Effect of an industrial chemical waste on the uptake J. Serb. Chem. Soc. 87 (0) 1–14 (2022) Original scientific paper JSCS–11718 Published 1 August 2022 1 Highly selective water-compatible molecularly imprinted polymers for benzophenone-4 MILOŠ P. PEŠIĆ1#, JUGOSLAV B. KRSTIĆ2 and TATJANA Ž. VERBIĆ1*# 1University of Belgrade – Faculty of Chemistry, Belgrade, Serbia and 2University of Belgrade – Institute of Chemistry, Technology and Metallurgy – National Institute of the Republic of Serbia, Belgrade, Serbia (Received 25 March, revised 6 May, accepted 9 May 2022) Abstract: Molecularly imprinting technology was applied for preparing selec- tive sorbents for benzophenone-4 (BP4), an organic UV filter used in sun- screens and cosmetics. Several imprinted polymers were prepared by bulk polymerization, using BP4 as template. Combination of stability (mechanical and chemical), selectivity and robustness of the imprinted polymers with BP4 properties resulted in a successful imprinting process (imprinting factors in range 1.05–2.60). The prepared polymers were characterised by infrared spec- troscopy, elemental analysis, conductometric titrations and nitrogen physisorp- tion at 77 K. Adsorption capacities and selectivity towards 7 other organic UV filters (benzophenone-3, benzophenone-8, homosalate, butyl methoxydibenzo- ylmethane, ethyl hexyl salicylate, ethyl hexyl p-dimethylamino benzoate and ethyl hexyl p-methoxycinnamate) were determined, proving high adsorption capacity and high selectivity for BP4 binding. The highest adsorption capacity was observed for 4-vinylpyridine/ethylene glycol dimethacrylate co-polymer prepared in dimethyl sulfoxide (1.108 mmol g-1). The imprinted polymer with the highest binding capacity was applied to solid phase extraction of BP4 from aqueous solutions with 98.5 % efficiency. Keywords: UV filters; imprinting factor; binding selectivity. INTRODUCTION Molecularly imprinted polymers (MIPs) are smart synthetic materials which can be used for several applications:1–4 as sorbents for solid phase extraction (SPE),5–9 as stationary phase for chromatography,10–15 as sensors,16–18 artificial enzymes,19–21 as systems for controlled drug delivery and release,22–24 etc. Molecular imprinting technology (MIT) is based on interactions between tem- plate molecule and functional monomer in the presence of porogen. The chosen * Corresponding author. E-mail: tatjanad@chem.bg.ac.rs # Serbian Chemical Society member. https://doi.org/10.2298/JSC22032540P 2 PEŠIĆ, KRSTIĆ and VERBIĆ monomer should have at least one functional group which is likely to interact with the template. During the polymerization process monomer molecules, as present in excess, surround the template molecule, forming a pre-polymerization complex. Then, in the presence of a cross-linker, a rigid polymer network is formed usually by free radical polymerization. After the template is washed out, selective binding sites complementary to template molecules in size, shape and chemical characteristics remain in the polymer, enabling the template rebinding.1 In some examples, only one monomer is used for the preparation of MIPs.25,26 The above-described approach is known as non-covalent imprinting and presents the most commonly used approach in MIT. Interactions like van der Waals, hydrogen and ionic bonding are involved in non-covalent imprinting process and, later on, in template rebinding. Thus, the higher the variety and strength of monomer-template interactions, the more selective polymer with higher binding capacity is obtained.1 Imprinted polymers have good selectivity, chemical and mechanical stab- ility; therefore, these materials can be universally applied in various fields of chemistry and biosciences.2 Although extensive research is being done within the field, and many advances have been achieved in MIT in recent decades, MIPs are generally still lacking high binding efficiency. Most of the non-covalent MIPs described in the literature can bind only a small percent of