Substitution structures of large molecules and medium range correlations in quantum chemistry calculations

Pate, Brooks

Permalink

https://hdl.handle.net/2142/97109 Copy

Description

Title

Substitution structures of large molecules and medium range correlations in quantum chemistry calculations

Author(s)

Pate, Brooks

Contributor(s)

Evangelisti, Luca

Issue Date

2017-06-19

Keyword(s)

Comparing theory and experiment

Abstract

A study of the minimally exciting topic of agreement between experimental and measured rotational constants of molecules was performed on a set of large molecules with 16-18 heavy atoms (carbon and oxygen). The molecules are: nootkatone (C$_{15}$H$_{22}$O), cedrol (C$_{15}$H$_{26}$O), ambroxide (C$_{16}$H$_{28}$O), sclareolide (C$_{16}$H$_{22}$O$_{2}$), and dihydroartemisinic acid (C$_{15}$H$_{24}$O$_{2}$). For this set of molecules we obtained 13C-subsitution structures for six molecules (this includes two conformers of nootkatone). A comparison of theoretical structures and experimental substitution structures was performed in the spirit of the recent work of Grimme and Steinmetz.[1] Our analysis focused the center-of-mass distance of the carbon atoms in the molecules. Four different computational methods were studied: standard DFT (B3LYP), dispersion corrected DFT (B3LYP-D3BJ), hybrid DFT with dispersion correction (B2PLYP-D3), and MP2. A significant difference in these theories is how they handle medium range correlation of electrons that produce dispersion forces. For larger molecules, these dispersion forces produce an overall contraction of the molecule around the center-of-mass. DFT poorly treats this effect and produces structures that are too expanded. MP2 calculations overestimate the correction and produce structures that are too compact. Both dispersion corrected DFT methods produce structures in excellent agreement with experiment. The analysis shows that the difference in computational methods can be described by a linear error in the center-of-mass distance. This makes it possible to correct poorer performing calculations with a single scale factor. We also reexamine the issue of the “Costain error” in substitution structures and show that it is significantly larger in these systems than in the smaller molecules used by Costain to establish the error limits._x000d_ _x000d_ [1] Stefan Grimme and Marc Steinmetz, “Effects of London dispersion correction in density functional theory on structures of organic molecules in the gas phase”, Phys. Chem. Chem. Phys. 15, 16031-16042 (2013)._x000d_

Publisher

International Symposium on Molecular Spectroscopy

Type of Resource

text

Language

eng

Permalink

http://hdl.handle.net/2142/97109

DOI

https://doi.org/10.15278/isms.2017.MI05

Copyright and License Information

Copyright 2017 Brooks Pate

Owning Collections