nitrogen lone pair is orientated perpendicular to the acceptor orbital on titanium.78 A similar orientation of donor and acceptor orbitals is expected also in the case of titanacyclobutenes, leading to substituent controlled reactions. For titanacyclobutenes of type 24f exhibiting large substituents (a-R = SiMe3, b-R = Ph), a long internal C]C single bond of 1.502(7) Å is found by Scheme 2 X-ray structural analysis, and the reactivity is characterised by cycloreversion processes at higher temperatures.On the other hand, with smaller substituents (R = R = Me) the internal C]C single bond is found to be shorter [1.434(4) Å] indicating a transition in the reactivity of the titanacyclobutene. Indeed, electrocyclic ring opening reactions are only observed for the non-substituted titanacyclobutene Cp*2TiCH]] CHC]] CH2 24f9 leading via 33 and 34 to the formation of trans-polyacetylene 35.51,79 The observed reactivity of the titanacycles 24 is in accordance with the general observation, that the reactivity of bent metallocene complexes is determined by the electron deficient character of the metal centre in combination with the orientation of the acceptor orbitals in the equatorial plane of the metallocene fragment.This can be additionally illustrated by the orientation of the substituents on nitrogen atoms in Cp*2- TiNR9R0 complexes. If there is no steric hindrance, the substituents on the nitrogen atom (R9 = R0 = H;80 R9 = H, R0 = Me81) are rotated out of the equatorial plane of the metallocene to maximise overlap between the nitrogen lone pair (located in the pz orbital) and the metal lowest unoccupied molecular orbital (LUMO). This effect is also found in the case of Cp*2Hf(H)(NHMe).82 If larger substituents are present, as in Cp*2TiN(CH3)(Ph) the methyl and the phenyl group are now located in the equatorial plane, and consequently a dp–pp interaction is no longer possible.83 If the heteroatoms are incorporated in a planar metallacyclic ring as in 24, such dp–pp interactions are the cause of the different reaction behaviour.In a similar manner, these reactivity patterns can also be observed in the insertion behaviour of the titanacycles 24 (Scheme 3). In the case of titanacyclobutane 1084 and cyclobutenes 51 24f, insertion of small molecules such as isocyanides is observed only into the Ti]C bond opposite the exo-methylene group.For1996 J. Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 Scheme 3 the titanacyclobutene 24d, a spontaneous ring enlargement by insertion of a further nitrile molecule is observed 24dÆ32, due to the activation of the Ti]C sp2 bond by the lone pair on the nitrogen centre.In a similar manner the oxetanes 24a and 24c show insertion reactions into this Ti]C bond forming mono- (38) or double-insertion products (39).85 On the other hand, azatitanacyclobutanes 24g are inert to ring enlargement reactions with isocyanides even at higher temperatures. In some cases, cycloaddition reactions involving a titana– allene intermediate are useful in organic synthesis. The preparation of allenylketenimines 40 becomes possible in high yields via complexes of type 37 (Cp instead of Cp*) by extrusion of ‘Cp2Ti’.86 Additionally, the intramolecular cycloaddition of the titanocene vinylidene complex, formed by dechloroalumination of 41,87 with an alkene (or alkyne) affords bicyclic titanacyclobutanes and 42 (or butenes).88 