A Mixed Quantum-Classical Molecular Dynamics Study of anti-Tetrol and syn-Tetrol Dissolved in Liquid Chloroform: Hydrogen-Bond Structure and Its Signature on the Infrared Absorption Spectrum

The conformation of chloramphenicol in the ordered and disordered phases

The conformational behavior of chloramphenicol (CHL) in the solid, liquid and vapor phases is revisited here by means of FTIR spectroscopy and QM methods. In the crystalline phase, both the IR analysis and QM computations discard the conformer proposed by Acharya et al. (Acta Cryst., 1979, B35:1360-1363) and support the one proposed by Chatterjee et al. (J. Cryst. Mol. Struct., 1979, 9:295-304), characterized by an intramolecular OH⋯O hydrogen bond in which the primary hydroxyl group acts as hydrogen bond donor. The conformational behavior of CHL in the liquid and gas phases has been analyzed using QM calculations. The Self-Consistent Reaction Field (SCRF) method with the Onsager solvation model has been used for the initial optimizations in solution, and the lowest energy conformers have been refined using the Solvation Model based on Density (SMD). In solution environment the intramolecular OH⋯O hydrogen bond in CHL is reversed so that the secondary hydroxyl group acts as hydrogen bond donor. In addition, the dichloroacetamide group folds back further over the phenyl ring to form an intramolecular CCl⋯π halogen bond. Two different halogen bonds are actually observed (each one with a different chlorine atom) resulting in two different stable conformers, that can be detected by FTIR spectroscopy due to the conformational sensitivity of the CO group to the conformation of the dichloroacetyl group. Finally, the stability of the conformers with the polarity of the medium is also discussed.

Nonadiabatic Dynamics via the Symmetrical Quasi-Classical Method in the Presence of Anharmonicity

The symmetrical quasi-classical (SQC) method recently proposed by Miller and Cotton allows one to simulate nonadiabatic dynamics based on an algorithm with classical-like scaling with respect to system size. This is made possible by casting the electronic degrees of freedom in terms of mapping variables that can be propagated in a classical-like manner. While SQC was shown to be rather accurate when applied to benchmark models with harmonic electronic potential energy surfaces, it was also found to become inaccurate and to suffer numerical instabilities when applied to anharmonic systems. In this paper, we propose an extended SQC (E-SQC) methodology for overcoming those discrepancies by describing the anharmonic nuclear modes, which are coupled to the electronic degrees of freedom, in terms of classical-like mapping variables. The accuracy of E-SQC relative to standard SQC is demonstrated on benchmark models with quartic and Morse potential energy surfaces.