scholarly journals Improved accuracy and transferability of molecular-orbital-based machine learning: Organics, transition-metal complexes, non-covalent interactions, and transition states

2021 ◽  
Vol 154 (6) ◽  
pp. 064108
Author(s):  
Tamara Husch ◽  
Jiace Sun ◽  
Lixue Cheng ◽  
Sebastian J. R. Lee ◽  
Thomas F. Miller
Molecules ◽  
2021 ◽  
Vol 26 (5) ◽  
pp. 1390 ◽  
Author(s):  
Ilya G. Shenderovich

Due to the rigid structure of 1,3,5-triaza-7-phosphaadamantane (PTA), its 31P chemical shift solely depends on non-covalent interactions in which the molecule is involved. The maximum range of change caused by the most common of these, hydrogen bonding, is only 6 ppm, because the active site is one of the PTA nitrogen atoms. In contrast, when the PTA phosphorus atom is coordinated to a metal, the range of change exceeds 100 ppm. This feature can be used to support or reject specific structural models of organometallic transition metal complexes in solution by comparing the experimental and Density Functional Theory (DFT) calculated values of this 31P chemical shift. This approach has been tested on a variety of the metals of groups 8–12 and molecular structures. General recommendations for appropriate basis sets are reported.


Sign in / Sign up

Export Citation Format

Share Document