Vapour-liquid equilibria from molecular simulations: some issues affecting reliability and reproducibility

摘要

Molecular simulation data of the vapour-liquid equilibria (VLE) published in the period 2005-2016 and listed in the Web-of-Science collection have been scrutinised and their correctness examined using the recently proposed exact compressibility factor criterion [I. Nezbeda, J. Chem. Eng. Data 61, 3964 (2016)]. It turns out that a large number of the examined data are very inaccurate, if not completely wrong, and this is illustrated by several examples. This problem goes, unfortunately, unnoticed and the data are further used by other researchers. The finding goes hand-in-hand with the becoming practice of ignoring the common etiquette of presenting (pseudo)experimental data, i.e. to provide sufficient information both on technical details and on data post-simulation processing which could enable anyone their independent check and further reliable use. Moreover, the problem of the correctness of published data does not concern only VLE data but any simulation results and these cases along with potential reasons are therefore also briefly discussed. An appeal is therefore made both to the community of simulators and users to examine data using available tools before publishing or using them. Similarly, an appeal is made also to reviewers to insist that the submitted papers with simulation data do contain all necessary details.