ISAPT partitioning that cuts 2 sigma bonds in 1 fragment


#1

Does anyone know how I might extend a straightforward ISAPT computation on 1,2-cyclopentatediol to an analogous system where an epoxide accepts the hydrogen bond? The latter is giving me fits because you effectively have to cut 2 sigma bonds to construct the epoxide fragment.

Here is my input for the 1,2-diol system that works as expected.
input.dat (2.6 KB)

However, if I try something similar for the closely related 2,3-epoxycyclopentanol structure I get a variety of error messages that seem to be associated with the charge and multiplicity of the fragment(s).
input.dat (2.3 KB)

I assume the problems stem from trying to make the epoxide O atom its own fragment. Playing with the multiplicity didn’t seem to fix it.

Any suggestions would be most welcome. Thanks!


#2

Maybe I’m misunderstanding your post, but the reason for the error message seems straightforward. You cannot possibly add a doublet, a triplet, and a singlet, and end with a singlet. Only a doublet and a quartet state would be possible.

In particular, shouldn’t the ring itself have an even multiplicity? You have to effectively cut an odd number of bonds, one for the alcohol and two for the epoxide.


#3

Thank you for helping resolve this issue. I was focusing on the fragment associated with the epoxide and lost track of the change(s) that needed to be made to the fragment for the cyclopentane ring.


#4

@tschumpr I’m sorry this reply is so late to the party, but I have to caution you against applying F-/ISAPT to study interactions with epoxides for the same reasons that I cautioned another user against using atomic fragments in this post; especially since you’re breaking two bonds to the same atom (even though they’re both sigma bonds), the charge assignment issues get magnified drastically and make the results unreliable and unpredictable.