GROMACS version: 2023.4
GROMACS modification: No
Hi all,
I have some questions about using the gmx dos command to calculate quantum-corrected heat capacities.
I want to compare Cp values obtained from simulations of a small organic molecule with experimental values at different temperatures. I followed the setup described in “Force Field Benchmark of Organic Liquids: Density, Enthalpy of Vaporization, Heat Capacities, Surface Tension, Isothermal Compressibility, Volumetric Expansion Coefficient, and Dielectric Constant” by Caleman et al. The simulations have a length of 100 ps with a timestep of 0.2 fs and an output frequency of 4 fs for velocities and 0.2 ps for positions, without applying any constraints. The CHARMM36 force field is used here.
However, the integral over the DoS(v) graph (from dos.xvg) does not result in 3*N_Atoms. Does that suggest a problem in my simulation settings? Or is it acceptable to address this issue by (simply) using the “-normdos” option to normalize the DoS(v)?
After adding the correction (Cp-Cv, calculated with gmx energy) to the Cv value obtained from the normalized DoS(v) graphs, the values were in relatively good agreement with molar heat capacities obtained from a thermal conductivity meter, but they were significantly different from values obtained by calorimetry. Has anyone else experienced a similar issue?
Thanks for an answer,
Marius