What the oxy-carbene looks like
When the aryllithium attacks a metal carbonyl, the first product is an anionic acylate. Alkylation or protonation at the oxygen gives an oxy-carbene, a Fischer-type carbene:
\[\text{M(CO)}n + \text{Ar–Li} \rightarrow [\text{M(CO)}{n-1}(\text{C(O)Ar})]^- \xrightarrow{\text{R'X}} (\text{CO})_{n-1}\text{M=C(OR')Ar}\]
The key structural feature is the metal–carbon double bond to the carbene carbon, with an alkoxy group on that same carbon.
Why the carbene carbon is electrophilic
Two electronic effects combine to make the carbene carbon electron-poor.
- The metal carbonyl fragment is a strong π-acceptor. The CO ligands pull electron density away from the metal, and the metal in turn back-donates only weakly to the carbene carbon. That leaves the carbene carbon with significant positive character.
- The alkoxy substituent donates through oxygen. The oxygen lone pair stabilizes the carbene by π-donation, but it also concentrates electron density on oxygen rather than on carbon. The carbon remains the electron-poor end of the M=C bond.
A useful way to think about it: the carbene carbon behaves like the carbonyl carbon of an ester or an amide. It is electrophilic and it reacts with nucleophiles.
Why butyl bromide can attack it
Butyl bromide is an alkyl halide. The carbon bearing the bromide is electrophilic, but the bromide itself is a leaving group, and the butyl fragment can be delivered to a nucleophile. In the presence of the oxy-carbene, the carbene carbon is the electrophile and the butyl group is the nucleophile. The result is a butyl group attached to the carbene carbon, which is the undesired adduct.
Why the aryllithium is not the vulnerable site
In this step the aryllithium has already done its job: it has attacked the metal carbonyl and become part of the carbene. The remaining aryllithium is a nucleophile, not an electrophile. The site that butyl bromide attacks is the electrophilic carbene carbon.