The issue of which proton is removed when there is a choice is not as simple in E2 as it is in the other mechanisms. Two empirical rules have been formulated which apply to different systems, and essentially say the opposite things. These rules were developed by Hofmann and Saytzev (working separately) in the 19th century. Hofmann’s work applied to quaternary amine compounds, and Saytzev’s to alkyl halides. The work of these two and the rules obtained is summarized below;
Hofmann
Applied to quaternary amine compounds (an example of which is shown in the reaction below), his rule states that the “alkene will predominate which has least alkyl substituents on the double bond carbons”.
This appears to be contrary to expectations where the most stable double bond (i.e. the most substituted and therefore thermodynamically stable) predominates – say in E1 reactions, for example.
The thing that is of particular note here is that the leaving group is positively charged.
This positively charged group is more electron withdrawing than a neutral group (e.g. a halide), and so will stabilise negative charge building up where the proton is removed. The result of this is a TS slightly earlier than the ‘typical’ E2 TS – where the proton has been partially removed, but no double bond character has yet built up.
Therefore the proton most likely to be removed is the most acidic, not the one which will lead to a more stable double bond, because double bond character is not involved in the TS. Of a choice between a proton on a more substituted, or less substituted carbon centre, the one on the less substituted centre will be slightly more acidic because the carbanion formed will be less destabilised by inductive electron donation from alkyl groups. Hence, in Hofmann E2, the alkene formed is the least substituted of the possibilities.
Saytzev
Applied to alkyl halide compounds, this rule states that the “alkene will predominate which has most alkyl substituents on the double bond carbons” – i.e. the complete opposite of Hofmann, and therefore more in line with what would generally be expected.
In order for the stability of the alkene product to enter into the regiochemistry of the reaction, the transition state must include some alkene character – so it is reasonable to suppose that the TS for Saytzev elimination is the all-encompassing TS mentioned in reference to the general E2 mechanism. For completeness, an example of Saytzev elimination is included below;