the template amount initially used during polymerization.25 Therefore, optimizing and fine tuning of polymerization conditions and the process itself to obtain highly selective poly- mers with high template binding efficiency is an on-going demand in MIT.2 The growing need for skin protection from the sun radiation has established UV filters as one of the emerging environmental contaminants.27,28 Regularly used sunscreen cosmetic products contain a high amount of UV filters (up to 15 %). UV filters can be divided into inorganic (zinc oxide and titanium oxide) and organic (benzophenones, cinnamates, dibenzoylmethanes, triazole derivatives, salicylates, etc).28 Benzophenone-type UV-filters are major ingredients of sunscreen products nowadays. According to EU regulations, BP4 or sulisobenzone (Fig. 1) and its sodium salt, BP5, can be used in sunscreens up to 5%.30 When released in the environment, several adverse effects have been noticed such as coral bleaching, estrogenic activity in fish and even endocrine disrupting effects are sus- pected.27,31,32 It is estimated that 40 % of coastal coral colonies worldwide are exposed to UV-filter pollution.28,33 BP3 can cause oxidative stress in zooxanthellae, org- anisms that live in symbiosis with corals. The decrease of zooxanthellae causes the lightening of the brown colour of the corals.27,32–34 BP4 was found in river, lake and sea water;35 the EC50 value (half maximal effective concentration) of MOLECULARLY IMPRINTED POLYMERS FOR BENZOPHENONE-4 3 BP4 is near 10 mg dm–3 in the Isochrysis galbana, Scorpaenopsella armata, Paracentrotus lividus and Mytilus galloprovincialis.27,36 OOH O SO3H Fig. 1. Structure of benzophenone-4. Although UV filters gained high attention with highly increased research interest in methods for pre-concentration and extractions from various samples with subsequent quantitation, only few papers describe MIT applied for develop- ment of UV filter imprinted polymers. Sun et al. described the application of MIP imprinted with BP2 for solid phase extraction (SPE) of four benzophenones from tap and river water. 4-vinylpyridine (4VP)/ethylene glycol dimetachrylate (EDMA) based polymer imprinted with benzophenone-2 (BP2) was successfully applied for the extraction of benzophenone-1 (BP1), BP2, benzophenone-4 (BP6) and benzophenone-8 (BP8).37 Ayadi et al. described molecularly imprinted poly- aniline on silica support for the selective adsorption of BP4 from aqueous media.38 To the best of our knowledge, no other examples of polymers imprinted using BP-type UV-filters as templates were published so far. Within this study, we used BP4 as the template molecule to prepare several imprinted polymers by bulk polymerization.39 Different polymer constituents were used: monomers (N,N-dimethylaminoethyl methacrylate-DMAEM and 4VP), cross-linkers (divinylbenzene-DVB and EDMA), porogenic solvents (ACN and DMSO). Polymers were characterized using FT-infrared spectroscopy (FTIR), nitrogen adsorption and adsorption isotherms. Binding capacity and sel- ectivity towards other UV-filters was determined. Elemental analysis (EA) and conductometric titrations were used in order to estimate total and accessible binding places. The goal was to prepare selective sorbents for BP4 with high binding capacity, which can be used as potential sorbents for BP4 extraction and pre-concentration during sample preparation. EXPERIMENTAL Apparatus Infrared spectra were recorded using Thermo Scientific Nicolet 6700 FT-IR Spectr- ometer, ATR technique. HPLC measurements were done using Agilent Technologies HPLC instrument Series 1260 with quaternary pump, on-line degasser, auto sampler and diode array detector. Conductometric titrations were done using WTWCond330i with TetraCon325 con- ductometric cell. Elemental analysis was performed by combustion analysis on a Vario EL III C,H,N,S/O elemental analyzer (Elementar Analysesysteme, GmbH, Hanau, Germany). Chro- mabond vacuum manifold (Macherey-Nagel) was used for SPE extraction. For nitrogen phys- isorption Sorptomatic 1990 Thermo Finnigan analyzer was used. MultiBio RS24 (BioSan) rotator and Microspin 12 centrifuge (BioSan) were used for conducting binding experiments. 