Two strategies, two product streams
Start both strategies from the same flask: aryllithium formed by Li–Br exchange, with residual nBuBr still present, and the metal carbonyl about to be added.
Strategy A — 2 eq nBuLi, overnight stir
- The added nBuLi has no reason to prefer nBuBr over the other electrophiles in the flask.
- It attacks the metal carbonyl directly, giving a butyl-acyl species that is not the desired oxy-carbene.
- It deprotonates or adds to the aryllithium-derived intermediate, destroying the desired precursor.
- The residual nBuBr is still there, and it still alkylates whatever oxy-carbene does form.
Net effect: the desired oxy-carbene stream narrows, and two side-product streams widen — the butyl-acyl adduct and the butyl–carbene adduct.
Strategy B — remove or neutralize nBuBr first
- Vacuum transfer, sacrificial aryl halide exchange, or a non-nucleophilic base lowers the nBuBr concentration below the threshold.
- The metal carbonyl is then added to an aryllithium solution that contains no free butyl nucleophile.
- The oxy-carbene forms and survives, because the only species that could alkylate it has been removed.
Net effect: the desired oxy-carbene stream widens, and the butyl–carbene adduct stream nearly closes.
Reading the divergence
The two strategies do not differ by a small percentage. They differ in which species is present when the carbene is formed. In Strategy A, a butyl nucleophile is guaranteed to be present, because it was just added. In Strategy B, the butyl bromide has been removed or converted, so the carbene has nothing to react with except the intended partners.
Caveat
Removal does not guarantee a quantitative yield. The aryllithium can still decompose thermally, and the carbene can still be lost to other pathways. Removal only removes one specific failure mode — the butyl–carbene adduct — and the yield gain is bounded by how much of that pathway was operating.