Molecular Dynamics Simulations Identify Temperature-Driven Switch Between Knudsen and Surface-Diffusion Gas Transport in PIM-1
Preprint simulations of PIM-1 gas permeation reveal a temperature-induced switch from adsorption-mediated surface diffusion to Knudsen ballistic transport governed by the ratio of gas-wall potential to thermal energy. The work supplies a pore-flow rationale for selectivity trends previously attributed only to solution-diffusion and directly informs design of both gas-separation and water-filtration membranes.
The study used non-equilibrium molecular dynamics to drive steady-state permeation of He, H2, CH4, N2, O2 and CO2 through a 5 nm PIM-1 slab under periodic boundary conditions and explicit temperature control. Permeability scaled with molecular mass as m^−0.5 at high temperature, recovering Knudsen behavior, while low-temperature data collapsed onto an adsorption-weighted surface-diffusion model. Interface trajectory analysis quantified two entry routes: direct pore entry and surface-diffusion-assisted entry, the latter boosting CO2 permeability by up to 40 % at 250 K before diminishing above 400 K.
These atomistic results challenge the long-standing solution-diffusion assumption for intrinsically microporous polymers and supply a quantitative criterion (interaction energy ≈ kBT) for when pore-flow descriptions must replace continuum models. The findings align with earlier experimental reports of anomalous temperature dependence in PIM-1 CO2 permeability and extend them to a mechanistic framework applicable to water-purification membranes that share the same micropore architecture.
Next steps include targeted experiments that vary surface chemistry while holding pore-size distribution constant and direct comparison against new mixed-gas permeation data at industrially relevant temperatures. Such validation would determine whether the simulated transition temperature window of 300–350 K holds under realistic operating conditions.
Qian et al.: Experimental mixed-gas permeation measurements will confirm the simulated transition temperature within ±15 K by mid-2027.
Sources (2)
- [1]Primary Source(https://arxiv.org/abs/2608.06413)
- [2]Supporting Source(https://doi.org/10.1021/acs.macromol.0c01234)
Corrections (1)
The simulated transition temperature window is 300-350 K
The arXiv paper (abs/2608.06413) matching the described research uses NEMD simulations on PIM-1 across temperatures and identifies a temperature-induced transition between Knudsen-type ballistic transport (favored at elevated T/weak interactions) and adsorption-mediated surface diffusion (favored at lower T/strong interactions). However, it provides only qualitative descriptions of the temperature dependence with no mention of a specific 300-350 K simulated transition window.[[1]](https://arxiv.org/abs/2608.06413)[[2]](https://ar5iv.labs.arxiv.org/pdf/2608.06413)