In the chemical zoo this month: calcium octacarbonyl as reported by Xuan Wu et al. in the journal science. (link) The habitat of this creature is very limited and confined to a laboratory in Shanghai and then even only at 4 degrees above the absolute freezing point and guarded by neon.
metal carbonyls are nothing special. Examples are nickel tetracarbonyl, iron pentacarbonyl, chromium hexacarbonyl but the problem is calcium is not a metal. Calcium should do simple things like being in calcium oxide or calcium sulfate. Calcium has 2 valence electrons in the outer s-orbital and therefore prefers a +2 ionic state and anionic bonding partners.
Ca(CO)8 was made by evaporating the calcium metal with a laser and shooting pulses of it at solid carbon monoxide and neon. Analysis was done with fourier transform infrared absorption spectroscopy. In the experimental sweet spot a single carbonyl stretching band emerging with increasing CO concentration was taken as evidence for a 18-electron octacarbonyl complex with cubic Oh symmetry. It also helped that the octacarbonyls showed up in mass spectroscopy.
The type of bonding is explained (relatively speaking, these articles are not written for the general public) in terms of spd orbital mixing. In the MO diagram the two calcium spd-electrons end up as the HOMO triplet state in a type of pi backbonding with the antibonding pi orbitals in CO with for this molecule the largest energy contribution. Two of the CO electrons do absolutely nothing and the rest of them (14) are huddled together sharing calcium antibonding p-orbitals. The d-orbital electrons in calcium are energetic and therefor calcium is a good electron donor. The atom is also large.