Reply to “Comment on ‘Efficient stress relaxation in molecular dynamics simulations of semiflexible n -alkanes’ ”

The main issue of our original paper [Phys. Rev. E 58, 6766 (1998)] is to demonstrate that the so-called atomic scaling, in which all available degrees of freedom are coupled to the pressure bath, is more efficient for stress relaxation in large molecules than the conventional molecular scaling in which the molecular centers of mass are coupled to the pressure bath, and internal degrees of freedom are left unchanged. Marchi and Procacci (MP) claim that this is not the case and try to demonstrate this by a simulation of the alkane system II (dotriacontanes, 32 monomers) treated in our paper, comparing atomic and molecular scaling with their R-RESPA integrator. They state that the stress-relaxation process should happen within a few picoseconds. As a possible explanation for their findings, they state an incorrect computation of the molecular pressure in our paper. Furthermore, MP claim there are further inconsistencies in our paper. In this Reply, it will be shown that contrary to the statements of MP, the virial has been computed correctly. Moreover, the inconsistency statement by MP results from the fact that MP have confused features in the figures of our paper. Finally, we do not agree that the stress relaxation of dotriacontanes seen in our simulations on the time scale of hundreds of picoseconds should happen within a few picoseconds. At room temperature, these systems form waxes and a slowing down of stress relaxation with respect to the liquid phase is to be expected.