This cyclisation has proven to be possible in all such complexes.The remaining carbon–metal bond in the metallacycles 42 offers great potential for further elaboration to give organic products. This is illustrated for complex 42 in reactions with N-bromosuccinimide (NBS), which gave the dibromide 45.The insertion reaction of isocyanide affords the imino complex 43 which gave the aldehyde 44 on acidic work-up; alternatively, reaction with carbon monoxide followed by acidic work-up gave the ketone 47, presumably via the ene–diolate complex 46.88 Regioisomers We have demonstrated that the formation of the titanacyclobutanes and butenes 24 occur in a regioselective manner. Big differences in the partial charge of the X, Y atoms, determined e.g.by 13C NMR measurements in the case of alkynes,51 lead to stereochemically pure compounds with the more negative carbon bonded to titanium. a- and b-Regioisomers are obtained by using substrates with small partial charge differences. This is the case for alkynes 51 or phosphaalkynes.60 As well as the electronically controlled reactions, leading mostly to kinetic products, sterically controlled reactions are also observed.Thus from a mixture of both cycloaddition products of 8 with the phosphaalkyne ButC]] ] P 24e and 48, only the product exhibiting the bulky tert-butyl group in the b position (24e) crystallised from the solution. In the case of mixtures of regioisomers obtained from unsymmetrical acetylenes, the more bulky group rearranged to the b position upon heating.51 The effect of electronic and steric control can be observed in the reaction of the titanocenevinylidene intermediate [Cp*2Ti]] C]] CH2] 8 with 1,3-diynes (see Fig. 5), yielding titanacyclobutene complexes (50).61 Only one regioisomer is formed, containing the acetylide group in the a position of the metallacyclic ring. The regioselectivity is in accordance with the polarities of the diynes and stereochemical conditions in the cyclobutene ring. This behaviour is in agreement with ab initio calculations and the results of molecular modelling. Using unsymmetric diynes, metallacyclobutenes exhibiting the larger substituent in the b position are formed, as shown for 50e.61 The only primary regioisomer formed by cycloaddition of 8 with isothiocyanates is 24b, which contains the sulfur atom in the a position of the metallacyclic ring.When heated in the presence of pyridine, 24b can be isomerised into the other possible regioisomer 51;58 C]] N cycloaddition products are not observed. The quantitative isomerisation 24bÆ51 shows that the lower polarity of the C]] S unit in the RNCS molecule allows the formation of a second isomer.In this respect, the behaviour of titanathietanes is quite different from the behaviour of titanaoxetanes, which normally react to give Ti]] O fragments (classical behaviour). In cycloaddition reactions of 8 with CS2, only the regioisomer with the sulfur atom in the bring position is observed.89 The inverse regiochemistry results from the difference in reactivity of a carbonyl and a thiocarbonyl group with strong carbanionic molecules and shows the thiophilic character of the nucleophilic carbene complex 8.90–92 Structure Isomerisation, Vinylidene–Acetylene Rearrangement The stabilisation of the vinylidene group H2C]] C:, tautomeric to acetylene, has enabled the investigation of vinylidene complexes of late