Two-dimensional (2D) nanofluidic membranes have demonstrated extraordinary promise in high-throughput ionic transport systems for water-energy nexus applications. Nevertheless, the inability to engineer macroscopic, structurally ordered 2D membranes with precisely aligned nanochannels severely restricts their practical implementation. Here, we demonstrate a scalable shear-flow alignment strategy to assemble MXene/ANF composite membranes with vertically ordered nanochannels, which outperforms ionic flux leading to giant osmotic energy generation. The viscosity contrast between components generates sufficient shear-force gradients in dilute suspensions, enabling parallel alignment of MXene nanosheets during vacuum-assisted assembly. Under a 50-fold salinity gradient, the resulting vertically ordered architecture establishes ultralow-tortuosity ion highways, achieving unprecedented ionic flux (2.4 × 105 S/m2) with near-perfect cation selectivity (0.94), leading to a power density of 81.8 W m−2 under a 500-fold concentration gradient (5 M/0.01 M NaCl). Remarkably, the MXene/ANF nanocomposite simultaneously achieves exceptional mechanical robustness (tensile strength: 348.9 MPa, Young's modulus: 19.9 GPa) and environmental stability (97.5% performance retention after 30-day operation). Theoretical simulations demonstrate that highly aligned channels can significantly reduce resistance and energy loss during ion transmembrane process. The combination of flow-directed assembly and nanoscale interface engineering presented here provides new insights into scalable fabrication of high-performance ion exchange membranes for sustainable energy applications.
Wang and colleagues recently reported that lipid nanoparticle (LNP) administration promotes experimental pulmonary tumor metastasis through TLR9-MyD88- and cGASSTING-mediated neutrophil extracellular trap (NET) formation, invoking mitochondrial DNA (mtDNA) released from necrotic muscle cells as the activating ligand. This Opinion challenges a fundamental assumption underlying that interpretation: because TLR9 recognizes unmethylated CpG dinucleotides shared equally by mtDNA and bacterial DNA, and cGAS detects double-stranded DNA in a sequence-independent manner, the genetic rescue experiments are equally consistent with bacterial nucleic acid or lipopolysaccharide (LPS) contamination of the LNP preparations. Standard endotoxin detection methods, including the Limulus amebocyte lysate assay, are demonstrably unreliable in lipid matrices, where LPS lipid A inserts stably into hydrophobic domains and evades detection. Total lipid A quantification by EndoQuant (acid hydrolysis/HPLC-MS/MS), 16S rDNA analysis, polymyxin B pretreatment, and amplicon sequencing are required to resolve ligand identity. Until these controls are applied, the mtDNA hypothesis and a manufacturing contamination hypothesis remain equally supported by the available data — a distinction with profoundly different regulatory and clinical implications for LNP-based therapeutics.
