The principle of microscopic reversibility (PMR), Curtin-Hammett principle (CHP) and Hammond postulate (HP) are the three pillars of organic reactivity. PMR enables us to construct a reasonable reaction path if the mechanism of the reverse process is known. CHP tells us that stability and reactivity do not necessarily correlate, whereas HP allows us to infer about the transition state structure using intermediates that are sometimes detectable and/or isolable. It is difficult to see what we would do without these fundamentals.
I am teaching my first year organic chemistry course and, while we do not go into the details of PMR and CHP, we do discuss the main elements of HP. The reason HP is useful is that it allows us to weed out the pervasive fallacy “Compound Z must be the reaction product because Z is more stable than the alternative Y”. While I have a habit of saying that it is generally incorrect to state that more stable things are kinetically preferred, it is not necessarily true and here is an example.
Professor Alabugin (http://www.chem.fsu.edu/~alabugin/), whose great talk I just heard at the conference in Moscow, recently published an influential paper in which he re-examined the course of some anionic cyclizations. It appears that there must be a major re-evaluation of some stomping grounds of cyclization mechanisms involving alkynes. The 4-endo/5-exo comparison presents a particularly interesting scenario: the reason for the lower barrier to form the 5-membered ring lies in the exothermicity of the reaction. In other words, the transition state that leads to the formation of the 5-membered ring is lower because the product is more stable. This is an important application of the Marcus theory to polar mechanisms.
amphoteros 
I’m not an expert on this topic so maybe the question is naïve but could there be a chance that the line of causality is not necessarily the one stated in the post? I mean, couldn’t we ever say that product 5-endo is more stable (aka the reaction is more exothermic) because the barrier is lower? I mean, I do know it sounds counterintuitive but couldn’t it possible be a case of post hoc reasoning to say that because the barrier is lower, then the corresponding product will be more stable (i.e. the reaction more exothermic) than if the barrier was higher? I’m thinking that maybe Marcus’ theory holds the answer to my question and probably those blue and red lines never cross.
I enjoy very much your thought provoking blog!
Well – I think that the paper has some answers to your questions, but maybe not all. I will think about it and will ask Igor.
It is a thought-provoking question, Joaquin.
The starting point for the dissection of reaction profiles is a matter of choice. This is a little bit of a chicken vs. egg question and one can argue forever what comes first. In fact, all species of interest (reactant, TS, product) lie on the same potential energy surface and influence each other.
If one follows the reaction coordinate from the reactant to product, then, of course, TS comes before the product and, yes, one can try to analyze the nature of product from the properties of the TS. Or if one moves from the product to the reactant and uses the principle of microscopic reversibility, then TS would control the starting material as well. And then everything will be controlled by just the TS! It is not quite a “reductio ad absurdum” but you can see how this may not be a very productive idea. 🙂
Since chemists understand the electronic structure of reactants and products much better than the properties of TS, it is generally more instructive to start with the product/reactant for modeling the TS than to use the alternative where analysis and predictions derive from an ephemeral saddle point for which we cannot even write a Lewis structure.
As long as one models a reaction potential energy surface as a superposition of two curves (reactant and product), the barrier comes naturally from their intersection. It does not matter how the two potential energy curves are approximated (they can be the Marcus parabolas or Bell-Evans-Polanyi/ Hammond-Leffler surfaces or Shaik-Hiberty VB curves), the origin of reaction barriers is clear. The useful corollary from such analysis is that opens the way for a systematically control over the magnitude of reaction barriers by introducing electronic factors that systematically stabilize or destabilize the products.
In this particular case, one of the products is strained (4-exo) whereas the other one is not (5-endo). The strain difference makes the first reaction considerably less exothermic than the other. The mathematic machinery of the Marcus theory allows one to estimate how much of this effect is present in the TS.
For a more detailed discussion of how Marcus theory can be useful for organic reactions and illustrative figures, see: Alabugin I.V.; Manoharan, M.; Breiner, B.; Lewis, F. “Control of Kinetics and Thermodynamics of [1,5]-Shifts by Aromaticity: A View Through the Prism of Marcus Theory” J. Am. Chem. Soc. 2003, 125, 9329-9342. Or send me an e-mail.