Crystallographic and Dynamic Aspects of Solid‐State NMR Calibration Compounds: Towards ab Initio NMR Crystallography

Abstract

The excellent results of dispersion‐corrected density functional theory (DFT‐D) calculations for static systems have been well established over the past decade. The introduction of dynamics into DFT‐D calculations is a target, especially for the field of molecular NMR crystallography. Four 13C ss‐NMR calibration compounds are investigated by single‐crystal X‐ray diffraction, molecular dynamics and DFT‐D calculations. The crystal structure of 3‐methylglutaric acid is reported. The rotator phases of adamantane and hexamethylbenzene at room temperature are successfully reproduced in the molecular dynamics simulations. The calculated 13C chemical shifts of these compounds are in excellent agreement with experiment, with a root‐mean‐square deviation of 2.0 ppm. It is confirmed that a combination of classical molecular dynamics and DFT‐D chemical shift calculation improves the accuracy of calculated chemical shifts.

Enhanced NMR Discrimination of Pharmaceutically Relevant Molecular Crystal Forms through Fragment-Based Ab Initio Chemical Shift Predictions

“…Of course, dynamics can also play an important role, and deviations from a unit slope might also result from neglecting those effects. 45,46 Because the 15 N benchmark set contains only 51 shifts, adding the dozen more shifts with a systematic error from this study does modestly alter the regression line and improves the errors for these 12 shifts by about 1 ppm at the fragment PBE0 level. Nevertheless, because including so many nitrogen shifts from only two systems might overly bias the overall test set, we continue to advocate use of the original regression line from ref (78).…”

The application of tailor-made force fields and molecular dynamics for NMR crystallography: a case study of free base cocaine

“…In this version of CASTEP, the Grimme-2010 (-D3) dispersion correction is not available. Thus, to keep the consistency of the protocols we used for two of our previous studies (Li et al, 2014(Li et al, , 2016, the semi-empirical type Grimme-2006 dispersioncorrection scheme (Grimme, 2006) with the PBE functional was used, referred to as PBE-D2. The protocol for the energy minimizations was described by Li et al (2014) with one exception in this particular case: energy minimizations with only the hydrogen atoms allowed to move were omitted, because all the crystal structure candidates were generated from PBE-D3 calculations; the positions of the hydrogen atoms are, in most cases, more accurate than the experimental X-ray diffraction data.…”

“…DFT calculations are conducted at zero kelvin, whereas SS-NMR experiments are usually performed at ambient temperature and represent an average of space and time, thus leading to an inconsistency between experimental and theoretical predictions. Therefore, methods that introduce vibrational averaging to the system of interest have been developed and applied in a series of case studies, including introducing the effects of temperature using a perturbative expansion within the harmonic approximation (Monserrat et al, 2014), classical molecular dynamics (MD) with transferable force fields (De Gortari et al, 2010;Li et al, 2016) or force field parameters derived from ab initio MD simulations (Robinson & Haynes, 2010), Born-Oppenheimer approximation-based ab initio MD (De Gortari et al, 2010;Dračínský & Bouř, 2012;Dračínský & Hodgkinson, 2013), Car-Parrinello MD (Wegner et al, 2011), path integral MD (Dračínský & Hodgkinson, 2014;Dračínský et al, 2016), quantum Monte Carlo (Monserrat et al, 2014) and so forth. The most popular method for integrating the thermal motion in ab initio NMR calculations is ab initio MD; however, it requires intensive computational resources to carry out a simulation on a timescale of picoseconds, and the simulation box is usually restricted to a single unit cell.…”

“…In this paper, we therefore combine the two approaches and run MD simulations for each of the predicted structures, requiring over 600 SS-NMR calculations. Previous investigations revealed that the usefulness of MD averaging depends on the quality of the molecular mechanical force field (Li et al, 2016;Nemkevich et al, 2010). The quality of the force field is addressed by using a TMFF, i.e.…”

Computational Dehydration of Crystalline Hydrates Using Molecular Dynamics Simulations

“…MD simulations have been successfully used to study diverse problems, such as protein folding 24 and drug discovery. 25,26 However, the use of MD simulations to investigate organic crystal structures have been relatively limited [27][28][29] and does to our knowledge not include any studies of dehydration processes.…”