Water diffusion within hydrated model grafted polymeric membranes with bimodal side chain length distributions

Abstract

The effect of bimodal side chain length distributions on pore morphology and solvent diffusion within hydrated amphiphilic polymeric membranes is predicted. Seven polymeric architectures are constructed from hydrophobic backbones from which at regular intervals side chains branch off that are alternatingly short (composed of p hydrophobic A fragments or beads) and long (q A fragments, q > p). The side chains are end-linked with a hydrophilic C fragment. Pore morphologies at a water volume fraction of 0.16 are calculated by dissipative particle dynamics (DPD). Water diffusion through the water containing pores is calculated by tracer diffusion calculations through 140 selected snapshots and from the water bead motions. Diffusion constants decrease with difference in side chain lengths, q - p. Overall, the distance between pores also decreases with q - p. The results are explained by counting for every architecture the average number of bonds 〈N(bond)〉 between an A and the nearest C fragment. These results are in line with a database that contains more than 60 architectures. Diffusion constants tend to increase linearly with 〈N(bond)〉|C|(-1)|A|, where |C| and |A| are the C and A bead fractions within the architecture. 〈N(bond)〉 is therefore expected to be an interesting design parameter for obtaining low percolation thresholds for solvent and/or proton diffusion.

Dissipative Particle Dynamics Modeling of Polyelectrolyte Membrane–Water Interfaces

“…Dissipative particle dynamics (DPD) is a coarse-grained simulation method used to model long-time temporal evolution and large spatial length scales. DPD simulations have been widely used to study PEMs [8][9][10][11][12][13][14][15]. Yamamoto et al [11] used DPD to analyze the phase-separated Nafion structure and found the simulated structure factors and pore radii close to the experimental values.…”

“…Dorenbos [14] found that the water diffusion increased with increasing difference between side-chain lengths for a bimodal distribution of side-chain lengths. Sepehr et al [15] simulated anion exchange membranes (AEM) using DPD and found water cluster shapes varying with hydration. They also reported larger water domains for chloride anions, as charge carriers, when compared to the hydroxyl anion charge carriers.…”

“…Comparison of the chain radius of gyration (R g ) in different layers(1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17), at varying distances from the bottom Z boundary, for (a) Dow, (b) Nafion, and (c) Aciplex PEMs. Layer 1 is nearest to the bottom Z-boundary and layer 17 is at the interface of water and the membrane.…”

Coarse-grained model of water diffusion and proton conductivity in hydrated polyelectrolyte membrane

“…The segregation morphology whose typical scale ranges from several to tens of nanometers is predicted by mesoscale methods, such as thermodynamic modeling, 2,3 self-consistent field theory, [4][5][6] MesoDyn, 7 coarse grained molecular dynamics (CG MD) [8][9][10][11][12] or DPD. [13][14][15][16][17][18][19][20][21][22][23][24][25][26] Then, the diffusion of water and protons is considered in a static structure obtained from mesoscale modeling. 25,27 Alternatively, water and ion diffusion is modeled in pre-determined "ideal" environments such as cylindrical channels 28,29 and then the results for ideal pores are extrapolated onto the macroscopic system using pore network models to predict overall transport properties 30,31 or to obtain insights on pore structure and diffusion mechanisms from comparison of simulation results with experiments.…”

“…37,[41][42][43][44][45][46][47][48] DPD models with electrostatics considered implicitly were applied to other polyelectrolytes, such as sulfonated poly(phenylene) sulfone (sPSO2) at different sulfonation levels (SLs), 49 SPEEK, 50 and grafted copolymers with varying type and the attachment of the side chain. 17 Explicit treatment of electrostatic interactions in DPD was enabled by the development of smeared charge models. In this approach, the charge is distributed around the bead center: linear, 51 Slater-type exponential, 52 Gaussiantype, 53 and Bessel-type 54 distributions of charge density were employed in the literature.…”