An efficient computational approach for the calculation of the vibrational density of states

The vibrational density of states is a very important quantity that allow for the calculation of partition functions and many other thermodynamic properties (1-6). In this work, we present a new code, called Paradensum, developed for the calculation of the vibrational density of states for high dimensional systems. Paradensum is based on the implementation of the Wang-Landau Monte Carlo algorithm (7) and is written for parallel architectures. The code accuracy was tested on several molecular systems, including some cases in which an exact values for the vibrational density of states was available through direct counting. Paradensum shows a significant computational speed-up respect to standard approaches and is able to handle systems with 150 degrees of freedom. These features make it a very promising tool for future calculations of thermodynamic properties and thermal rate constants of complex systems.