Accurate determination of binary gas diffusion coefficients by using laser-optical measurement methods and molecular dynamics simulations

• Applicant: Professor Dr.-Ing. Andreas Paul Fröba
• Subject Area: Technical Thermodynamics ; Chemical and Thermal Process Engineering
• Term: from 2016 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 289947578

The major aim of the proposed research project is the accurate determination of the binary diffusion coefficient in binary gas mixtures by experimental and theoretical methods. For this, the laser-optical measurement techniques dynamic light scattering (DLS) and holographic interferometry applied for a Loschmidt cell (HILC) as well as Molecular Dynamics (MD) simulations should be used for the systems methane-carbon dioxide and methane-propane. To improve the potential of MD simulation in predicting diffusion coefficients of such mixtures, it is the aim to further develop the method by using new optimized force fields (FFs) for the specific mixtures. These FFs should be derived from highly accurate ab initio quantum calculations in the zero-density limit by Dr. Hellmann from the University of Rostock in a parallel proposal. To verify the results obtained from the MD calculations, the Fick binary diffusion coefficients obtained from DLS and HILC are required. The application of the two experimental methods provides information on the diffusive mass transport over a broad range of mixture densities, aiming at typical uncertainties of the binary diffusion coefficients below 1%. The data base obtained from HILC should also be compared with Dr. Hellmann's theoretical binary diffusion coefficient data to validate his calculations at the limit of zero density. These theoretical data are then employed by Dr. Hellmann to develop suitable FFs for our MD simulations performed as a function of density in the gas phase. Based on these FFs, the MD simulations should allow for a reliable calculation of mass diffusivities by studying microscopic fluctuations, which is similar to the principle of the DLS method. With MD simulations, the self-diffusion coefficients of the gases in their pure and mixed states as well as the Maxwell-Stefan diffusion coefficient and the thermodynamic factor for the mixtures should be analyzed separately. Combining the two latter properties results in the prediction of the Fick binary diffusion coefficient which should be compared with the experimental data. By decoupling the molecular diffusion process in the MD simulations, it can also be proven to which extent the different diffusivities agree and how they can be related to each other and to common predictive models in literature. A comparison of the simulated diffusivity data obtained from the ab initio calculations-derived FFs for the pure components and the specific mixtures with those obtained by using established literature FFs for the pure components should be carried out for a broad range of thermodynamic states. By this, it should be revealed whether particularly the use of pair-specific FFs is preferable for more accurate computations of mass diffusivities and how well these results agree with those obtained from quantum calculations for the zero-density limit as well as those from the experiments over an extended density range.

• DFG Programme: Research Grants
• Co-Investigator: Dr.-Ing. Michael Heinrich Rausch