2.2. Methods

First, the corresponding assay solution is prepared. After introducing a known volume of this solution into the feeding tank, the digital balance and computer are turned on. Once the main nitrogen valve is opened, the experimental unit is powered on, and the pump is started. Then, using the needle valves and pressure gauge, the feed pressure is adjusted to 5 bar. When the pressure stabilizes, as observed on the computer, the stopwatch is started, and permeate samples are taken every 3 min.

After taking two samples at 5 bar pressure, it is increased to 10 bar. In each assay, the feed pressure is varied at 5, 10, 15, and 20 bars, and two permeate samples are collected for each pressure. At 20 bars, samples are taken every 2 min due to the high flow rate.

At the end of the assay, a sample is taken from the feeding tank using the three-way valve. As the tank has been receiving the reject stream, the concentration of the sample will be higher than the initial concentration. Finally, the collected permeate, reject, and feeding solution samples are analysed using either the conductivity metre or the spectrophotometer, depending on the specific procedure. The tests were carried out in duplicate, and each one lasted 20–30 min. All the experimentation was carried out with the same membrane, and when changing the experimental series, after each contaminant, a wash with distilled water was carried out to condition the membrane and eliminate the remains of the previous contaminant.

The collected samples from the experimental unit are analysed differently depending on the type of assay they originate from.

The wavelengths of maximum absorption obtained for carbamazepine, ketoprofen, and bisphenol A are 290 nm, 260 nm, and 300 nm, respectively. These λ_m values are similar to those used by other authors [19,20]. After determining the λ_m for each contaminant, different calibration curves were obtained.