4 PEŠIĆ, KRSTIĆ and VERBIĆ Reagents Benzophenone-4 (BP4), benzophenone-3 (BP3), ethyl hexyl salicylate (OS), ethyl hexyl p-dimethylamino benzoate (PABA-O), N,N-dimethylaminoethyl methacrylate (DMAEM), di- vinylbenzene (DVB), azobisisobutyronitrile (AIBN) were purchased from Sigma Aldrich. HPLC grade methanol and acetonitrile were purchased from Fisher Scientific. Benzophenone- 8 (BP8) was purchased from the Tokyo Chemical Industry. 4-vinylpyridine (4VP) and ethyl- ene glycol dimethacrylate (EDMA) were purchased from Acros Organics. Butyl methoxydi- benzoylmethane (AVB) and homosalate (HMS) were purchased from Merck. Inhibitor remover (Sigma Aldrich) was used to remove polymerization inhibitors from 4VP, DMAEM and EDMA (5 mL of monomer/cross-linker was passed over 250 mg of inhib- itor remover). DVB was purified using alumina, in the same way as EDMA. Commercially available BP4 is declared to have some percent of water bound. It was dried at 60 °C under vacuum, until no further change in mass was observed (color change from pale to clear lemon yellow was observed during drying). Other reagents were of analytical grade and used as received. Preparation of imprinted and non-imprinted polymers The imprinted polymers were prepared using bulk polymerization: template, monomer, cross-linker, porogen and polymerization initiator were placed in a glass vial. Upon dissol- ution, argon was purged through a pre-polymerization mixture for 5 min. AIBN was used as initiator (30 mg). Vials were tightly closed and left in an oil bath at 60 °C for 24 h. After polymerization was completed, the glass vials were broken and the polymer grounded in a mortar with pestle. The non-imprinted polymers were prepared as the imprinted ones, omitting the template presence. Composition of the prepared polymers is given in Table I. TABLE I. Composition of prepared polymers Polymer label mBP4 template / mg (n / mmol) Vmonomer / L (n / mmol) Vcross linker / mL Vporogen / mL NIP1 – DMAEM 337 (2.0) EDMA 1.888 ACN 2.730 MIP1 154.0 (0.5) DMAEM 337 (2.0) EDMA 1.888 ACN 2.730 NIP2 – DMAEM 337 (2.0) DVB 1.424 ACN 2.730 MIP2 154.0 (0.5) DMAEM 337 (2.0) DVB 1.424 ACN 2.730 NIP3 – 4VP 213 (2.0) EDMA 1.888 DMSO 2.730 MIP3 154.0 (0.5) 4VP 213 (2.0) EDMA 1.888 DMSO 2.730 NIP4 – 4VP 213 (2.0) DVB 1.424 DMSO 2.130 MIP4 154.0 (0.5) 4VP 213 (2.0) DVB 1.424 DMSO 2.130 NIP5 – 4VP 106.4 (1.0) DVB 1.424 DMSO 2.000 MIP5 308.3 (1.0) 4VP 106.4 (1.0) DVB 1.424 DMSO 2.000 Template was removed from the polymer by exhaustive washing using 2 % ammonia in methanol/water 1:1 volume ratio (process was monitored by HPLC supernatant analysis). The polymers were then washed several times with methanol and dried in an oven at 60 °C. Binding experiments Batch binding experiments were performed as follows: polymer 10.0±0.2 mg was weighed in a micro tube, and 2.000 mL of template solution in acetonitrile was added. The concentrations of template solution were 0.1–10 mmol dm-3 (0.1, 0.2, 0.5, 1.0, 2.0, 4.0, 6.0, MOLECULARLY IMPRINTED POLYMERS FOR BENZOPHENONE-4 5 8.0 and 10.0). Micro tubes were shaken using a rotator for 60 min, centrifuged (10 min, 14,500 rpm) and template concentration in supernatant was determined by HPLC. The adsorption capacity was determined according to: 0 f( )c c V Q m − = (1) where Q stands for adsorption capacity (μmol g-1), c0 for total concentration of template at the beginning (mM), cf – free concentration of template after binding (mmol dm -3), V – volume of template solution used for binding experiment (mL) and m – mass of the polymer used for binding experiment (g)25. The imprinting factors were determined according to:40 ( ) ( ) bound free MIP bound free NIP / / n n IF n n = (2) where IF stands for imprinting factor, nbound / μmol – amount of the template bound to NIP/ /MIP and nfree / μmol – amount of