Combined atomistic simulations and machine learning for probing pressure-driven transport and resulting electrokinetic effects in cationic-polymer-chain-grafted nanochannels

Pressure-driven transport in charged nanofluidic channels have known to generate most remarkable electrokinetic effects such as streaming potential and electroviscous effects. In a recent study, our group demonstrated that in anionic-polymer-chain-grafted nanochannels, coion-driven induced electroosmotic transport may lead to electroslippage effect (where the resulting flow increases instead of being reduced). In this study, we investigate the same pressure-driven transport and induced electrokinetic effects in cationic-polymer-chain-grafted nanochannels. The specifics of the cationic polymer are captured through all-atom molecular dynamic simulations, while the resulting mechanism of the fluid flow (a combination of pressure-driven and induced electroosmotic transport) is analyzed by machine learning (e.g., linear discriminant analysis) approach. The findings reveal highly interesting interplay of the brush deformation and counterion localization (as a function of the strength of the pressure-gradient driving the flow) in dictating the overall liquid transport and the induced electrokinetic effects.