Abstract
We performed fixed-node diffusion Monte Carlo (DMC) calculations to investigate structural and energetic properties of graphenylene (GPNL), a two-dimensional network of sp2-bonded carbon atoms with large near-circular pores, and its H2 separation performance for gas mixtures. We have found that the energetic stability of a GPNL monolayer is comparable to that of γ-graphyne, as evidenced by its large cohesive energy of 6.755(3) eV/atom. Diffusion barriers of several gas molecules, including hydrogen, through a GPNL membrane were determined from the analysis of their adsorption energies depending on the adsorption distance, which led to our estimation for hydrogen selectivity with respect to other target molecules. DMC hydrogen selectivity of a GPNL monolayer was found to be exceptionally high at 300 K, as high as 1010-1011 against CO and N2 gases. This, along with high hydrogen permeance due to its generic pore structure, leads us to conclude that GPNL is a promising membrane to be used as a high-performance hydrogen separator from gas mixtures. We find that when compared to our DMC results, DFT calculations tend to overestimate H2 selectivity, which is mostly due to their inaccurate description of short-range repulsive interactions.
| Original language | English |
|---|---|
| Article number | 144703 |
| Journal | Journal of Chemical Physics |
| Volume | 157 |
| Issue number | 14 |
| DOIs | |
| State | Published - Oct 14 2022 |
Funding
This paper was supported by Konkuk University in 2020. We also acknowledge the support from the Supercomputing Center/Korea Institute of Science and Technology Information with supercomputing resources (Grant No. KSC-2020-CRE-0126) that were used for DFT-PBE calculations. H.S. and A.B. were supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Materials Sciences and Engineering Division, as part of the Computational Materials Sciences Program and Center for Predictive Simulation of Functional Materials. An award of computer time was provided by the Innovative and Novel Computational Impact on Theory and Experiment (INCITE) program and was used to generate all DMC results. This research used resources of the Argonne Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract No. DE-AC02-06CH11357.