the template free after binding. Selectivity studies 10.0±0.2 mg of the imprinted polymer was weighed in a micro tube and 2.000 mL of 5.0 mmol dm-3 solution of selected UV filter was added. After 1 h of mixing, the polymer was removed by centrifugation, and UV filter concentration in supernatant was determined by HPLC according to procedure described in Section HPLC analysis. HPLC analysis BP4 was analyzed on Purospher Star RP18e 55-4 column, particle size was 3 μm, flow rate 1 mL min-1, solvent A (1 % AcOH in water), solvent B (ACN), solvent volume ratio 80:20. Injection volume was 1.00 μL, column temperature 25 °C, detection wavelength 286 nm, retention time 3.30 min. Total method duration time was 4 min. BP3 and BP8 were analyzed using Waters Symmetry C18 column, 3 mm×100 mm, 5 μm particle size. Flow rate was 0.6 mL/min, solvent A (water), solvent B (ACN), solvent volume ratio 30:70. Injection volume was 1.00 μL, column temperature 25 °C, total method duration time – 4 min. Retention times for BP3 and BP8 were: 2.27 and 1.46 min, respectively. For other UV filters, detection wavelengths were: 310 nm (OCT and PABA-O), 360 nm (AVB), 238 nm (OS and HMS). Retention times were OCT – 2,25 min, AVB – 2.09 min, OS – 2.64 min, HMS – 2.85 min and PABA-O – 2.32 min. Solid phase extraction 100.0 mg of MIP4 was placed in an empty SPE cartridge with a frit on the bottom, PTFE filter (0.45 μm) was placed on the top of the polymer. Packed sorbent was washed with 1.000 mL of ACN, then 3 times with HPLC water (3×0.500 mL). 50 mL of BP4 solution (5.00 mg L-1) was passed through the SPE cartridge at flow rate of approx. 1 mL min-1. The sorption was repeated using the same solution (once already passed). BP4 concentration was deter- mined in filtrate using HPLC. Conductometric titrations 30–50 mg of polymer was placed in a titration cell, 20.0 mL of ACN added, suspension was stirred using a magnetic stirrer and titrated with 0.1024 M HCl solution at 25.0±0.5 °C. The amount of the N-containing group was calculated according to end-point HCl volume. 6 PEŠIĆ, KRSTIĆ and VERBIĆ FTIR FTIR spectra were recorded for NIPs, washed MIPs and MIPs that were washed and subsequently tested for template re-binding. Samples of polymers with bound template were obtained after binding experiments when a 10.0 mM template solution was used. After cen- trifugation, polymer was washed with a minimal volume of ACN, and dried in oven at 60 °C and atmospheric pressure for 24 h. Nitrogen physisorption Nitrogen physisorption was determined at 77 K. Samples were degassed for 1 h at room temperature and kept at 50 °C under vacuum for 24 h. The value of specific surface area of samples was determined using Braunauer, Emmet, Teller method (BET). RESULTS AND DISCUSSION Benzophenone 4 (Fig. 1) is a diprotic acid containing the sulfonic (–SO3H) and the phenolic –OH group. Acting as a strong acid, sulfonic group can form ionic pairs with basic monomers. Hydroxyl group in ortho position is forming a pseudo six-membered ring with benzophenone group and, as such, it most pro- bably doesn’t significantly contribute to the stability of the pre-polymerization complex. Methoxy (–OCH3) group in the position 4 can act as H-bond acceptor (HBA). Two basic monomers were used in our study – DMAEM and 4VP. Both were expected to form ionic pairs and hydrogen bonds with template molecules, as well as to establish π–π stacking interactions (in case of 4VP). EDMA and DVB were used as cross-linkers, expecting that through π–π interactions DVB would contribute more to the binding capacity of the polymers than EDMA. If so, polymers prepared with DVB would have higher adsorption capacity than EDMA polymers. This was confirmed by binding experiments (Section Binding isotherms). ACN is a commonly used solvent for MIP preparation because it favours non-covalent interactions.40 Thus, ACN was our first choice for porogenic sol- vent, but was not applicable to all polymers preparation. When a combination of 4VP and EDMA/DVB was