Whereas a number of homoleptic metal(III) tetramethylaluminates M(AlMe(4))(3) of the rare earth metals have proven accessible, the stability of these compounds varies strongly among the metals, with some even escaping preparation altogether. The differences in stability may seem puzzling given that this class of metals usually is considered to be relatively uniform with respect to properties. On the basis of quantum chemically obtained relative energies and atomic and molecular descriptors of homoleptic tris(tetramethylaluminate) and related compounds of rare earth metals, transition metals, p-block metals, and actinides, multivariate modeling has identified the importance of ionic metal-methylaluminate bonding and small steric repulsion between the methylaluminate ligands for obtaining stable homoleptic compounds. Low electronegativity and a sufficiently large ionic radius are thus essential properties for the central metal atom. Whereas scandium and many transition metals are too small and too electronegative for this task, all lanthanides and actinides covered in this study are predicted to give homoleptic compounds stable toward loss of trimethylaluminum, the expected main decomposition reaction. Three of the predicted lanthanide-based compounds Ln(AlMe(4))(3) (Ln = Ce, Tm, Yb) have been prepared and fully characterized in the present work, in addition to Ln(OCH(2)tBu)(3)(AlMe(3))(3) (Ln = Sc, Nd) and [Eu(AlEt(4))(2)](n). At ambient temperature, donor-free hexane solutions of Ln(AlMe(4))(3) of the Ln(3+)/Ln(2+) redox-active metal centers display enhanced reduction to [Ln(AlMe(4))(2)](n) with decreasing negative redox potential, in the order Eu ≫ Yb ≫ Sm. Whereas Eu(AlMe(4))(3) could not be identified, Yb(AlMe(4))(3) turned out to be isolable in low yield. All attempts to prepare the putative Sc(AlMe(4))(3), featuring the smallest rare earth metal center, failed.

Tebbe reagent [Cp 2 Ti{(m-CH 2 )(m-Cl)Al(CH 3 ) 2 }] (A; Cp= cyclopentadienyl) belongs to the most enigmatic organometallic compounds.[1] Its successful synthesis, resulting from the careful investigation of the reaction of [Cp 2 TiCl 2 ] with two equivalents Al(CH 3 ) 3 , was triggered by important discoveries in two fundamentally different areas of homogeneous catalysis. Indeed, the initial studies of methane (and methylidene) formation from [Cp 2 TiCl 2 ]/Al(CH 3 ) 3 mixtures were conducted in the context of Ziegler-Natta polymerization catalysis, [2] but the methylene unit was structurally characterized by X-ray crystallography for the first time in tantalum alkylidene complexes, such as [Cp 2 Ta(CH 2 )(CH 3 )], [3] and tungsten methylene compounds, related to proposed catalysts for olefin metathesis.[4] Although catalytically active in olefin metathesis, [5] the Tebbe reagent is currently used for efficient carbonyl methylenation reactions.[ [8] have been reported, there are no X-ray structures of the Tebbe reagent nor of discrete metallacycles of the type [M(m-CH 2 )(m-R)Al(CH 3 ) 2 ] (R = Cl, CH 3 ). [9,10] Previous studies from our laboratories on rare-earthmetal(III) tetramethylaluminate complexes [L x Ln{Al-(CH 3 ) 4 } y ] (y = 1, 2, 3; x + y = 3, L = monovalent ancillary ligand, Ln = lanthanides and Sc, Y, La) as polymerization catalysts [11,12] led to the isolation of Ln III clusters with methylene, [13,14] methine, [15] and carbide functionalities. [16] We also found that complex [Cp* 3 Y 3 (m-Cl) 3 (m 3 -Cl)(m 3 -CH 2 )-(thf) 3 ] (Cp* = C 5 (CH 3 ) 5 ) displayed Tebbe-like reactivity.[13]

Homoleptic tetramethylgallate Lu(GaMe 4 ) 3 reacts selectively with superbulky (Tp tBu,Me )H according to a methane elimination reaction, affording the quantitative formation of monomeric base-free low-coordinate (Tp tBu,Me )LuMe 2 . Weak interaction of the strongly basic lanthanide methyl groups with the comparatively weak Lewis acid trimethylgallium in Lu-(GaMe 4 ) 3 is indicated by their separation into [LuMe 3 ] n and GaMe 3 even at -35 °C.

Complexes [NNN]Ln(AlMe(4))(2) (Ln = Y, La, Nd, Lu) bearing the sterically demanding aryl-substituted triazenido ligand [(Tph)(2)N(3)] (Tph = [2-(2,4,6-iPr(3)C(6)H(2))C(6)H(4)]) can be obtained from homoleptic complexes Ln(AlMe(4))(3) in moderate yields, both via protonolysis with [(Tph)(2)N(3)]H and a salt metathesis reaction pathway utilizing [(Tph)(2)N(3)]K. In the solid state the Y and Lu derivatives are isostructural, with both tetramethylaluminate groups coordinated in an eta(2) fashion, while one of the [AlMe(4)] ligands of the Nd derivative features a distorted eta(2) coordination mode. Due to the high affinity of the triazenido ligand toward the more Lewis-acidic and harder aluminium cation compared to the softer rare-earth metal centres, ligand redistribution is observed in solution and formation of byproduct [(Tph)(2)N(3)]AlMe(2) is prominent. While the monoanionic triazenido ligand coordinates the rare-earth metal centres in an asymmetrical syn/anti fashion, it adopts an almost symmetric syn/syn configuration in the aluminium complex. Attempts were also made to produce putative dimethyl complexes {[(Tph)(2)N(3)]LnMe(2)} (Ln = Y, Lu) via cleavage of the aluminate moieties with diethyl ether. Furthermore, the intrinsic redistribution reactions are proposed to affect the performance of complexes [(Tph)(2)N(3)]Ln(AlMe(4))(2) in isoprene polymerization.

Quinolyl-substituted half-sandwich complexes (Cp Q )Ln(AlMe 4 ) 2 (Cp Q = 2,3,4,5-tetramethyl-1-(8quinolyl)cyclopentadienyl; Ln = Y, La) were obtained in quantitative yield via protonolysis reactions utilizing HCp Q and homoleptic tetramethylaluminates Ln(AlMe 4 ) 3 . X-ray structure analyses revealed that the quinolyl-substituted cyclopentadienyl ligand coordinates to the rare-earth metal centers in an η 5 :η 1 fashion through the Cp ring carbon atoms and the N atom of the quinolyl substituent. The complexes (Cp Q )Ln(AlMe 4 ) 2 show good activity and high trans-1,4-stereoselectivity (maximum 93%) in the polymerization of isoprene upon activation with the organoborates [PhThe effects of metal size, cocatalyst, temperature, and solvent were assessed in these polymerization reactions, and the performance of such N-donor-functionalized complexes was compared to that of (Cp R )Ln(AlMe 4 ) 2 containing different types of nondonor-functionalized cyclopentadienyl ligands.