Dienes and MO Theory
By James Ashenhurst
Why Are Endo vs Exo Products Favored in the Diels-Alder Reaction?
Last updated: April 19th, 2019
[Advanced] Secondary Orbital Interactions – A Rationale For Why Endo Products Are Favored In The Diels-Alder Reaction
In our last post, we noted that endo products tend to be favored over exo products in the Diels-Alder reaction. [We also introduced a quick-and-dirty rule for telling the difference between endo- and exo- products in the Diels-Alder.]
The preference for endo versus exo is especially curious since the endo products appear to be more sterically hindered.
For example, here’s the Diels-Alder reaction of cyclopentadiene with maleic anhydride. The ratio of endo to exo products in this reaction is about four to one:
Note that in the endo product above, the anhydride is on the underside of the new six-membered ring, whereas in the exo, it points away. This is indeed less sterically hindered.
So why is the endo typically favored?
[To jump ahead, here’s a fact we’ll cover in more detail in a later post: most exo products are in fact more stable than the endo products for steric reasons, but the endo product tends to be formed faster. Furthermore, given enough heat, the Diels-Alder product can revert back to starting materials . Long story short: the Diels-Alder is another example of a reaction that can be run under kinetic or thermodynamic control, where the “endo” is the kinetic product and the “exo” is the thermodynamic product. ]
Our last post showed the mechanism of the Diels-Alder reaction through the orbital interactions of the highest occupied molecular orbital (HOMO) of the diene and the lowest unoccupied molecular orbital (LUMO) of the dienophile.
When the orbital interactions are drawn out in the two different transition states, we see something like this.
Note that in the endo transition state the electron-withdrawing groups point toward the two carbons that will eventually comprise the C2-C3 alkene, whereas in the exo transition state, the electron-withdrawing groups point away from the C2-C3 alkene:
Here’s another example with (E,E)-2,4-hexadiene and acrolein. Again, note how the endo transition state has the electron withdrawing group positioned over the diene:
- In the endo transition state, the carbons which become the C2–C3 double bond are positioned right below the carbonyl carbon of the dienophile.
- In the exo transition state, the electron-withdrawing group points away from the C2–C3 carbons.
Secondary Orbital Interactions Are Possible In The Endo Transition State (But Not The Exo)
By itself, this doesn’t seem to offer any explanation as to why the endo transition state might be favored.
IF we just confine ourselves to examining the “primary” molecular orbitals – i.e. the molecular orbitals involved in bond formation.
The key difference comes when we extend our view and look the “secondary” molecular orbitals of the diene and dienophile that are not directly involved in bond formation, but might still interact with each other.
In this view, there is something special about the endo transition state that isn’t present in the exo.
Because the C2-C3 orbitals of the diene HOMO are positioned close to the C=O orbitals of the dienophile LUMO, they can interact. This isn’t possible in the exo transition state.
This is not a bond-forming interaction (that would be a “primary orbital interaction”), but it is a stabilizing interaction nonetheless. We call it a “secondary orbital interaction”. [We haven’t written about “hyperconjugation” here, but you can think of the interaction as being similar. Essentially, it’s an interaction between an occupied orbital with an unoccupied orbital to form what you can think of as a very weak “partial bond”, and the overall interaction is stabilizing].
This interaction stabilizes the endo transition state to an extent that compensates for the slightly greater steric hindrance.
Here’s what these orbitals look like:
That’s really it. If you know how to draw the complete HOMO of the diene and the complete LUMO of the dienophile, then you can sketch out how they might interact. [We described how to draw out pi molecular orbitals in this post].
We’ve drawn some molecular orbitals for this reaction in the endnotes.
Next Post: Kinetic and Thermodynamic Control of the Diels Alder Reaction
As alluded to above, the endo product tends to be the “kinetic” product, that is, the one that is formed the fastest. Under typical reaction conditions at relatively low temperature, the product distribution reflects the difference in energy between the exo and endo transition states – which is not necessarily the same thing as the difference in energy between the products!
If heated sufficiently, Diels-Alder products can revert to their starting materials, and an equilibrium between the reactants and products can be established. Under these conditions, the product distribution will reflect the difference in energy between the exo and endo products (which tends to favor the exo. )
We’ll go into more detail on the reversibility of the Diels-Alder (and kinetic vs. thermodynamic control) in a future post.
In fairness, there is some debate as to whether secondary orbital effects actually exist.
Endnote #1 (Advanced): Molecular Orbital Diagram for Secondary Orbital Interactions
In the endo transition state, we can have a donation of electron density from electrons in the HOMO of the diene to the empty C=O pi* orbital in the LUMO of the dienophile. This isn’t possible in the exo transition state.
From a molecular orbital perspective, we can draw a pair of electrons in the diene HOMO on the left, and the C=O LUMO on the right.
Any time there is donation from an occupied orbital to an unoccupied orbital there is a lowering of energy.
If the two orbitals in the transition state interact, we can imagine a slight lowering of energy of the electrons (along with a corresponding raise in the energy of the LUMO).
This slight lowering of energy is responsible for the slight preference for the endo transition state.
Endnote #2 (Advanced) Lewis Acid Catalysis Increases endo : exo Selectivity
Here’s an extension of the same idea. It’s known that Lewis acids (such as ZnCl2, TiCl4, SnCl4, and many others) can accelerate the rate of Diels-Alder reactions. It can also increase the endo : exo selectivity!
For instance, compare the rate of non-Lewis acid catalyzed versus Lewis-acid catalyzed ratios in this reaction. [Reference]
What happens is that the LUMO of the C=O bond is lowered when a Lewis acid coordinates to the carbonyl.
For reasons we won’t explain, the interaction between orbitals strengthens as they become closer together in energy. Therefore the secondary orbital interaction is strengthened and a greater stabilization energy is obtained for the endo transition state.
That’s probably too much for this site, but there you go.
For more information, I highly suggest hunting down Dave Evans’ Chemistry 206 notes from Harvard, if you can find them. They are excellent.