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.

Regenerable Polyelectrolyte Membrane for Ultimate Fouling Control in Forward Osmosis.

This study demonstrated the feasibility of using regenerable polyelectrolyte membranes to ultimately control the irreversible membrane fouling in a forward osmosis (FO) process. The regenerable membrane was fabricated by assembling multiple polyethylenimine (PEI) and poly(acrylic acid) (PAA) bilayers on a polydopamine-functionalized polysulfone support. The resulting membrane exhibited higher water flux and lower solute flux in FO mode (with the active layer facing feed solution) than in PRO mode (with the active layer facing draw solution) using trisodium citrate as draw solute, most likely due to the unique swelling behavior of the polyelectrolyte membrane. Membrane regeneration was conducted by first dissembling the existing PEI-PAA bilayers using strong acid and then reassembling fresh PEI-PAA bilayers on the membrane support. It was found that, after the acid treatment, the first covalently bonded PEI layer and some realigned PAA remained on the membrane support, acting as a beneficial barrier that prevented the acid-foulant mixture from penetrating into the porous support during acid treatment. The water and solute flux of the regenerated membrane was very similar to that of the original membrane regardless of alginate fouling, suggesting an ultimate solution to eliminating the irreversible membrane fouling in an FO process. With a procedure similar to the typical membrane cleaning protocol, in situ membrane regeneration is not expected to noticeably increase the membrane operational burden but can satisfactorily avoid the expensive replacement of the entire membrane module after irreversible fouling, thereby hopefully reducing the overall cost of the membrane-based water-treatment system.