transition metals.The 1,2-proton shift is a characteristic feature in the synthesis of vinylidene complexes formed from acetylenes.93–95 Depending on the relative stability of theJ. Chem.Soc., Dalton Trans., 1997, Pages 1991–2001 1997 vinylidene complex compared to the acetylene derivative, the formation of vinylidene complexes is often preferred by late transition metals (Scheme 4). On the other hand, a reverse proton shift, from a vinylidene to an acetylene intermediate, is observed on heating solutions of the bimetallic oxetanes 24c, generating the oxatitanacyclopentene complexes 52.38,57,59 This transformation corresponds to the behaviour of a ‘free’ vinylidene C]] CH2 molecule, in the gas phase.The five-membered dinuclear complexes 52 are crystalline materials of high thermal stability (up to 220 8C). In contrast to the vinylidene–acetylene rearrangement observed for the bimetallic ‘non-classical’ oxetanes 24c, the ‘classical’ oxetanes 24a are not able to rearrange to fivemembered rings. That means that there must be a destabilising effect by the second transition metal, which can be explained by a cycloreversion process in the first step and formation of an h2- C]C bonded bridged vinylidene complex, which initiates the vinylidene–acetylene rearrangement.In this regard, the structure of heterodinuclear vinylidene bridged complexes is of general interest. Remarkably an unusual vinylidene–acetylene rearrangement is also observed for vinylidene bridged homodinuclear molybdenum complexes.96 Using intermediate 8, the formation of different vinylidene bridged complexes should be possible. These are the symmetrically bridged 1,1-dimetallaethylene and the unsymmetrically bridged structures, characterised by different types of dp–pp interaction with the Ti]] C or the C]] C double bond of 8 (Fig. 6).97 These structure types are in accordance with CO- and CS-bridged binuclear complexes. The s,s-bridging mode is generally dominant.36,98–102 Examples of side-on bridged vinylidene complexes are rare.96,97,103–109 The semi-bridging type complexes are known for CO,110,111 CS112 but for vinylidene to our knowledge only one example of a Mo]Ru complex is published.105 Indeed in reactions of 17 with copper and gold complexes, dinuclear vinylidene bridged complexes can be isolated as crystalline materials [m.p.(decomp.) 93 8C 53a, 180 8C 53b].113 In the case of 53a–53d, the 13C NMR data are indicative of the semi-bridging structural type, due to the low field shift of the bridging a-vinylidene carbon atom in the range d 330 (53a)–300 (53d). Intermolecular Carbene–Carbene Coupling By reacting the nucleophilic Schrock-type carbene intermediate 8 with other carbene complexes 54, the formation of the dinuclear 1,3- as well as 1,2-dimetallacyclobutanes 55 and 56 should be possible.Remarkably, in reactions of thermally generated 8 from 17 with the Fischer carbene complex (OC)5Cr]] C(OMe)Me, a new type of C]C coupling reaction is found. The products 59 and 60 could be separated by chromatography. The structure of 59 was Fig. 5 Schematic drawing of electronically determined regiochemsitry (left) and sterically controlled reactions (right)1998 J. Chem.Soc., Dalton Trans., 1997, Pages 1991–2001 determined by X-ray diffraction.114 In 59 the octahedral geometry at the chromium centre is realised by co-ordination of the methoxy ligand, whereas in 60 an