Presentation Title

Fragment-Based Approach for Predicting 51V Nuclear Magnetic Resonance Chemical Shifts

Faculty Mentor

Joshua D. Hartman

Start Date

17-11-2018 12:30 PM

End Date

17-11-2018 2:30 PM

Location

CREVELING 113

Session

POSTER 2

Type of Presentation

Poster

Subject Area

physical_mathematical_sciences

Abstract

Nuclear magnetic resonance (NMR) spectroscopy plays a central role in determining molecular structure for biological and pharmaceutical compounds. The complex nature of these systems presents unique challenges in the interpretation of experimental spectral data. Thus, NMR investigations are becoming increasingly reliant on computation for mapping spectral features to chemical structures. In particular, fragment-based chemical shielding calculations have proven highly successful in predicting NMR chemical shieldings for a variety of pharmaceutically relevant systems. However, the application of fragment-methods has largely neglected elements beyond the second row. Such elements are of tremendous biological and pharmaceutical interest. For example, 51V is a quadrupolar, NMR-active nuclei present in numerous enzyme active sites and pharmaceutically relevant compounds. Challenges associated with high-accuracy prediction of 51V chemical shifts using conventional computation methods has limited their role in 51V NMR studies. In the present work, we demonstrate that fragment methods, coupled with our novel electrostatic embedding model, faithfully reproduce experimental shifts with modest computational cost. Using a carefully constructed test set of 11 pharmaceutically relevant 51V-containing molecular crystals, we show 51V isotropic shieldings are strongly dependent on long-range electrostatic interactions, and accuracy is improved through the application of hybrid density functionals. Specifically, our combined cluster/fragment methods using the PBE0 density functional give an overall test set root-mean-square error of only ~7.5 ppm, on par with the uncertainty in the experimental values. The results of this study suggest that our unique approach is capable of producing high-resolution structural refinement on both periodic and non-periodic systems alike.


Keywords:

  • 51V Nuclear Magnetic Resonance
  • Fragment-based Methods
  • Vanadium
  • Solid-state NMR
  • Crystal Structure Determination
  • Isotropic Chemical Shift Prediction
  • Density Functional Theory

This document is currently not available here.

Share

COinS
 
Nov 17th, 12:30 PM Nov 17th, 2:30 PM

Fragment-Based Approach for Predicting 51V Nuclear Magnetic Resonance Chemical Shifts

CREVELING 113

Nuclear magnetic resonance (NMR) spectroscopy plays a central role in determining molecular structure for biological and pharmaceutical compounds. The complex nature of these systems presents unique challenges in the interpretation of experimental spectral data. Thus, NMR investigations are becoming increasingly reliant on computation for mapping spectral features to chemical structures. In particular, fragment-based chemical shielding calculations have proven highly successful in predicting NMR chemical shieldings for a variety of pharmaceutically relevant systems. However, the application of fragment-methods has largely neglected elements beyond the second row. Such elements are of tremendous biological and pharmaceutical interest. For example, 51V is a quadrupolar, NMR-active nuclei present in numerous enzyme active sites and pharmaceutically relevant compounds. Challenges associated with high-accuracy prediction of 51V chemical shifts using conventional computation methods has limited their role in 51V NMR studies. In the present work, we demonstrate that fragment methods, coupled with our novel electrostatic embedding model, faithfully reproduce experimental shifts with modest computational cost. Using a carefully constructed test set of 11 pharmaceutically relevant 51V-containing molecular crystals, we show 51V isotropic shieldings are strongly dependent on long-range electrostatic interactions, and accuracy is improved through the application of hybrid density functionals. Specifically, our combined cluster/fragment methods using the PBE0 density functional give an overall test set root-mean-square error of only ~7.5 ppm, on par with the uncertainty in the experimental values. The results of this study suggest that our unique approach is capable of producing high-resolution structural refinement on both periodic and non-periodic systems alike.


Keywords:

  • 51V Nuclear Magnetic Resonance
  • Fragment-based Methods
  • Vanadium
  • Solid-state NMR
  • Crystal Structure Determination
  • Isotropic Chemical Shift Prediction
  • Density Functional Theory