used, not all components were soluble in ACN. It seems that the ionic pair between 4VP and BP4 (or complex in general) is not soluble in EDMA/ACN or DVB/ACN. Monomer/template complex forms imme- diately upon mixing of these two components, which is easily observed by the colour change – mixture turns red (Fig. 2), 4VP is colourless to pale yellow liquid and BP4 is yellow solid. Fig. 2. MIP2, MIP4 and MIP5: a) pre-polymerization complexes and b) prepared polymers. MOLECULARLY IMPRINTED POLYMERS FOR BENZOPHENONE-4 7 As this complex was not soluble in ACN, ACN/toluene mixtures, nor in acetone, other solvent had to be used. Therefore, dry DMSO was used even though it doesn’t enable hydrogen bonding and is not the best choice for porogen. Pre-polymerization mixtures and obtained monolith polymers are shown in Fig. 2. Binding isotherms Binding of BP4 for NIPs and MIPs was conducted by batch binding exp- eriments.25 Binding isotherms were obtained as Q = f(cf) 40 where Q stands for the adsorption capacity (μmol g–1) and cf for the free concentration of template after binding (mmol dm–3), as defined in Eq. (1). Polymers showed very high capacities for binding BP4. As it can be seen on Fig. 3, BP4 binding capacity for each NIP is much lower than for the corresponding MIPs. Polymer MIP4 showed the highest binding capacity. 0 2 4 6 8 0 200 400 600 800 1000 1200 NIP2 MIP2 NIP4 MIP4 NIP5 MIP5 Q /  m o l g -1 c f / mmol dm -3 Fig. 3. Binding isotherms for selected polymers (Table I). Obtained results imply that the imprinting process was successful. If tem- plate – monomer ionic pairs are formed, the formation of high-affinity binding sites is expected, as ionic interactions prevail as the dominant ones. Each avail- able nitrogen-containing part of the polymer will bind the template. So, the only difference between the imprinted and the non-imprinted polymer is that MIP has more available monomer units than NIP because the template was present during polymerization. Imprinting effect Imprinting factor is calculated as ratio of the amount of template bound to imprinted vs. non-imprinted polymer. The imprinting factors for polymer pairs 8 PEŠIĆ, KRSTIĆ and VERBIĆ MIP/NIP 2, 4 and 5, in ACN/water mixtures and ACN are shown in Table II. It is interesting that the presence of water in rebinding medium has a more prominent effect on imprinted polymers. TABLE II. Imprinting factors (IF) for MIP/NIP pairs in ACN (IFACN) and in aqueous sol- utions containing 20 vol. % of ACN (IFACN-water) Polymer IFACN IFACN-water (20 vol. % ACN) Improvement (IFACN-water/IFACN) MIP1:NIP1 1.05 1.14 1.10 MIP2:NIP2 2.53 2.64 1.04 MIP3:NIP3 1.40 1.75 1.25 MIP4:NIP4 2.60 3.17 1.22 MIP5:NIP5 2.10 3.43 1.63 The imprinting factors in aqueous medium, for DVB polymers, are higher than in ACN, but the binding percentage (Table III) of the template to the poly- mer is lower. If water is present in porogen, some interactions between the tem- plate and the monomer, like hydrogen bonding, are disrupted. This leads to lower binding efficiency, but to higher imprinting factor. However, as the imprinted polymers have specific (imprinted) binding sites, the effect of water is lower than in non-imprinted ones. The improvement of imprinting factors in aqueous media compared to ACN is also shown in Table II (calculated as IFACN-water/IFACN). High imprinting factors in aqueous medium are of considerable importance when the application of these polymers in aqueous samples is considered. TABLE III. Binding of BP4 to studied polymers calculated as the amount of bound BP4 rel- ative to starting BP4 concentration (10 mmol dm-3) Polymer Binding of BP4 (ACN) / % Binding of BP4 (20 % ACN, V/V) / % NIP1 29.54 26.64 MIP1 30.52 29.38 NIP2 15.57 11.05 MIP2 31.94 24.69 NIP3 30.88 18.89 MIP3 38.61 28.96 NIP4 32.25 16.31 MIP4 55.41 38.22 NIP5 16.89 10.70 MIP5 29.54 29.10 Adsorption capacities The measured adsorption capacities for imprinted polymers were in the range 0.591–1.108 mmol g–1 of BP4. These capacities are much higher than those described in literature for “regular” MIPs for other templates. Ayadi et al.39 MOLECULARLY IMPRINTED POLYMERS FOR BENZOPHENONE-4 9 described polymers for BP UV filters binding which had a capacity of 27.90 or 96.06 μmol g–1. The adsorption capacities obtained from binding experiments were com- pared to those calculated from the polymer composition (based on monomer composition), but also those obtained from elemental analysis and conducto- metric titrations (Fig. 4). 