How to accurately model IR spectra of nanosized silicate grains

Abstract

We assess the accuracy of various computational methods for obtaining infrared (IR) spectra of nanosized silicate dust grains directly from their atomistic structure and atomic motions. First, IR spectra for a selection of small nanosilicate clusters with a range of sizes and chemical compositions are obtained within the harmonic oscillator approximation employing density functional theory (DFT) based quantum chemical calculations. To check if anharmonic effects play a significant role in the IR spectra of these nanoclusters, we further obtain their IR spectra from finite temperature DFT-based ab initio molecular dynamics (AIMD). Finally, we also study the effect of temperature on the broadening of the obtained IR spectra peaks in larger nanosilicate grains with a range of crystallinities. In this case, less computationally costly classical molecular dynamics simulations are necessary due to the large number of atoms involved. Generally, we find that although DFT-based methods are more accurate, surprisingly good IR spectra can also be obtained from classical molecular dynamics calculations.