Storable sources of transition-metal atoms, or "naked-metalatom" reagents, were first discovered in the pioneering research of Wilke and co-workers, which involved spectacular homoleptic alkene nickel(0) complexes, including [Ni-(cod) 2 ], [1] (cod = 1,5-cyclooctadiene) and [Ni(C 2 H 4 ) 3 ]. [2] Subsequently, Jonas et al. isolated related anionic alkene complexes, [Co(cod) 2 ] À [3] and [Co(C 2 H 4 ) 4 ] À , [4] which proved to be extremely useful synthons for the atomic cobalt anion. Kündig and Timms [5] and Elschenbroich and Möckel [6] then independently discovered that early-transition-metal "nakedatom" reagents could be isolated in the form of quite labile neutral homoleptic naphthalene complexes, [M(C 10 H 8 ) 2 ] (M = V, Cr, Mo). Anionic early-transition-metal homoleptic naphthalene complexes were later established for zirconium ([Zr(h 4 -C 10 H 8 ) 3 ] 2À ) [7] and tantalum ([Ta(h 4 -C 10 H 8 ) 3 ] À ) [8] and have been shown to function as sources of atomic Zr 2À and Ta À ions in their reactions with carbon monoxide and similar ligands. However, no compounds of this class are presently known for the later transition metals. Indeed, even heteroleptic naphthalene complexes containing formally negativevalent metals are unknown for d-block elements to the right of manganese, [9] and also unknown for the corresponding heavier metals within each triad. For these reasons, we report herein on the isolation, structural characterization, and some chemical properties of the first compound of this type, [Co(h 4 -C 10 H 8 )(h 4 -cod)] À (1). Also, two other naphthalenecobaltate(1À) complexes were unexpectedly obtained in the course of developing a "noncobaltocene" route to alkenecobaltates(1À).
Recently, we disclosed that the reaction of cobalt(II) bromide with three equivalents of potassium naphthalene gave a thermally unstable deep-red species, which functioned