Interfacial Barriers for the Transport of Nanoconfined Fluids
The aim of the project is to investigate the interfacial resistance to transport of gases in nanoporous materials and in polymer-filler composites used in mixed matrix membranes. There will be four specific aims: ● Develop and validate a new EMD-based simulation method for simultaneously determining both interfacial and internal transport resistances using an atomistic structural model. ● Determine system length-dependent interfacial and internal transport coefficients of gases and in zeolites and carbon nanotubes using the new EMD-based technique. ● Perform simulations of polymer structure at the polymer-zeolite interface in mixed matrix composites ● Perform simulations of gas transport through the composite and determine the transport properties of the interfacial region This project aims to make advances in the modelling of transport in nanoporous materials applied in emerging energy storage and gas separation technologies, by determining the interfacial barriers critical to the entry and exit of molecules from their nanostructure. The relative importance of the interfacial resistance scales inversely with system size, which is a significant impediment to developing efficient systems at the nanoscale where it becomes governing. During 2019 we have investigated multicomponent gas transport in polymer membranes using molecular dynamics simulations, with the aim of understanding the mechanisms influencing separation properties. We have also investigated interfacial resistance in carbon nanotubes and disordered carbons, finding large entry lengths in finite systems, hitherto overlooked in conventional simulations considering infinite system size through periodic boundary conditions. Such large entry lengths and associated resistance have important consequences for emerging nanoscale membranes, for which transport coefficients based on infinite systems will not be applicable
Area of science
Chemical Sciences, Engineering, Geosciences
Systems used
Magnus
Applications used
(1) LAMMPS, (2) DL_MONTE, and (3) VASPThe Challenge
The principal challenge is to develop a method of determining the interfacial resistance and internal transport coefficient of a finite system between bulk reservoirs, which is significantly more efficient that the conventional dual control volume grand canonical molecular dynamics method. Since any conventional technique will provide the transport coefficient of the whole simulation box including bulk reservoirs, the method developed must be able to isolate the system. To determine this resistance it is also necessary to investigate equilibrium and dynamic properties of the periodic unit cell representing an infinitely large system. Further the method must be adaptable to mixtures, so that the resistance for any component can be determined.
The Solution
To address the above challenge we have implemented a technique based on EMD, which uses the centre of mass mean squared displacements of the molecules in the finite system while accounting for molecules entering and leaving the system into bulk reservoirs. This has allowed the determination of the transport coefficient in the finite system, and the interfacial resistance upon comparing this transport coefficient with that in an infinitely large (i.e. periodic) system
The Outcome
Our EMD simulations have been implemented using LAMMPS, which is highly scalable in both shared and distributed memory architectures. The structure of initial (finite) zeolite has been prepared by performing energy minimisation using VASP software. In addition, we have conducted electronic structure calculations on short polymer chains using VASP to determine the atomic charges. Further, we have conducted extensive EMD simulations on long chain molecules such as 6FDA-durene polymer, and ionic liquids in the presence of inorganic materials including ZIF-8 with different adsorbates on 48-144 procs, for 24 hrs, depending on the system size, expending 2.3 to 6.9 KSU per run. Each EMD job has been repeated for 10 different initial configurations. We have also conducted GCMC simulations using DL_MONTE to investigate the performance of gas adsorption isotherms in the polymers as well as polymer-filler composites. These simulations run on 48 processors for about 24 hrs, requiring 2.3 KSU, with 12 GB RAM and 50 GB of storage space. The availability of a large number of processors has been very beneficial to our work.
A hybrid membrane model representing mixed matrix membrane consisting of 6FDA-durene polymer and ZIF-8, where the interfacial voids that exist between the organic and inorganic phases are filled with ionic liquid (BMIM-BF4), to achieve better gas separation performance.
List of Publications
1. Liu, L., D. Nicholson and S.K. Bhatia, “Influence of Morphology on Transport Properties and Interfacial Resistance in Nanoporous Carbons”, J. Phys. Chem. C. 123, 21050-21058 (2019).
2. Dutta, R.C. and S.K. Bhatia, “Interfacial Barriers to Gas Transport: Probing Solid-Gas Interfaces at the Atomistic Level”, Molecular Simulation, 45, 1148-1162 (2019).
3. Dutta, R.C. and S.K. Bhatia, “Atomistic Investigation of Mixed-Gas Separation in a Fluorinated Polyimide Membrane”, ACS Appl. Polym. Mater. 1, 1359-1371 (2019).
4. Monsalve-Bravo, G.M., Dutta, R.C., and S.K. Bhatia, “Multiscale Simulation of Gas Transport in Mixed-Matrix Membranes with Interfacial Polymer Rigidification”, Microporous Mesoporous Mater. 296, 109982 (2020).
5. Dutta, R.C., and S.K. Bhatia, “Interfacial Engineering of MOF-Based Mixed Matrix Membranes through Atomistic Simulations”, J. Phys. Chem. C 124, 594-604 (2020).