additional carbon monoxide molecule is co-ordinated. Overall, C]C bond formation occurs between carbon atoms of a vinylidene, carbene and carbonyl ligand. The most remarkable feature of the structures of 59 and 60 is the fact that both metal atoms are bonded to different Scheme 4 Fig. 6 Different structure types of vinylidene bridged heterodinuclear complexes atoms than in the starting materials. The formation of 59 and 60 is in accordance with a carbene–carbene coupling via intermediate generation of the allene complex 58. This reaction step becomes clear considering the primary interaction of the nucleophilic and electrophilic carbene carbon atoms, supported by CO co-ordination to the oxophilic titanium as shown in 57.Titanium-centred cycloaddition of the allene molecule with the remaining Cr(CO)x fragment leads directly to 59 and, after fast CO addition, to 60. The described reaction represents to our knowledge the first example of an intermolecular coupling of inversely polarised carbene ligands 63Æ64, although intramolecular carbene–carbene coupling reactions 61Æ62 have been reported.115 This new method of achieving metal-centred coupling ofJ.Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 1999 several carbon atoms of different substrates can potentially be used in syntheses. With other carbon electrophiles further types of C]C coupling reactions are observed. The reaction of 17 with the aminocarbene complex 65 leads to the dinuclear titanium complex 66 after chromatographic work-up. The dimeric structure is confirmed by the mass spectrum and the crystal-structure analysis.116 The surprising feature of 66 is its dark blue colour.In contrast, the monomeric units of the type 24f and 50 are dark red compounds. As observed in the crystal structure of 66 there is a twist angle between the cyclobutene planes of nearly 508. Other carbon electrophiles are also able to undergo C]C coupling reactions at the nucleophilic carbon centre of 8. By treating allylpalladium chloride 67 with 8 in the presence of PMe3 the formation of the coupling product 68 can be observed.117 Closing Remarks and Outlook The high reactive selectivity of alken-1-yl ligands in the coordination sphere of electron-deficient metals has proved to be a powerful tool in the classification of organometallic reactivity.From this, we can learn what is easy and what is difficult for a transition-metal–carbon s-bond to do. The transformation of titanium vinyl complexes to titanium vinylidene intermediates under mild thermal conditions in particular has significantly improved the access to short-lived carbene complexes of early transition metals.This has widened the spectrum of preparative and catalytic applications of [Ti]] C] generating derivatives considerably. Different types of stable cycloaddition products of 8 are isolable, allowing investigation of structures and reactivity relationships. However, many questions still remain open. First, what have we learned from the easy proton elimination from vinyl groups (which does not occur in comparable alkyl ligands), and are useful Heliminations from other substrates in the co-ordination sphere of metallocene complexes also possible under similar conditions? As well as the discussed Ti]] C systems, Ti]] Si,118 M]] N119 (M = Ti 120,121 or Zr 120,122), Zr]] P,123,124 Zr]] O,125,126 Zr]] S,125,126 and even intermediates