Juan J. De Pablo | Searcharxiv

arXiv2022

Ionic transport in electrostatic Janus Membranes. An explicit solvent molecular dynamic simulation

Janus --or two-sided, charged membranes offer promise as ionic current rectifiers. In such systems, pores consisting of two regions of opposite charge can be used to generate a current from a gradient in salinity. The efficiency of \textcolor{black}{nanoscale} Janus pores increases dramatically as their diameter becomes smaller. However, little is known about the underlying transport processes\textcolor{black}{, particularly under experimentally accessible conditions}. In this work, we examine the molecular basis for rectification in Janus nanopores using an applied electric field. Molecular simulations with explicit water and ions are used to examine the structure and dynamics of all molecular species in aqueous electrolyte solutions. \textcolor{black}{For several macroscopic observables, the results of such simulations are consistent with experimental observations on asymmetric membranes. Our analysis reveals a number of previously unknown features, including a pronounced local re-orientation of water molecules in the pores, and a segregation of ionic species that has not been anticipated by previously reported continuum analyses of Janus pores. Using these insights, a model is proposed for ionic current rectification in which electric leakage at pore entrance controls net transport.

physics.chem-ph↗

arXiv2020

Machine learning active-nematic hydrodynamics

Hydrodynamic theories effectively describe many-body systems out of equilibrium in terms of a few macroscopic parameters. However, such hydrodynamic parameters are difficult to derive from microscopics. Seldom is this challenge more apparent than in active matter where the energy cascade mechanisms responsible for autonomous large-scale dynamics are poorly understood. Here, we use active nematics to demonstrate that neural networks can extract the spatio-temporal variation of hydrodynamic parameters directly from experiments. Our algorithms analyze microtubule-kinesin and actin-myosin experiments as computer vision problems. Unlike existing methods, neural networks can determine how multiple parameters such as activity and elastic constants vary with ATP and motor concentration. In addition, we can forecast the evolution of these chaotic many-body systems solely from image-sequences of their past by combining autoencoder and recurrent networks with residual architecture. Our study paves the way for artificial-intelligence characterization and control of coupled chaotic fields in diverse physical and biological systems even when no knowledge of the underlying dynamics exists.

cond-mat.soft↗