0.0 0.3 0.6 0.9 1.2 1.5 4 3 2 1 MIP5NIP5MIP4NIP4MIP2 Q /  m o l g -1 NIP2 Polymer Fig. 4. 1: Calculated content of N – based on monomer composition, 2: N content determined by EA, 3: adsorption capacity – determined from adsorption isotherms, 4: monomer content determined by conductometric titration – HCl consumption. According to the elemental analysis results (Supplementary material to this paper, Table S-V), polymers have adsorption capacity in the range 0.714 mmol g–1 (MIP5) up to 1.454 mmol g–1 (MIP4). Capacity was calculated according to the content of nitrogen (%) in the polymer, as all nitrogen in the polymer originates from the functional monomer. As can be seen in Fig. 4, all polymers have lower experimentally determined adsorption capacity than obtained from elemental analysis or based on the monomer and cross-linker composition. The only excep- tion is polymer MIP4, where the adsorption capacity is very close to the deter- mined content of N (EA data) and similar to the calculated nitrogen content, pro- ving the most successful imprinting process for MIP4. To the best of our know- ledge, this represents a unique example in the literature published so far. Conductometric titrations were done to check the accessibility of monomer units in the polymer. Accessible binding sites were calculated according to 0.1 mol dm–3 HCl titration volume (titration graphs shown in Supplementary material, Figs. S-1 and S-2). The values obtained by conductometric titrations are 10 PEŠIĆ, KRSTIĆ and VERBIĆ in between the calculated and the determined N, on one side, and the adsorption capacity on the other side. This can be explained as follows: not all monomer units are available for the template binding, but also, there are units not available for titration with HCl during the applied equilibration time. Again, there is one exception – MIP4. BET surface (Supplementary material, Table S-VIII) areas are larger for MIPs compared to NIPs. However, it can be seen that binding capacity is not in direct correlation with specific surface area, i.e., MIP4 (663 m2 g–1) and MIP5 (634 m2 g–1) have very similar BET areas, but the binding capacity for these two polymers differ almost twice (1.108 mmol g–1 for MIP4, 0.591 mmol g–1 for MIP5). MIP2/NIP2 are found to be mesoporous materials, while MIP4/NIP4 and MIP5/NIP5 seem to be microporous. Selectivity studies Selectivity of the prepared polymers was studied by comparison of the bind- ing of UV filters from few groups of organic UV filters. In Fig. 5, the adsorption capacities are shown for template and other studied UV filters when binding was done using 5 mM solutions. 0 200 400 600 800 BP4BP8PABA-OHMSOSAVBOCT Q /  m o l g -1 MIP2 MIP4 MIP5 BP3 UV filter Fig. 5. Selectivity of imprinted polymers towards other UV filters. As it can be seen, the binding of BP4 (55.4 %) compared to other UV filters (<5.5 %) is much higher due to different type of interactions between template and polymer: the template is bound to the polymer by strong ionic interactions between sulfonic group of the template and pyridine core (or amino group of MOLECULARLY IMPRINTED POLYMERS FOR BENZOPHENONE-4 11 DMAEM) in the polymer. Stronger binding is also a consequence of hydrogen bonding between the –OH group of the template and pyridine core, even though the –OH group is at least partially bound in pseudo-six-membered ring with ben- zophenone oxygen. None of other UV filters have the acidic group as BP4 has, so the rest of the examined compounds bind to polymers mainly by