Zinc (0) chemistry: does the missing 18-electron zinc tricarbonyl really exist?

Recently, a combined laser ablation and density functional theory study (Jiang and Xu, J. Am. Chem. Soc. 2005, 127, 8906) claimed the existence of the long-sought 18-electron member of the first-row transition metal carbonyl complex, Zn(CO)(3). In this paper, we systematically investigate the thermodynamic and kinetic stability of Zn(CO) 3 towards CO-extrusion at the BP86, B3PW91, BPW91, PBEPBE, BH&HLYP, B3LYP, MP2, MP4SDQ, QCISD, CCSD and CASPT2 levels as well as the Born-Oppenheimer molecular dynamic (BOMD) simulation. All these calculations consistently reveal that the 18e Zn(0) complex Zn(CO)(3) is neither a genuine minimum point nor kinetically stable with negligibly low barriers. In particular, Zn(CO) 3 is quite thermodynamically unstable with respect to the fragments Zn-1 + 3CO by around 40 kcal mol(-1) at all the three sophisticated correlation levels, i.e., MP4SDQ, QCISD and CCSD. We thus conclude that the tricarbonyl Zn(0) complex, Zn(CO)(3), should not exist even for spectroscopic characterization. Interestingly, our extensive structural search predicts that two triplet di-zinc carbonyls, i.e., (3)(CO)ZnZn and (3)(CO)(2)ZnZn, have noticeable kinetic stability (10.41 and 8.11 kcal mol(-1) at the CCSD level) against the respective CO- and Zn-extrusion, which can be compared with the value 8.70 kcal mol(-1) for the already detected Zn-3(CO)(2) (Jiang and Xu, J. Phys. Chem. A 2006, 110, 7092). Our designed (3)(CO) ZnZn and (3)(CO)(2)ZnZn together with the experimentally known (ZnCO)-Zn-3 and Zn-3(CO)(2) are formally associated with the zinc (0) "spin-based zinc carbonyls'' and should be considered as remarkable, since most of the known zinc complexes usually contain + 2 or + 1 oxidation state Zn.