Hexamethyltungsten was first reported in 1973 by Wilkinson and Shortland, who described its preparation by the reaction of methyllithium with tungsten hexachloride in diethyl ether. The synthesis was motivated in part by previous work which indicated that tetrahedral methyl transition metal compounds are thermally unstable, in the hopes that an octahedral methyl compound would prove to be more robust. In 1976, Wilkinson and Galyer disclosed an improved synthesis using trimethylaluminium in conjunction with trimethylamine, instead of methyllithium. The stoichiometry of the improved synthesis is as follows: Alternatively, the alkylation can employ dimethylzinc:
Molecular geometry
W6 adopts a distorted trigonal prismatic geometry with C3vsymmetry for the WC6 framework and C3 symmetry including the hydrogen atoms. The structure can be thought of as consisting of a central atom, capped on either side by two eclipsing sets of three carbon atoms, with one triangular set slightly larger but also closer to the central atom than the other. The trigonal prismatic geometry is unusual in that the vast majority of six-coordinate organometallic compounds adopt octahedral molecular geometry. In the initial report, the IR spectroscopy results were interpreted in terms of an octahedral structure. In 1978, a study using photoelectron spectroscopy appeared to confirm the initial assignment of an Oh structure. The octahedral assignment remained for nearly 20 years until 1989 when Girolami and Morse showed that was trigonal prismatic as indicated by X-ray crystallography. They predicted that other d0 ML6 species such as,, and W6 would also prove to be trigonal prismatic. This report prompted other investigations into the structure of W6. Using gas-phase electron diffraction, Voldenet al. confirmed that W6 is indeed trigonal prismatic structure with either D3h or C3v symmetry. In 1996, Seppelt et al. reported that W6 had a strongly distorted trigonal prismatic coordination geometry based on single-crystalX-ray diffraction, which they later confirmed in 1998. As shown in the top figure at right, the ideal or D3htrigonal prism in which all six carbon atoms are equivalent is distorted to the C3v structure observed by Seppelt et al. by opening up one set of three methyl groups to wider C-W-C angles with slightly shorter C-W bond lengths, while closing the other set of three methyls to 75-78° with longer bond lengths. Deviation from octahedral geometry can be ascribed to an effect known as a second-order Jahn-Teller distortion. In 1995, before the work of Seppelt and Pfennig, Landis and coworkers had already predicted a distorted trigonal prismatic structure based on valence bond theory and VALBOND calculations. The history of the structure of W6 illustrates an inherent difficulty in interpreting spectral data for new compounds: initial data may not provide reason to believe the structure deviates from a presumed geometry based on significant historical precedence, but there is always the possibility that the initial assignment will prove to be incorrect. Prior to 1989, there was no reason to suspect that ML6 compounds were anything but octahedral, yet new evidence and improved characterization methods suggested that perhaps there were exceptions to the rule, as evidenced by the case of W6. These discoveries helped to spawn re-evaluation of the theoretical considerations for ML6 geometries. Other 6-coordinate complexes with distorted trigonal prismatic structures include ,, and. All are d0 complexes. Some 6-coordinate complexes with regular trigonal prismatic structures include , , and the aforementioned .
Reactivity and potential uses
At room temperature, hexamethyltungsten decomposes, releasing methane and trace amounts of ethane. The black residue is purported to contain polymethylene and tungsten, but the decomposition of W6 to form tungsten metal is highly unlikely. The following equation is the approximate stoichiometry proposed by Wilkinson and Shortland: Like many organometallic complexes, WMe6 is destroyed by oxygen. Similarly, acids give methane and unidentified tungsten derivatives, while halogens give the methyl halide and leave the tungsten halide. A patent application was submitted in 1991 suggesting the use of W6 in the manufacture of semiconductor devices for chemical vapor deposition of tungsten thin films; however, to date it has not been used for this purpose. Rather, tungsten hexafluoride and hydrogen are used instead. Treatment of W6 with F2 diluted with Ne at −90 °C affords W6 in 50% yield as an extremely volatile white solid. Hexamethyltungsten reacts with trimethylphosphine in light petroleum to give WMe6, which in neat PMe3, with U.V. irradiation gives the carbyne complex trans-WMe4 in high yield.
