α-halo ketones when treated with a strong base (generally alkoxides) lead to the formation of esters with rearrangement of the carbon frame work.
The mechanism
Through product studies it was proposed (by Favorskii himself) that the rearrangement resembles Benzyl -benzilic acid rearrangement (this mechanism had been established at that time)
Applying this to Favorskii rearrangement, Compound 1 and 2 should give the corresponding products as shown below.
In actual fact both compounds give the same product hydrocinnamic acid. Hence it was thought that some other mechanistic pathway to be operative.
Loftfield Mechanism
The University of New Mexico Department of Biochemistry & Molecular Biology
Loftfield, in 1951 described a pathway that goes through a cyclopropanone derivative as an intermediate which is the most satisfactory general mechanism to date.
Applying this mechanistic route to the reactions involving compound 1 and 2.
Both go through the same intermediate hence the same product.
It is possible for an anion to be formed at the second α-carbon, but the presence of halogen would destabilise it.
So far the mechanistic probe of product identification has been used. This was then confirmed by extensive isotopic labelling studies. Some of those are described in the following schemes.
Synthesis labelled α-chloro cyclohexanone.
Degradation studies. Identification and estimation of the labelled carbon in the products.