Correction of vibrational broadening in molecular dynamics clusters with the normal mode optimization method

Area of Science:

• Computational Chemistry
• Spectroscopy
• Molecular Dynamics

Background:

• Simulating vibrational properties of solutions commonly uses molecular dynamics (MD) clusters.
• Raw cluster geometries from MD simulations often result in unrealistic vibrational band broadening.
• This broadening affects both ab initio and empirical force field calculations.

Purpose of the Study:

• To address and reduce unrealistic vibrational band broadening in simulated solution spectra.
• To investigate the origin of broadening errors in molecular dynamics simulations of solutions.
• To improve the accuracy of simulated Raman and Raman optical activity (ROA) spectra.

Main Methods:

• Developed and applied a partial optimization technique in normal-mode coordinates.
• Analyzed a simplified two-dimensional system to understand the source of broadening errors.
• Applied the partial optimization procedure to simulated Raman and ROA spectra of solvated lactamide.

Main Results:

• Identified anharmonicity, mode coupling, and neglected quantum effects as causes of broadening.
• Demonstrated that partial optimization in normal-mode coordinates significantly reduces broadening.
• Achieved the most realistic simulated spectra for lactamide using a vibrational wavenumber cutoff of approximately 200 cm(-1).

Conclusions:

• Partial optimization in normal-mode coordinates is an effective empirical method to correct vibrational band broadening in MD simulations.
• The technique improves the comparison between experimental and simulated vibrational spectra.
• This approach enhances the reliability of computational studies on the vibrational properties of solutions.