exhibiting Ti]] ] C bonds,127,128 are available.Secondly, what about axial chirality in the titana–allene building block? From the preferred Cs symmetric ground state and the calculated Ti]C rotation energy (134 kJ mol21) 37 for the titana–allene compound, axial chiral complexes must be available. Thirdly, which types of further cycloaddition products are available? What about the possibility of syntheses of new types of molecules, like radialenes derived from reactions of 8 and butatrienes or from molecules of type 23.What further types of molecules can be prepared by new C]C coupling reactions? Finally, are acetylene–vinylidene rearrangements possible in the case of electron-deficient transition-metal complexes? The fact that the polymerisation of acetylene seems to proceed via a vinylidene intermediate suggests that this could be the case, but up to now, there is no definitive proof of such a rearrangement.79,129 Many reactions discussed before seem to be sterically controlled. This means that other ligand systems must also be able to realise the role of the Cp* ligand.130,131 Let’s go and find them! Acknowledgements I wish to express my particular thanks to my active co-workers, Dr.Jürgen Oster, Dr. Javier Sang, Dipl.-Chem. Jürgen Heinrichs, Dipl.-Chem. Martin Wagner and Dipl.-Chem. Isabelle Strauß, who are working on the titana–allene project. I am indebted to all my colleagues for constructive discussions and suggestions, in particular Dr. Uwe Böhme (University Bergakademie Freiberg) for ab initio calculations. I wish to express my gratitude to Mr.Edward Pritchard (York), for checking the English version of the manuscript. I especially wish to thank the Institute of Inorganic Chemistry at Aachen Technical University, where I was a guest and have done research for the last five years, the research group for crystal-structure determination, and in particular to Dr. Trixie Wagner for solving the numerous crystal-structure determinations. I gratefully acknowledge that my research was supported by the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie.The BAYER AG Leverkusen and the DEGUSSA AG are also acknowledged for generous financial support. References 1 C. Elschenbroich and A. Salzer, Organometallchemie, B. G. Teubner, Stuttgart, 3rd edn., 1990. 2 R. H. Crabtree, The Organometallic Chemistry of the Transition Metals, Wiley, New York, 2nd edn., 1994. 3 K. H. Dötz, H. Fischer, P. Hofmann, F. R. Kreissl, U. Schubert and K. Weiss, Transition Metal Carbene Complexes, Verlag Chemie, Weinheim, 1983. 4 M. Regitz, Angew. Chem., 1996, 108, 791; Angew. Chem., Int. Ed. Engl., 1996, 35, 725. 5 M. Regitz, Angew. Chem., 1991, 103, 691; Angew. Chem., Int. Ed. Engl., 1991, 30, 674. 6 W. A. Herrmann, C. Köcher, L. J. Gooßen and G. R. J. Artus, Chem. Eur. J., 1996, 2, 1627. 7 R. Beckhaus, Angew. Chem., 1997, 109, 694; Angew. Chem., Int. Ed. Engl., 1997, 36, 687.2000 J. Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 8 R. Beckhaus, J. Oster, J. Sang, I.Strauß and M. Wagner, Synlett, 1997, 241. 9 D. L. Hughes, J. F. Payack, D. Cai, T. R. Verhoeven and P. J. Reider, Organometallics, 1996, 15, 663. 10 J. A. van Doorn, H. van der Heijden and A. G. Orpen, Organometallics, 1995, 14, 1278. 11 S. De Angelis, E. Solari, C. Floriani, A. Chiesi-Villa and C. Rizzoli, Angew. Chem., 1995, 107, 1200; Angew. Chem., Int. Ed. Engl., 1995, 34, 1092. 