Predicting Extraction Selectivity of Acetic Acid in Pervaporation by Machine Learning Models with Data Leakage Management

The extraction of acetic acid and other carboxylic acids from water is an emerging separation need as they are increasingly produced from waste organics and CO₂ during carbon valorization. However, the traditional experimental approach can be slow and expensive, and machine learning (ML) may provide new insights and guidance in membrane development for organic acid extraction. In this study, we collected extensive literature data and developed the first ML models for predicting separation factors between acetic acid and water in pervaporation with polymers' properties, membrane morphology, fabrication parameters, and operating conditions. Importantly, we assessed seed randomness and data leakage problems during model development, which have been overlooked in ML studies but will result in over-optimistic results and misinterpreted variable importance. With proper data leakage management, we established a robust model and achieved a root-mean-square error of 0.515 using the CatBoost regression model. In addition, the prediction model was interpreted to elucidate the variables' importance, where the mass ratio was the topmost significant variable in predicting separation factors. In addition, polymers' concentration and membranes' effective area contributed to information leakage. These results demonstrate ML models' advances in membrane design and fabrication and the importance of vigorous model validation.

Enhanced Pervaporation Properties of PVA-Based Membranes Modified with Polyelectrolytes. Application to IPA Dehydration

In this work, dense and supported pervaporation polyvinyl alcohol (PVA)-based membranes modified with poly(allylamine hydrochloride) (PAH) and poly(sodium 4-styrenesulfonate)(PSS)/PAH top nanolayers were synthesized. Two main points were investigated: the role of the polyelectrolyte PAH on water selectivity of the selective polymer matrix and the impact of the porous substrate based on polyacrylonitrile (PAN) and aromatic polysulfone amide (UPM-20®), used to get supported high-performance membranes. Various methods of analysis (fourier-transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), atomic force microscopy (AFM), small-angle X-ray scattering (SAXS), porosity, contact angles, ultrafiltration) were applied to study the developed membranes. Transport characteristics of the developed membranes were studied in isopropanol dehydration by pervaporation. Obtained results are discussed in the light of the structure and physicochemical characteristics of these PVA/PAH membranes and the types of porous substrate. It was shown that the PAN-supported membrane with the selective layer based on PVA/PAH modified by 10 polyelectrolyte PSS/PAH bilayers possessed ~4.5 times higher permeation flux with the same high selectivity level (99.9 wt % water in the permeate) for the dehydration of the isopropanol (20 wt % water) at 60 °C compared to the commercial analog PERVAPTM 1201.

PEGylated polyvinylidene fluoride membranes via grafting from a graphene oxide additive for improving permeability and antifouling properties

Polyvinylidene fluoride (PVDF) porous membranes with enhanced hydrophilicity and antifouling performance were developed via surface PEGylation (PEG, polyethylene glycol) via a reactive graphene oxide (GO) additive. PVDF/GO blended membranes were first fabricated via a non-solvent-induced phase separation process. Then the carboxyl groups of GO sheets immobilized on the membrane surface acted as initiating sites for grafting amine-functionalized PEG (PEG-NH₂) chains via an amination reaction. Analysis of the X-ray photoelectron spectroscopy and Fourier transform infrared spectroscopy-attenuated total reflectance results confirmed the successful grafting of hydrophilic PEG molecular chains on PVDF membrane surfaces. The water contact angle of the PEGylated PVDF membrane decreased to 59.9°, indicating improved hydrophilicity. As a result, the antifouling performance was enhanced significantly. After surface PEGylation, the flux recovery rate is reached 90.2%, the total fouling ratio was as low as 20.7%, and reversible fouling plays a dominant role during the membrane fouling process. This work provides a valuable strategy to fabricate PEGylated membranes via the introduction of a reactive GO additive.

Polyvinylidene fluoride (PVDF) porous membranes with enhanced hydrophilicity and antifouling performance were developed via surface PEGylation (PEG, polyethylene glycol) via a reactive graphene oxide (GO) additive.

Development and Characterization of New Pervaporation PVA Membranes for the Dehydration Using Bulk and Surface Modifications

In the present work, the novel dense and supported membranes based on polyvinyl alcohol (PVA) with improved transport properties were developed by bulk and surface modifications. Bulk modification included the blending of PVA with chitosan (CS) and the creation of a mixed-matrix membrane by introduction of fullerenol. This significantly altered the internal structure of PVA membrane, which led to an increase in permeability with high selectivity to water. Surface modification of the developed modified dense membranes, based on composites PVA-CS and PVA-fullerenol-CS, was performed through (i) making of a supported membrane with a thin selective composite layer and (ii) applying of the layer-by-layer assembly (LbL) method for coating of nano-sized polyelectrolyte (PEL) layers to increase the membrane productivity. The nature of polyelectrolyte type-(poly(allylamine hydrochloride) (PAH), poly(sodium 4-styrenesulfonate) (PSS), poly(acrylic acid) (PAA), CS), and number of PEL bilayers (2⁻10)-were studied. The structure of the composite membranes was investigated by FTIR, X-ray diffraction, and SEM. Transport properties were studied during the pervaporation separation of 80% isopropanol⁻20% water mixture. It was shown that supported membrane consisting of hybrid layer of PVA-fullerenol (5%)⁻chitosan (20%) with five polyelectrolyte bilayers (PSS, CS) deposited on it had the best transport properties.