<style type="text/css">P { margin-bottom: 0.08in; direction: ltr; color: rgb(0, 0, 10); }P.western { font-family: "Times New Roman",serif; font-size: 12pt; }P.cjk { font-family: "WenQuanYi Micro Hei"; font-size: 12pt; }P.ctl { font-family: "Lohit Hindi"; font-size: 12pt; }</style>
Permeance of H2 through porous graphene from molecular dynamics
Hongjun Liu a, Sheng Dai a,b, De-en Jiang a,*
a Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
b Department of Chemistry, University of Tennesse, Nnoxville, TN 37966, USA
Summary
The experimental study on the ability of single layer graphene in molecular sieving was confirmed by simulation method. In molecular dynamics simulation study, H2 gas was used as a model gas and its permeance through porous graphene was studied. H2 flux, pressure drop across membrane, and H2 permeance were obtained by trajectories. Permeance was of the order of 1015 GPU in a pressure range of 2 to 163 atm. Pore density was obtained by permeation data.
Gas separation by membrane is energy efficient and has advantages over traditional separation techniques. The inverse relation of gas permeance with membrane thickness led to use single layered graphene in the simulation of H2 permeance through porous graphene. Punching of graphene was carried out, but not for its use as a membrane. Uniformly distributed porous membranes have been prepared and simulation studies showed their use in gas separation. Experimental proofs are also available for gas separation by porous graphene. In an experimental study a highly pressurized microchamber of gas was attached to graphene and it bulged out because of gas pressure. Pore in graphene sheet was made by UV/ozone and the leak rate of the gas was measured. It was found that the leak rate of H2 and CO2 was orders of magnitude higher than other gases. This can be cross checked by simulation study. The approach of first principle and density functional theory led to small timescale simulation and accurate measures of permeance was not possible. In this paper, force field based molecular dynamics simulation to study H2 permeance through porous graphene.
The dimension of graphene membrane was 100 ? x 100 ? with nanopore density of 2500 ?2. Removal of two benzene rings, replacement of four dangling bonds with hydrogen and remaining four dangling bonds along with their carbon atoms with nitrogen gave the nanopore named 4N4H pore. All force field parameters were obtained from literature. The simulation was carried out in LAMMPS in NVT ensemble. Porous graphene was kept fixed during the course of simulation. Hydrogen gas was kept at different initial pressures and permeance of H2 was monitored. Virial equation was used to measure pressure of H2 on both sides of the membrane. The simulation time step was 0.1 fs at 300 K.
The permeance of H2 increased with the increase of pressure. The permeance of H2 was roughly linear fore pressure of 41 atm and above. At low pressures, there were few pass-through events. It required long simulation time. It was found that the flux of H2 was linear with pressure drop and pressure was calculated from Virial equation and also from ideal gas law. It showed that linear regression can be used to estimate the flux against pressure. It was also found that flux difference was 10% at and below 100 atm against virial equation and ideal gas law. It suggests that one can use ideal gas below 100 atm, but at higher pressure differences over 100 atm, virial equation should be used. The simulation results were not different for pressures of 163 atm even for longer time scales. The gas permeance in the pressure range of 2-163 atm was calculated. It was found that the error bars were longer at low time frames because of less permeance events. The permeance was low compared to first principle molecular dynamics study because of low pore density, low temperature, and low pressure. It was found that H2 passes through pore vertically while CO2 passes perpendicularly.
Comparison of this study with experimental study showed that pore density can be measured accurately in simulation and permeance is also known. Comparison of simulation with experiment can help to find pore density in experiments. The discrepancy between simulation and experiment can be thought because of more accurate pore size and density in simulation study. Overall, experimental and simulation were found similar.
Reviewer: Aamir Alaud Din