12 J. A. van Doorn, H. van der Heijden and A. G. Orpen, Organometallics, 1994, 13, 4271. 13 M.D. Fryzuk, S. S. H. Mao, M. J. Zaworotko and L. R. MacGillivray, J. Am. Chem. Soc., 1993, 115, 5336. 14 R. K. Minhas, L. Scoles, S. Wong and S. Gambarotta, Organometallics, 1996, 15, 1113. 15 H. van der Heijden and B. Hessen, J. Chem. Soc., Chem. Commun., 1995, 145. 16 L. Scoles, R. Minhas, R. Duchateau, J. Jubb and S. Gambarotta, Organometallics, 1994, 13, 4978. 17 K. Utimoto and K. Takai, NATO ASI Ser., Ser. C, 1989, 269, 379. 18 K. Takai, T. Kakiuchi, Y. Kataoka and K. Utimoto, J.Org. Chem., 1994, 59, 2668. 19 R. H. Grubbs and R. H. Pine, Comprehensive Organic Synthesis; Alkene Metathesis and Related Reactions, in Comprehensive Organic Synthesis, ed. B. M. Trost, Pergamon, New York, 1991, vol. 5, ch. 9.3, pp. 1115–1127. 20 R. H. Grubbs, Prog. Inorg. Chem., 1978, 24, 1. 21 S. H. Pine, Carbonyl Methylenation and Alkylidation using Titanium- Based Reagents, in Organic Reactions, ed. L. A. Paquette, Wiley, New York, 1993, vol. 43, pp. 1–91. 22 N.A. Petasis and E. I. Bzowej, J. Am. Chem. Soc., 1990, 112, 6392. 23 D. A. Straus and R. H. Grubbs, J. Mol. Catal., 1985, 28, 9. 24 R. R. Schrock, Acc. Chem. Res., 1979, 12, 98. 25 J. Feldman and R. R. Schrock, Prog. Inorg. Chem., 1991, 39, 1. 26 R. R. Schrock, Pure Appl. Chem., 1994, 66, 1447. 27 J. L. Polse, R. A. Andersen and R. G. Bergman, J. Am. Chem. Soc., 1996, 118, 8737. 28 P. Binger, P. Müller, R. Benn and R. Mynott, Angew. Chem., 1989, 101, 647; Angew. Chem., Int. Ed.Engl., 1989, 28, 610. 29 P. Binger, P. Müller, R. Wenz and R. Mynott, Angew. Chem., 1990, 102, 1070; Angew. Chem., Int. Ed. Engl., 1990, 29, 1037. 30 A. Maercker, Angew. Chem., 1993, 105, 1072; Angew. Chem., Int. Ed. Engl., 1993, 32, 1023. 31 A. Müller, M. Marsch, K. Harms, J. C. W. Lohrenz and G. Boche, Angew. Chem., 1996, 108, 1639. 32 G. Boche, M. Marsch, A. Müller and K. Harms, Angew. Chem., 1993, 105, 1081; Angew. Chem., Int. Ed. Engl., 1993, 32, 1032. 33 G. Boche, K.Harms, M. Marsch and A. Müller, J. Chem. Soc., Chem. Commun., 1994, 1393. 34 R. Beckhaus, K.-H. Thiele and D. Ströhl, J. Organomet. Chem., 1989, 369, 43. 35 R. Beckhaus, Habilitation Thesis, Technische Hochschule Leuna-Merseburg, 1989, RWTH Aachen, 1993. 36 M. I. Bruce, Chem. Rev., 1991, 91, 197. 37 R. Beckhaus, S. Flatau, S. I. Troyanov and P. Hofmann, Chem. Ber., 1992, 125, 291. 38 R. Beckhaus, Methylidentitanacyclobutane vs. Titanocene- Vinylidene-Versatile Building Blocks in Organic Synthesis via Organometallics, eds.D. Enders, H.-J. Gais and W. Keim, Vieweg Verlag, Braunschweig, 1993, pp. 131–149. 39 N. A. Petasis and I. Akritopoulou, Synlett, 1992, 665. 40 N. A. Petasis and E. I. Bzowej, Tetrahedron Lett., 1993, 34, 943. 41 N. A. Petasis, J. P. Staszewski and D.-K. Fu, Tetrahedron Lett., 1995, 36, 3619. 42 C. McDade, J. C. Green and J. E. Bercaw, Organometallics, 1982, 1, 1629. 43 R. Beckhaus and U. Böhme, unpublished work. 44 B. Akermark and A.Ljungqvist, J. Organomet. Chem., 1979, 182, 59. 45 J. M. Brown and N. A. Cooley, Chem. Rev., 1988, 88, 1031. 46 P. Courtot, V. Labed, R. Pichon and J. Y. Salaün, J. Organomet. Chem., 1989, 359, C9. 47 S. L. Buchwald and R. B. Nielsen, Chem. Rev., 1988, 88, 1047. 48 R. Beckhaus and K.-H. Thiele, J. Organomet. Chem., 1986, 317, 23. 49 R. Beckhaus, J. Sang, J. Oster and T. Wagner, J. Organomet. Chem., 1994, 484, 179. 