Membrane technologies are nowadays widely used; especially various types of filtration or reverse osmosis in households, desalination plants, pharmaceutical applications etc. Facing water pollution, they are also applied to eliminate emerging contaminants from water. Incomplete knowledge directs the composition of membranes towards more and more dense materials known for their higher selectivity compared to porous constituents. This paper evaluates advantages and disadvantages of well-known membrane materials that separate on the basis of particle size, usually exposed to a large amount of water, versus dense hydrophobic membranes with target transport of emerging contaminants through a selective barrier. In addition, the authors present several membrane processes employing the second type of membrane.

Ibuprofen separation from water by adsorption and pertraction processes has been studied, comparing 16 different membranes. Tailor-made membranes based on Matrimid, Ultem, and diaminobenzene/diaminobenzoic acid with various contents of zeolite and graphene oxide, have been compared to the commercial polystyrene, polypropylene, and polydimethylsiloxane polymeric membranes. Experimental results revealed lower ibuprofen adsorption onto commercial membranes than onto tailor-made membranes (10–15% compared to 50–70%). However, the mechanical stability of commercial membranes allowed the pertraction process application, which displayed a superior quantity of ibuprofen eliminated. Additionally, the saturation of the best-performing commercial membrane, polydimethylsiloxane, was notably prevented by atomic layer deposition of (3-aminopropyl)triethoxysilane.

The hydrophilic and hydrophobic single-walled carbon nanotube membranes were prepared and progressively applied in sorption, filtration, and pertraction experiments with the aim of eliminating three antibiotics—tetracycline, sulfamethoxazole, and trimethoprim—as a single pollutant or as a mixture. The addition of SiO₂ to the single-walled carbon nanotubes allowed a transparent study of the influence of porosity on the separation processes. The mild oxidation, increasing hydrophilicity, and reactivity of the single-walled carbon nanotube membranes with the pollutants were suitable for the filtration and sorption process, while non-oxidized materials with a hydrophobic layer were more appropriate for pertraction. The total pore volume increased with an increasing amount of SiO₂ (from 743 to 1218 mm³/g) in the hydrophilic membranes. The hydrophobic layer completely covered the carbon nanotubes and SiO₂ nanoparticles and provided significantly different membrane surface interactions with the antibiotics. Single-walled carbon nanotubes adsorbed the initial amount of antibiotics in less than 5 h. A time of 2.3 s was sufficient for the filtration of 98.8% of sulfamethoxazole, 95.5% of trimethoprim, and 87.0% of tetracycline. The thicker membranes demonstrate a higher adsorption capacity. However, the pertraction was slower than filtration, leading to total elimination of antibiotics (e.g., 3 days for tetracycline). The diffusion coefficient of the antibiotics varies between 0.7–2.7 × 10⁻¹⁰, depending on the addition of SiO₂ in perfect agreement with the findings of the textural analysis and scanning electron microscopy observations. Similar to filtration, tetracycline is retained by the membranes more than sulfamethoxazole and trimethoprim.

Pharmaceutical wastewater pollution has reached an alarming stage, as many studies have reported. Membrane separation has shown great performance in wastewater treatment, but there are some drawbacks and undesired byproducts of this process. Selective membranes could be used for pollutant investigation sensors or even for pollutant recovery. The polydimethylsiloxane (PDMS) membrane was first tested on separated and mixed antibiotic (ATB) water solutions containing sulfamethoxazole (SM), trimethoprim (TMP), and tetracycline (TET). Then, the bare and ultra-violet grafted (UV-grafted) PDMS membranes (MMA-DMAEMA 10, GMA-DMAEMA 5, and GMA-DMAEMA 10) were tested in tramadol (TRA) separation, where the diffusion coefficient was evaluated. Finally, the membranes were tested in pertraction with a mixture of SM, TMP, TET, and TRA. The membranes were characterized using the following methods: contact angle measurement, FTIR, SEM/EDX, and surface and pore analysis. The main findings were that TET was co-eluted during mixed ATB pertraction, and GMA-DMAEMA 5 was found to selectively permeate TRA over the present ATBs.