Current
...
thoughts:
Major contributions which determine stability of tran trans conformation of amides are (1) conjugation effect from a planar geometry and (2) nonbonded non-bonded interactions (carbonyl oxygen --- aromatic hydrogen)
From a small test set with six different but similar amides, it was found that even among the similar molecules, cis/trans energy differences vary.
To avoid defining additional torsion terms with very specific SMIRKS pattern, → better description of nonbonded non-bonded interaction and different strength of conjugation effect → (1) vdW parameter fitting to QM torsionscan torsion scan or (2) using WBO interpolated parameters for the spring constant of one-fold cosine function or (3) both (how to fit the two types of parameters should be considered. either step by step or simulataneouslysimultaneously)
(1) Validation of the accuracy of the level of theory(B3LYP-D3BJ/DZVP) for the test set
...
Except for one torsion scan (scan1), other torsion scans from the smaller basis set(DZVP) agree well enough with those from larger basis sets(def2-TZVP(-f), cc-pVTZ(-f))
Current choice of method is not unreliable, but some cases(such as scan1) support our concerns on the size of the current basis set to some extent.
(2) Scatter plots of cis/trans energy differences w.r.t OFF AM1 WBO/ DFT WBO
To see if the usage of WBO interpolated parameter for the spring constant of one-fold cosine function(which is for describing the cis/trans conformation preference), linear dependences of cis/trans energy differences and WBOs from different methods(OFF AM1-Wiberg method, DFT) were plotted.
...
Note that there’s a disagreement of the order of AM1 WBOs with the order of WBOs obtained from DFT.
(3) Two approaches to estimate covalent contribution of QM energies and scatter plots of covalent contribution of cis/trans energy differences w.r.t OFF AM1 WBO/ DFT WBO
...