CREST Paper: https://doi.org/10.1063/5.0197592
Motivation:: In the previous studies we saw that the catalysts can be in the "active" (syn) or "inactive" (anti) conformation. Active means the nitrogen lone pair and hydrogen are forming the catalytic pocket. Inactive means they are twisted 180 degrees relative to each other. I am presuming that once the catalyst gets deprotonated it is somewhat locked into one of these conformations due to the double bond character of the nitrogen-carbon bond. I also want to consider more broadly the conformational landscape of these catalysts to see if we can uncover thing we weren't previously considering.
Goal: Run CREST on the two U1- anion catalyst conformations--active and inactive. We aim to learn a few things:
Possible other studies with CREST:
We start one CREST run 2 with the "syn" structure and one with the "anti" to see how much the intial conditions affect the sampling.
Sorted by energy. The conformers on the left come from starting the run with the "syn" conformation. The conformers on the right come from starting the run with the "anti" conformation. Both CREST runs essentially came up with the same structures, so CREST seems somewhat unsensitive to initial conditions (this is good). The only real difference seems to be that some of the structures that are actually the same by rotational symmetry get cataloged as unique structures because CREST's internal RMSD algorithm for determine structure similarity depends (unwisely) upon the atom labeling (indices).
0
and 1
from Syn) ARE the same structure. They are identified by RMSD as being different structures because of how the atoms are labeled (different indices) and evidently CREST's RMSD calculator relies upon these indices. This is unfortunate. Specifically, 1
is a 180 degree rotation of the left aryl wing which produces the same structure but with different labeling of the atoms (it is actaully a rotamer with the left -CF3 groups rotated by 60 degrees, but it is not a new conformer). CREST does not do a good job of actually determine which structures are unique--I need additional filtering to get a unique set of conformers.~.6 kcal/mol
we saw previously. We now want to know the energy barrier between these structures.Goal:: I'd like to dive in a bit more hands-on with the structures and see if I can narrow us down to a bucket of actually unique structures by collapising strucutres that were identified as unique but are clearly the same though more careful RMSD comparison.
Let's start with the structures from the anti side. The "from syn" side has the same structure but with more duplicates, it appears.
These are clearly the same structure (mild rotamers).
These structures are essentially the same object but with the aryl wing rotated by +/- 60 degrees--essentially mirrored structures. After using VMD to align the structures they appear essentially the same.
These all appear basically the same too.
We essentially have 4 confromer motifs:
Initial questions with our results added: