Structural Evolution of Ternary Amphiphilic Block Copolymer Solvent Systems for Phase Inversion Membrane Formation

Abstract

Block copolymer (BCP)-derived asymmetric ultrafiltration membranes combine the BCP self-assembly with nonsolvent-induced phase separation (SNIPS). To understand the structural evolution in membrane top separation layers made from polyisoprene-b-polystyrene-b-poly­(4-vinylpyridine) (ISV) in dioxane (DOX) and tetrahydrofuran (THF) all the way to the final membrane, we combined solution small-angle X-ray scattering (SAXS), estimated solution concentrations and compositions upon solvent evaporation, in situ grazing-incidence SAXS (GISAXS), spin–spin relaxation time (T 2) analysis by solution 1H NMR, and scanning electron microscopy (SEM). Above the critical micelle concentration (<1 wt % ISV), solvent evaporation drives micelles with poly­(4-vinylpyridine) (P4VP) in the core across disorder-to-order and order-to-order transitions, the latter in part driven by the segregation of polyisoprene (PI)- from polystyrene (PS)-blocks. Extended to polystyrene-b-poly­(4-vinylpyridine) (SV) in dimethylformamide (DMF) and THF, results suggest that, in particular, T 2 relaxation analysis by 1H NMR is a powerful tool in analyzing which blocks form micelle core and which form corona chains. We expect insights to help develop next-generation SNIPS membranes for applications, e.g., in clean water and biopharmaceutical separations.

Structurally Asymmetric Porous Carbon Materials with Ordered Top Surface Layers from Nonequilibrium Block Copolymer Self-Assembly

“…These in situ GISAXS results obtained for our 95 kg mol –1 ISV-terpolymer-based studies are consistent with those reported by Gu et al for pure ISV cast films with varying molar masses and similar terpolymer composition . They corroborate the overall picture that order in the top surface layer of these membranes evolves from disorder to BCC and then to SC lattices, the former being consistent with solution structures at higher concentrations as observed with SAXS studies (see Figure ) and the latter being consistent with the pore structure of the final ISV membranes ( vide infra ) . The quiescent solution SAXS data in Figure was obtained by keeping the DOX/THF solvent ratio constant.…”

“…We hypothesize that in the simultaneous case, the addition of resols prior to dissolution of the ISV hinders the ISV polymer from forming homogeneously sized micelles and therefore from forming well-ordered micellar BCC lattices in solution. In the meantime, we know from a set of independent nuclear magnetic resonance (NMR) spectroscopy studies that the P4VP block rather than the PI block of ISV in 7:3 DOX/THF forms the micelle core . The reason why the presence of resols hinders highly ordered micelle lattice formation might be that they hydrogen-bond to the P4VP blocks upon dissolution early on, thereby locking the system into large structures that cannot equilibrate into well-defined micellar lattices.…”

Surface Segregation and Self‐Assembly of Block‐Copolymer Separation Layers on Top of Homopolymer Substructures in Asymmetric Ultrafiltration Membranes from a Single Casting Step

“…Although the SV micellar structure in such an amphiphilic solvent system is not trivial and still remains controversial, [ 37–39 ] our recent 1 H nuclear magnetic resonance (NMR) study of SV in tetrahydrofuran: N , N ‐dimethylformamide (THF:DMF) of 9:1 weight ratio based on spin–spin relaxation analysis supported that V is in the core and S is in the corona with a critical micelle concentration (CMC) somewhere between 0.1 and 0.5 wt%. [ 27 ] For the SV/DOX/NMP system investigated here we will assume V‐core S‐corona micelles in analogy to the SV/THF/DMF system. Note that each of χ DOX‐S and χ NMP‐S is lower than χ THF‐S and χ DMF‐S , respectively (Table 1), predicting that the DOX/NMP solvent system should prefer S‐corona micelles even more than the THF/DMF solvent system.…”

“…Note that even for pure SV membranes casted from binary solvent systems ordered self‐assembled surface were only observed when hydrophilic solvent (NMP or DMF) content was higher than 30 wt%. [ 27 ] Here, SV2 was casted from more than 98 wt% NMP‐containing solutions (DOX is less than 2 wt%). Yet resulting membranes showed self‐assembled surfaces.…”

“…The BCP self-assembly based top surface pore layer is obtained by immersing a doctor-bladed BCP-solution in a water bath after a certain amount of waiting time postdoctor blading (typically of order of a minute), which quenches the transient evolution of BCP self-assembly into a porous structure induced by the solvent evaporation. [25][26][27] BCP self-assembly minimizes pore size distribution and maximizes pore number-density at the surface, giving rise to higher resolution and flux in separation applications, respectively, as compared to conventional NIPS-derived homopolymer membranes. [28] Despite these favorable performance characteristics, a significant obstacle to the wider use of BCP-based SNIPS-derived UF membranes may be their higher costs relative to homopolymer NIPS derived conventional UF membranes.…”