50 R. Beckhaus, I. Strauß and T. Wagner, J. Organomet. Chem., 1994, 464, 155. 51 R. Beckhaus, J. Sang, T. Wagner and B. Ganter, Organometallics, 1996, 15, 1176. 52 J. Sang, Ph.D. Thesis, RWTH Aachen, 1996. 53 G. A. Luinstra and J. H. Teuben, Organometallics, 1992, 11, 1793. 54 R. Beckhaus, J. Oster and B. Ganter, unpublished work. 55 R. Beckhaus and J. Oster, Z. Anorg. Allg. Chem., 1995, 621, 359. 56 R. Beckhaus and I. Strauß, unpublished work. 57 R. Beckhaus, I. Strauß, T. Wagner and P. Kiprof, Angew. Chem., 1993, 105, 281; Angew. Chem., Int.Ed. Engl., 1993, 32, 264. 58 R. Beckhaus, J. Sang, T. Wagner and U. Böhme, J. Chem. Soc., Dalton Trans., in the press. 59 R. Beckhaus, J. Oster and T. Wagner, Chem. Ber., 1994, 127, 1003. 60 R. Beckhaus, I. Strauß and T. Wagner, Angew. Chem., 1995, 107, 738; Angew. Chem., Int. Ed. Engl., 1995, 34, 688. 61 R. Beckhaus, J. Sang, U. Englert and U. Böhme, Organometallics, 1996, 15, 4731. 62 M. Wagner, Diplomarbeit, RWTH Aachen, 1996. 63 R. Beckhaus, I. Straub and T. Wagner, Z.Anorg. Allg. Chem., 1997, 623, 654. 64 B. Schiott and K. A. Jorgensen, J. Chem. Soc., Dalton Trans., 1993, 337. 65 K. M. Doxsee, J. K. M. Mouser and J. B. Farahi, Synlett, 1992, 13. 66 K. M. Doxsee, J. B. Farahi and H. Hope, J. Am. Chem. Soc., 1991, 113, 8889. 67 K. M. Doxsee and J. B. Farahi, J. Am. Chem. Soc., 1988, 110, 7239. 68 K. M. Doxsee and J. B. Farahi, J. Chem. Soc., Chem. Commun., 1990, 1452. 69 K. A. Jorgensen and B. Schiott, Chem. Rev., 1990, 90, 1483. 70 K. A.Jorgensen, Chem. Rev., 1989, 89, 431. 71 L. L. Whinnery, jun., L. M. Henling and J. E. Bercaw, J. Am. Chem. Soc., 1991, 113, 7575. 72 G. C. Bazan, R. R. Schrock and M. B. O’Regan, Organometallics, 1991, 11, 1062. 73 J. Sundermeyer, K. Weber and H. Pritzkow, Angew. Chem., 1993, 105, 751. 74 D. J. Schwartz, M. R. Smith III and R. A. Andersen, Organometallics, 1996, 15, 1446. 75 C. P. Gibson and D. S. Bem, J. Organomet. Chem., 1991, 414, 23. 76 S. L. Buchwald and R. H. Grubbs, J.Am. Chem. Soc., 1983, 105, 5490. 77 J. W. Lauher and R. Hoffmann, J. Am. Chem. Soc., 1976, 98, 1729. 78 R. Beckhaus and M. Wagner, unpublished work. 79 A. Ohff, V. V. Burlakov and U. Rosenthal, J. Mol. Catal., 1996, 108, 119. 80 E. Brady, W. Lukens, J. Telford and G. Mitchell, Acta Crystallogr., Sect. C, 1994, 51, 558. 81 W. W. Lukens, jun., M. R. Smith III and R. A. Andersen, J. Am. Chem. Soc., 1996, 118, 1719. 82 G. L. Hillhouse, A. R. Bulls, B. D. Santarsiero and J. E.Bercaw, Organometallics, 1988, 7, 1309. 83 J. Feldman and J. C. Calabrese, J. Chem. Soc., Chem. Commun., 1991, 1042. 84 R. Beckhaus, C. Zimmermann, T. Wagner and E. Herdtweck, J. Organomet. Chem., 1993, 460, 181. 85 R. Beckhaus and I. Strauß, unpublished work. 86 R. D. Dennehy and R. J. Whitby, J. Chem. Soc., Chem. Commun., 1992, 35. 87 T. Yoshida and E. Negishi, J. Am. Chem. Soc., 1981, 103, 1276. 88 R. D. Dennehy and R. J. Whitby, J. Chem. Soc., Chem. Commun., 1990, 1060. 89 R.Beckhaus and J. Sang, unpublished work. 90 J. March, Advanced Organic Chemistry, Wiley, New York, 1985. 91 P. Beak and J. W. Worley, J. Am. Chem. Soc., 1972, 94, 597. 92 A. Ohno, K. Nakamura, Y. Shizume and S. Oka, Bull. Chem. Soc. Jpn., 1977, 50, 1003. 93 H. Werner, Nachr. Chem. Tech. Lab., 1992, 40, 435. 94 H. Werner, Angew. Chem., 1990, 102, 1109; Angew. Chem., Int. Ed. Engl., 1990, 29, 1077. 95 J. Silvestre and R. Hoffmann, Helv. Chim. Acta, 1985, 68, 1461. 96 S. F. T.Froom, M. Green, R. J. Mercer, K. R. Nagle, A. G. Orpen and R. A. Rodrigues, J. Chem. Soc., Dalton Trans., 1991, 3171. 97 H. El Amouri and M. Gruselle, Chem. Rev., 1996, 96, 1077. 98 M. I. Bruce and A. G. Swincer, Adv. Organomet. Chem., 1983, 22, 59. 99 H. Werner, J. Wolf, G. Müller and C. Krüger, J. Organomet. Chem., 1988, 342, 381. 100 J. E. Huheey, E. A. Keiter and R. L. Keiter, Anorganische Chemie– Prinzipien von Struktur und Reaktivität, Walter de Gruyter, Berlin, 2nd edn., 1995. 101 C. P. Horwitz and D. F. Shriver, Adv. Organomet. Chem., 1984, 23, 219. 102 P. V. Broadhurst, Polyhedron, 1985, 4, 1801. 103 U. Kern, C. G. Kreiter, S. Müller-Becker and W. Frank, J. Organomet. Chem., 1993, 444, C31.J. Chem. Soc., Dalton Trans., 1997, Pages 1991–2001 2001 104 M. Akita, S. Sugimoto, A. Takabuchi, M. Tanaka and Y. Moro-oka, Organometallics, 1993, 12, 2925. 105 R. Boese, M. A. Huffmann and K. P. C. Vollhardt, Angew. Chem., 1991, 103, 1542. 106 F. J. G.Alonso, V. Riera, M. A. Ruiz, A. Tripicchio and M. T. Camellini, Organometallics, 1992, 11, 370. 107 N. M. Doherty, C. Elschenbroich, H.-J. Kneuper and S. A. R. Knox, J. Chem. Soc., Chem. Commun., 1985, 170. 108 F. A. Cotton, Prog. Inorg. Chem., 1976, 21, 1. 109 R. A. Doyle, L. M. Daniels, R. J. Angelici and F. G. A. Stone, J. Am. Chem. Soc., 1989, 111, 4995. 110 R. H. Crabtree and M. Lavin, Inorg. Chem., 1986, 25, 805. 111 M. J. Winter, Adv. Organomet. Chem., 1989, 29, 101. 112 H. P. Kim, S. Kim, R. A. Jacobson and R. J. Angelici, J. Am. Chem. Soc., 1986, 108, 5154. 113 J. Oster, Ph.D. Thesis, RWTH Aachen, 1996. 114 R. Beckhaus, J. Oster, R. Kempe and A. Spannenberg, Angew. Chem., 1996, 108, 1636; Angew. Chem., Int. Ed. Engl., 1996, 35, 1565. 115 H. Kücükbay, B. Cetinkaya, S. Guesmi and P. H. Dixneuf, Organometallics, 1996, 15, 2434 and refs. therein. 116 R. Beckhaus and J. Oster, unpublished work. 117 R. Beckhaus and J. Oster, unpublished work. 118 E. Hengge and M. Weinberger, J. Organomet. Chem., 1993, 443, 167. 119 D. E. Wigley, Prog. Inorg. Chem., 1994, 42, 239. 120 C. P. Schaller, C. C. Cummins and P. T. Wolczanski, Organometallics, 1996, 118, 591. 121 C. C. Cummins, C. P. Schaller, G. D. Van Duyne, P. T. Wolczanski, A. W. E. Chan and R. Hoffmann, J. Am. Chem. Soc., 1991, 113, 2985. 122 P. J. Walsh, F. J. Hollander and R. G. Bergman, Organometallics, 1993, 12, 3705. 123 J. Ho, Z. Hou, R. J. Drake and D. W. Stephan, Organometallics, 1993, 12, 3145. 124 Z. Hou and D. W. Stephan, J. Am. Chem. Soc., 1992, 114, 10 088. 125 M. J. Carney, P. J. Walsh, F. J. Hollander and R. G. Bergman, Organometallics, 1992, 11, 761. 126 M. J. Carney, P. J. Walsh, F. J. Hollander and R. G. Bergman, J. Am. Chem. Soc., 1989, 111, 8751. 127 R. Andres, P. Gomez-Sal, E. de Jesus, A. Martin, M. Mena and C. Yelamos, Angew. Chem., 1997, 109, 72; Angew. Chem., Int. Ed. Engl., 1997, 36, 115. 128 R. Andres, M. V. Galakhov, A. Martin, M. Mena and C. Santamaria, Organometallics, 1994, 13, 2159. 129 H. G. Alt, H. E. Engelhardt, M. D. Rausch and L. B. Kool, J. Organomet. Chem., 1987, 329, 61. 130 See for example, R. R. Schrock, C. C. Cummins, T. Wilhelm, S. Lin, S. M. Reid, M. Kol and W. M. Davis, Organometallics, 1996, 15, 1470 and refs. therein. 131 J. S. Freundlich, R. R. Schrock and W. M. Davis, Organometallics, 1996, 15, 2777. Received 10th February 1997; Paper 7/00920H