Removal of pharmaceutical compounds from wastewater by adsorption on commercial materials and molecularly imprinted polymers

Removal of pharmaceutical compounds from wastewater by adsorption on commercial materials and molecularly imprinted polymers D. Frascari*, E. Girometti*, L. Sisti*, E. Savigni*, F. Benstoem**, D. Pinelli* * University of Bologna, Department of Civil, Chemical, Environmental and Materials Engineering, Via Terracini 28, 40131 Bologna, Italy (E-mail: [email protected]; [email protected]; laura.sis[email protected]; [email protected]; davide.p[email protected]) ** ATD GmbH, Krefelder Straße 147, D-52070 Aachen, Germany (E-mail: [email protected] ) Abstract The main goal of this research was to develop a continuous flow adsorption/desorption process of pharmaceutical compounds from municipal wastewater, comparing the performances obtained with activated carbon, a polymeric neutral resin and several molecularly imprinted polymers (MIPs). The target micropollutants were carbamazepine, diclofenac and ibuprofen. After an initial screening of several sorbents based on isotherm tests, an activated carbon (Norit 1240 W), a polymeric resin (XAD16N) and a MIP were tested in continuous flow tests of carbamazepine adsorption. XAD16N resulted the best material, with 18360 bed volumes treated at the 20% breakpoint and a 12.4 mgCBZ/gdry resin sorption capacity, at a 3-min EBCT. In continuous flow desorption tests conducted with carbamazepine, diclofenac and ibuprofen, the best material resulted again XAD16N, with the attainment of a 90% desorption after 10 to 17 bed volumes of solvent, depending on the target micropollutant. Among the tested solvents, ethanol performed better than methanol and 2-propanol. These results indicate that XAD16N is a very promising material for the removal of pharmaceuticals, and that an in-situ desorption of pharmaceuticals can be effectively performed with a very limited amount of ethanol. Keywords Activated carbon; adsorption; molecularly imprinted polymers; pharmaceuticals; wastewater treatment. INTRODUCTION The presence of pharmaceutical contaminants in wastewater (WW) represents a significant environmental concern due to their persistence, bioaccumulation potential, and adverse effects on ecosystems and human health (Fawzi et al., 2023). These contaminants include analgesics, antibiotics and hormones. They often resist conventional wastewater treatment processes, necessitating the development of advanced methods for their removal. Adsorption is a promising technique for this purpose, thanks to its efficiency and cost-effectiveness (Vinayagam, et al., 2022). This work targets the removal of diclofenac (DCF) and ibuprofen (IBU), two widespread anti-inflammatory drugs, and carbamazepine (CBZ), a widely used anticonvulsant and analgesic. Isotherms and continuous flow tests of both adsorption and desorption were utilized to compare the performances of activated carbon, the most commonly used material for the removal of micropollutants from wastewater, to those of a polymeric resin and molecularly imprinted polymers (MIPs). MATERIALS AND METHODS Adsorbent materials and analytical methods. The tested commercial adsorbent materials are Amberlite XAD16N, a non-ionic styrene-divinylbenzene adsorbent, and activated carbon Norit GAC 1240 W. MIPs were synthetized by bulk polymerization as described previously (Girometti et al., 2024). Several MIPs were produced for each target pharmaceutical by combining different monomers and cross-linkers. Each MIP was labelled with the acronym MIP_J_X_Y, where: J indicates the target pharmaceutical (CBZ, DCF or IBU), X indicates the monomer (methacrylic acid (MAA), 2-vinylpyridine (2VP), 2-hydroxyethyl methacrylate (HEMA), methacrylamide (MAAM), itaconic acid (ITA)eugenol (EUG)), Y indicates the cross-linker (a difunctional one that is ethylene glycol methacrylate (EG), and a trifunctional one that is trimethylolpropane triacrylate (TRIM)). CBZ, DCF and IBU concentrations were determined by a Waters UPLC-MS. The UPLC-MS method employs an Acquity UPLC BEH C18 column and a mobile phase composed of 60:40 water:acetonitrile (v/v) and 0.1 vol% formic acid. Adsorption isotherms and continuous flow adsorption and desorption tests The batch isotherm tests were performed with real wastewater treatment plant (WWTP) effluent, spiked