Answers are grounded in this paper and cite their sources.
Controlling polymer network architecture in nanofiltration membranes at the molecular scale remains a central challenge, as conventional interfacial polymerization yields structurally heterogeneous films with limited tunability of free-volume pathways. Here, we demonstrate that positional isomerism with varied substitution of hydroxyl groups in dihydroxybenzene while preserving atomic composition. This acts as molecular programming tool to encode crosslinking density, rigidity, and transport pathways in polyester TFC membranes using catechol, resorcinol, and hydroquinone; ortho/meta/para with an angular separation θ ≈ 60°, 120°, 180°. We fabricated ultrathin selective layers (11–25 nm) of positional isomers with trimesoyl chloride that exhibit markedly different water permeances (63 ± 1 to 92 ± 2 Lm −2 h −1 bar −1 ) despite converging on nearly identical molecular weight cut-offs (∼500–520 gmol -1 ). Buried volume (%V Bur ) analysis and steric mapping reveal that increasing angular separation progressively relieves reactive site occlusion from %V Bur = 64.4% (catechol, with intramolecular H-bonding) through 66.7% (resorcinol) to a markedly reduced 53.7% (hydroquinone, free volume ≈ 46.3%). This directly programs crosslinking density ( poly -Res ∼81% ≈ poly -Cat ∼82% > poly -Hyd ∼56%), glass transition temperature (90.2 °C > 66.1 °C > 46.7 °C), and mechanical stiffness (2.5, 0.2, 1.6 GPa). The para-substituted architecture ( poly -Hyd) generates the most interconnected free-volume transport network, delivering significantly higher permeance without compromising solute selectivity. All membranes achieve NaCl/Congo red separation factors of 1.3-1.6 × 10 3 , surpassing state-of-the-art nanofiltration membranes. These findings establish a compositionally invariant design principle: shifting from pore-size engineering to pathway topology programming enables independent control over transport resistance and molecular sieving within a single polymer family.
No takes yet. Share an insight, caveat, or question.
Amrutha et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: