(149i) Designing Functionalized 2D Material Membranes for Seawater Desalination: New Insights from Ab Initio Force Fields

Ultrathin membranes composed of two-dimensional (2D) materials exhibit exceptional properties, including high specific surface areas and superior separation efficiency, making them promising candidates for seawater desalination and osmotic power harvesting. Among these materials, molybdenum disulfide (MoS₂), hexagonal boron nitride (hBN), and graphene oxide (GO) stand out due to their stability and excellent desalination performance. While recent studies highlight the potential of 2D materials for seawater desalination to produce drinking water, molecular dynamics (MD) simulations have primarily focused on unrealistic bare nanopore edges in MoS₂ and hBN. This major limitation arises from the lack of force fields that can accurately describe functionalized nanopores, where interactions between water molecules and edge atoms (i.e., B and N in hBN and Mo and S in MoS₂) enable functionalization with hydrogen (H), oxo (O), and hydroxyl (OH) groups.

To address this gap, we employ density functional theory (DFT)-based ab initio molecular dynamics (AIMD) simulations to investigate the functionalization tendencies of hBN and MoS₂ nanopores in aqueous environments. Our results reveal a strong tendency for hydrogenation at S sites and hydroxylation at Mo sites in MoS₂, while boron edges in hBN favor hydrogen and hydroxyl functionalization, and nitrogen edges exhibit hydrogenation and occasional oxygen functionalization. Our study thus establishes functionalization as a key tuning parameter for 2D membranes. Additionally, we demonstrate the role of the Grotthuss mechanism in the functionalization process of MoS₂ and hBN edges in water.

Moving further, to enable realistic MD simulations of functionalized 2D membranes, we developed high-fidelity force fields for H-, O-, and OH-functionalized MoS₂ and hBN using potential energy surfaces from DFT calculations. These force fields enable stable MD simulations of water and ion transport through functionalized nanopores. Our findings indicate that previous studies, which considered only bare nanopores, likely overestimated water flux and underestimated ion rejection in MoS₂ and hBN membranes. Furthermore, we demonstrate that functional groups modulate charge distributions at nanopore edges, allowing for highly selective membranes for salt and boron rejection and revealing edge charges as a new paradigm for designing 2D material-based membranes.

Finally, while GO-based membranes are extensively studied for water purification, they face challenges such as nanochannel swelling due to water intercalation and membrane disintegration under high pressure, limiting their scalability. We addressed these issues by investigating aligned GO-based liquid crystals via a sequential interpenetrating polymeric network (IPN) employing electrostatic anchorage. Work by experimental collaborators showed that these highly ordered and structurally robust membranes exhibited high water flux and long-term separation efficiency for monovalent and divalent salts, dyes, and antibiotics. Our detailed MD simulations of GO interacting with the polymer network unraveled the underlying mechanisms of extreme ion retention in the membrane, thus demonstrating nanocomposite membranes as a highly scalable alternative to nanoporous membranes. Overall, our comprehensive in silico investigation establishes a framework for accurate modeling of functionalized 2D membranes, paving the way for the predictive design of highly efficient membranes for sustainable water treatment applications.