with CBZ, DCF or IBU to reach the desired concentrations. The conditions were: sorbent concentration 1 g/L, temperature 22 °C, contact time 24 hours. For CBZ, 4 continuous flow adsorption breakthrough (BT) tests were conducted with actual WWTP effluent: BT1, sorbent Norit GAC 1240 W, empty bed contact time (EBCT) 23 min, bed height 40 cm, particle size 0.7 – 1.4 mm; BT2, sorbent Norit GAC 1240 W, EBCT 3.1 min, bed height 21 cm, particle size 0.35 – 0.71 mm; BT3, sorbent XAD 16N, EBCT 3.0 min, bed height 18 cm, particle size 0.35 – 0.71 mm; BT4, sorbent MIP CBZ_MAA_TRIM, EBCT 3.5 min, bed height 20 cm, particle size 0.35 – 0.71 mm. In BT1 the inlet CBZ concentration was 0.90 mg/L, whereas in BT1, BT2 and BT3 it was 0.20 mg/L. Continuous flow desorption tests were conducted by bringing an 8-cm packed bed of sorbent in equilibrium with 1.5 L of WWTP effluent spiked with CBZ, DCF and IBU. The packed bed was then flushed with ethanol, methanol or 2-propanol, at a 14-min EBCT. RESULTS AND DISCUSSION Isotherms Fig. 1a shows the CBZ adsorption isotherms. With a linear sorption constant of 73 L/g, Norit activated carbon performed best, followed by one of the MIPs (MIP_CBZ_MAA_TRIM) with 65 L/g and XAD16N (44 L/g). These 3 sorbents were, therefore, selected for further evaluation through continuous flow adsorption tests conducted - in the initial phase of the research - only with CBZ. The other MIPS performed significantly worse. MIPs with trifunctional cross-linkers outperformed the other ones with a bifunctional one. Figure 1. Adsorption isotherms relative to CBZ (a) and DCF (b). The DCF isotherms are illustrated in Fig. 1b. Again, Norit resulted superior to the other materials, with a linear constant of 14 L/g, followed by a MIP (MIP_DCF_2VP_TRIM) with 4.7 L/kg and by XAD16N with 4.2 L/kg. The IBU adsorption isotherms, not shown in Fig. 1, yielded similar results: with a linear sorption factor of 8.9 L/g, Norit outperformed the other materials, followed by the XAD16N (2.3 L/g). Based on these results, it was decided to further investigate the process through continuous flow adsorption/desorption tests, conducted with only CBZ in the initial phase, with Norit, XAD16N and MIP_CBZ_MAA_TRIM. Continuous flow adsorption/desorption tests Tests BT1, conducted with Norit GAC 1240 W at a 23-min EBCT represents the benchmark condition, since the adsorption of micropollutants from WW or groundwater is typically conducted with activated carbon at 20-30 min EBCTs (Benstoem et al., 2017). The results are shown in Fig. 2 in terms of normalized effluent concentrations, defined as (CBZ effluent concentration) / (CBZ inlet concentration). After 9 months of monitoring, BT1 was stopped, in correspondence with a 4% normalized breakpoint (BP). BT1 confirmed the effectiveness of Norit 1240 W for pharmaceutical removal, with the treatment of 15190 bed volumes (BVs) at the 4% BP and a 34 mgCBZ/gdry resin sorption capacity. Test BT2 was conducted again with Norit GAC 1240 W at a significantly lower EBCT (3.1 min), in an effort to reduce the column size and the amount of sorbent required in full-scale applications. As shown in Fig. 2, as expected, the 7.7-fold reduction in EBCT resulted in a considerably lower performance, with the attainment of the 4% BP after the treatment of about 5000 BVs, instead of 15190 in BT1. At the same time, BT2 reached the 20% BP at 18690 BVs. This is considered a remarkable result, considering that in most pilot-plant applications the 20% BP is reached after 5000-15000 BVs, at EBCTs typically > 10 min (Benstoem et al., 2017). On the basis of this encouraging result, Norit was compared to XAD16N (test BT3) and MIP_CBZ_MAA_TRIM (test BT4) maintaining the EBCT in the 3-3.5 min range. The results of these 3 tests are shown in Fig. 2 in terms of normalized effluent concentrations, and in Table 1 in terms of operational conditions and main performance parameters evaluated at a 20% normalized BP. Figure 2. Adsorption breakthrough curves relative to tests BT1, BT2, BT3, BT4. Experimental data and bestfitting simulations performed with Aspen adsorption. Even though XAD16N (BT3, EBCT 3 min) and Norit (BT2, EBCT 3.1 min) featured similar BVs of treated WW at the 20% BP (about 18500), XAD16N (BT3) resulted in a delayed breakthrough curve, leading to significantly better performances