Concentration gradient driven molecular dynamics: a new method for simulations of membrane permeation and separation

Aydın Özcan, Claudio Perego, Matteo Salvalaglio, Michele Parrinello, A. Özgür YazaydınChemical 2017 · DOI

Abstract

In this study, we introduce a new non-equilibrium molecular dynamics simulation method to perform simulations of concentration driven membrane permeation processes. The methodology is based on the application of a non-conservative bias force controlling the concentration of species at the inlet and outlet of a membrane. We demonstrate our method for pure methane, ethane and ethylene permeation and for ethane/ethylene separation through a flexible ZIF-8 membrane. Results show that a stationary concentration gradient is maintained across the membrane, realistically simulating an out-of-equilibrium diffusive process, and the computed permeabilities and selectivity are in good agreement with experimental results.

Introduction

Membrane separation is a critical industrial process utilized in applications ranging from water desalination to natural gas sweetening. Designing more efficient membranes requires a deep understanding of how molecules navigate nanoscale pores, often studied via Molecular Dynamics (MD). However, real-world membrane separation is an out-of-equilibrium process driven by pressure or concentration gradients, which is notoriously difficult to simulate accurately.

Traditional methods like Dual Control Volume Grand Canonical MD (DCV-GCMD) rely on stochastic Monte Carlo steps to insert or delete molecules, which fails in dense fluid environments. Other approaches using moving "pistons" suffer from feed depletion, where the simulation ends once the source molecules are exhausted. This paper introduces Concentration Gradient Driven MD (CGD-MD), a deterministic approach that maintains steady-state gradients without these drawbacks.

Methodology

The CGD-MD method replaces physical walls or random particle insertions with localized "Control Regions" at the membrane's inlet and outlet. A non-conservative bias force is applied to molecules within these regions to regulate concentration. If the local density drops below a target value, the force pulls molecules in from the bulk; if it exceeds the target, it pushes them out.

Schematic representation of the CGD-MD simulation setup, showing the membrane and control regions (ICR/OCR) where bias forces are applied to maintain the concentration gradient.

By utilizing Periodic Boundary Conditions (PBCs), molecules that permeate the membrane and exit the outlet loop back to the inlet bulk fluid. This creates a continuous "conveyor belt" effect, allowing for indefinite steady-state simulations. The method was validated using a flexible ZIF-8 metal-organic framework (MOF) to separate methane, ethane, and ethylene.

Results and Discussion

The CGD-MD approach successfully maintained stable target concentrations (e.g., ~40 bar at the inlet and ~0 bar at the outlet) with minimal fluctuations. For pure gas simulations, the predicted permeability order—Ethylene > Methane > Ethane—perfectly matched experimental observations.

In a more challenging test involving a 50/50 equimolar mixture of ethylene and ethane, the simulation predicted a permeation selectivity of 2.0. This is in excellent agreement with the experimental value of 2.6 at 3.5 bar, especially considering the nearly identical kinetic diameters of the two molecules (4.16 Å vs 4.44 Å). Unlike previous methods, CGD-MD held the feed mixture at a strict, constant ratio throughout the simulation, mirroring industrial conditions.

Conclusion

CGD-MD offers a robust, deterministic alternative for simulating membrane processes. By avoiding the computational bottlenecks of particle insertion, it is particularly well-suited for dense fluid phases where traditional methods fail. While maintaining a perfect vacuum at the outlet remains a minor challenge due to statistical fluctuations at low densities, the method's ability to simulate long-term, steady-state diffusive behavior makes it a powerful tool for screening new membrane materials. Future work will extend this methodology to liquid-phase separations, such as water purification.

Summary updated

Abstract

In this study, we introduce a new non-equilibrium molecular dynamics simulation method to perform simulations of concentration driven membrane permeation processes. The methodology is based on the application of a non-conservative bias force controlling the concentration of species at the inlet and outlet of a membrane. We demonstrate our method for pure methane, ethane and ethylene permeation and for ethane/ethylene separation through a flexible ZIF-8 membrane. Results show that a stationary concentration gradient is maintained across the membrane, realistically simulating an out-of-equilibrium diffusive process, and the computed permeabilities and selectivity are in good agreement with experimental results.
BibTeX
@inproceedings{ozcan2017concentration,
  author = {Aydın Özcan and Claudio Perego and Matteo Salvalaglio and Michele Parrinello and A. Özgür Yazaydın},
  title = {{Concentration gradient driven molecular dynamics: a new method for simulations of membrane permeation and separation}},
  booktitle = {Chemical},
  year = {2017},
  doi = {10.1039/c6sc04978h},
}
Figure