Finding chair and boat transition states during ring-forming reactions

Question

Reactions that form 6 membered rings most commonly go via either boat or chair transition states but the products themselves do not necessarily adopt the same conformation as the TS.

I'm attempting to model a cyclisation in which an amine attacks a pi system to form a new ring, giving rise to two diasteromers depending on which face of the pi system is attacked.

The products have been explored (MM conformational search then QM optimisation of all low energy conformers) and are both chairs (the boat being significantly high in energy as to be essentially unpopulated).

In the reaction of interest there is precedence in the literature from computational studies that the reactions can go via either boats or chairs (depending on the exact structure) and so I would ideally like to find both possible transition states leading to each diastereomer.

My initial plan was to start with the product and work backwards (the transition state is thought to be late, meaning it should resemble the product) , elongating the bond by x angstroms (optimising at each step whilst constraining the distance) and then looking for a 'maximum' from which I would then optimise to a TS however this doesn't provide a method for me to find the boat TS's.

How does one go about finding all possible transition states? Is it valid for me to find a high energy boat conformation of the product and apply the same bond lengthening exercise?

Answer 1

As you have figured out, this is a non-trivial problem.

It's okay to find a boat conformation that might be a bit higher in energy and see if it can map to a TS. Keep in mind that for 6-membered rings, there are many conformations beyond chair and boat. A half-chair, for example, might be a useful test.

In general, for 6-membered rings, using ring-puckering searches are useful. Calculate the plane of best-fit for the 6-membered ring and displace atoms up or down -- this is similar to sampling the out-of-plane vibrational modes of the ring.

So the ideal situation for the TS would be to consider a set of conformations, that might be higher in energy than the reactant and product. After all, it's a transition state. Then use a nudged elastic band or similar method (your bond-lengthening) to look for the real TS.

Good luck!