in terms of capacity at 20% BP. Conversely, the selected MIP (CBZ_MAA_TRIM; BT4, EBCT 3.5 min) resulted in inferior performances, with just 270 BVs at the 20% BP and a CBZ adsorption capacity of 0.19 mgCBZ/gdry resin, remarkably lower than that obtained in the corresponding isotherm at 0.2 mg/L in the liquid. The last part of the work was dedicated to investigating the feasibility of a chemical desorption/regeneration process after CBZ, DCF and IBU adsorption on XAD16N or Norit, and to identifying the most suitable solvent. To this purpose, ethanol, methanol and 2-propanol were compared. For XAD16N, ethanol resulted the most effective solvent, featuring a 90% desorption after a number of BVs ranging between 10 (CBZ and IBU) and 17. For Norit, 30 to 40 BVs of either ethanol, methanol or 2-propanol resulted necessary in order to achieve desorption rates in the 20-70% range for IBU and CBZ and in the 0-10% range for DCF. Table 1. Operational conditions and main performance parameters relative to adsorption tests BT2, BT3, and BT4, conducted with WWTP effluent spiked with CBZ at 0.2 mg/L. Performance parameters were evaluated at a 20% normalized breakpoint (BP). Test ID BT2 BT3 BT4 Sorbent material NORIT 1240 W XAD16N MIP CBZ MAA TRIM EBCT (min) 3.1 3.0 3.5 Superficial velocity (m/h) 4.0 3.5 3.4 Treated BVs at BP (-) 18690 18360 270 Adsorption capacity (mgCBZ/gdry resin) 7.1 12.4 0.19 CONCLUSIONS This work showed the effectiveness of both Norit 1240 W and XAD16N in removing CBZ, IBU and DCF from real WWTP effluent. In tests of continuous flow adsorption conducted with Norit, a 7.7-fold decrease in EBCT (from 23 to 3 min) determined a decrease in adsorption performance. At the same time, the tests conducted at a 3-min EBCT with both the polymeric resin XAD16N and Norit resulted in interesting performances, with the attainment of the 20% BP after the treatment of about 18500 BVs. In these tests, XAD16N outperformed Norit, thanks to a higher adsorption capacity (12.4 mgCBZ/gdry resin) and yield (96.1%). In the continuous flow desorption tests, conducted with CBZ, IBU and DCF, the best material resulted again XAD16N, with the attainment of a 90% desorption after 10 to 17 BVs of solvent, depending on the target micropollutant. Among the tested solvents, ethanol performed better than methanol and 2-propanol. Conversely, in the tests conducted with Norit, all the tested solvents resulted in poor desorption performances for CBZ, IBU and DCF. These results point to XAD16N as a promising material for the adsorption of pharmaceuticals, paving the way for a relevant reduction in column size, amount of sorbent required and volume of regeneration solvent consumed in full-scale processes, in comparison to the benchmark process with activated carbon. Acknowledgment. This work received funding from the European Union’s Horizon Europe research and innovation program under grant agreement No 101082048 (MAR2PROTECT project). List of acronyms. BP, breakpoint. BT, breakthrough. BVs, bed volumes. CBZ, carbamazepine. DCF, diclofenac. EBCT, empty bed contact time. EG, ethylene glycol methacrylate. HEMA, 2-hydroxyethyl methacrylate. IBU, ibuprofen. ITA, itaconic acid. MAA, methacrylic acid. MAAM, methacrylamide. MIP, molecularly imprinted polymer. TRIM, trimethylolpropane triacrylate. 2VP, 2-vinylpyridine. WW, wastewater. WWTP, wastewater treatment plant. REFERENCES Benstoem, F., Nahrstedt, A., Boehler, M., Knopp, G., Montag, D., Siegrist, H., Pinnekamp, J., 2017 Performance of granular activated carbon to remove micropollutants from municipal wastewater - A metaanalysis of pilotand large-scale studies. Chemosphere 185, 105–118. Fawzi, O., Khasawneh, S., Palaniandy, P., 2021 Occurrence and removal of pharmaceuticals in wastewater treatment plants. Process Safety and Environmental Protection 150, 532–556. Vinayagam, V., Murugan, S., Kumaresan, R., Narayanan, M., et al. 2022. Sustainable adsorbents for the removal of pharmaceuticals from wastewater: A review. Chemosphere 300,134597. Girometti, E., Maggetti, C., Frascari, D., Pinelli, D., Sisti, L., Savigni, E. Development of an Adsorption Process for the Removal of Pharmaceuticals from Wastewater Treatment Plant Effluents by Means of Molecularly Imprinted Polymers and Commercial Adsorbents. (2024) Lecture Notes in Civil Engineering, 524 LNCE, pp. 263-268.