hydrogen bonding. Solid-phase extraction of BP4 from aqueous solutions With MIP4 as a sorbent, 95.0 % of BP4 was adsorbed after the first elution. If the same BP4 solution was passed over MIP4 for the second time, the total of 98.5 % of BP4 adsorption was achieved. The obtained results indicate that the prepared imprinted polymer can be successfully applied for BP4 removal from aqueous solutions. The possibility of extraction from aqueous solutions is crucial for the application in real samples analysis as BP4 can be found in water ecosys- tems.36 Infrared spectroscopy There is no difference between MIP4 (Fig. 6a) and NIP4 (Fig. 6c) in FTIR spectra, indicating that the removal of the template molecule from the imprinted polymer was fully achieved by the applied washing procedure. 4000 3500 3000 2500 2000 1500 1000 e) d) c) b)  / cm -1 a) T / % Fig. 6. Infrared spectra for selected polymers: a) MIP4; b) MIP4 with re-bound BP4; c) NIP4; d) NIP4 with bound BP4; e) BP4. The difference within 1500–1700 cm–1 region between polymers with re- -bound template (Fig. 6b and d) and unbound ones (Fig. 6a and c) confirms the 12 PEŠIĆ, KRSTIĆ and VERBIĆ presence of the template when compared to the spectra of pure template (Fig. 6e). Furthermore, the presence of the re-bound template in spectra of Fig. 6b and d is visible within hydrogen bonds region (3300–3500 cm–1). This is probably due to moisture traces present in BP4. CONCLUSION Benzophenone 4 (BP4) is one of the most commonly used organic UV filter in sunscreen lotions and other cosmetic products. As an emerging environmental contaminant, the need for quantitative determination upon environmental sample preconcentration has pointed out the need for BP4 selective sorbents develop- ment. Five polymer pairs of molecularly imprinted (MIPs) and nonimprinted polymers (NIPs) were prepared within this study. The polymers were charac- terized by template binding studies, infrared spectroscopy, elemental analysis, conductometric titrations and nitrogen adsorption analysis. All MIPs were suc- cessfully imprinted with BP4. However, the imprinting factor was higher in poly- mers containing DVB (2.10–2.60) compared to EDMA (1.05–1.40) polymers. The prepared polymers have high adsorption capacity for template binding – up to 1.108 mmol g–1 (MIP4). Binding of the template to prepared MIPs was com- pared to binding of 7 other organic UV filters, proving high BP4 MIPs select- ivity. The nitrogen physisorption showed that differences in adsorption capacity are not the consequence of different specific surface areas. Finally, MIP4 was used as a sorbent for solid phase extraction of BP4 from aqueous solution, with 98.5 % sorption efficiency. The obtained results are promising for the application of MIPs as solid phase extraction sorbents for BP4 preconcentration from real samples, which is the subject of our ongoing study. SUPPLEMENTARY MATERIAL Additional data and information are available electronically at the pages of journal website: https://www.shd-pub.org.rs/index.php/JSCS/article/view/11718, or from the corres- ponding author on request. Acknowledgement. The authors would like to thank the Ministry of Education, Science and Technological Development of Republic of Serbia (Grant No: 451-03-68/2022-14/200168 and 451-03-68/2022-14/200026) for financial support. И З В О Д ВИСОКО СЕЛЕКТИВНИ ВОДОКОМПАТИБИЛНИ МОЛЕКУЛСКИ ОБЕЛЕЖЕНИ ПОЛИМЕРИ ЗА БЕНЗОФЕНОН-4 МИЛОШ П. ПЕШИЋ1, ЈУГОСЛАВ Б. КРСТИЋ2 и ТАТЈАНА Ж. ВЕРБИЋ1 Универзитет у Београду – Хемијски факултет, Београд и 1Универзитет у Београду – Институт за хемију, технологију и металургију – Институт од националног значаја за Републику Србију, Београд Технологија молекулског обележавања примењена је у синтези селективних сорбе- ната за бензофенон-4 (BP4), органски UV филтер који се користи у кремама за сунчање и другим козметичким производима. Полимери су добијени полимеризацијом у маси, https://www.shd-pub.org.rs/index.php/JSCS/article/view/11718 MOLECULARLY IMPRINTED POLYMERS FOR BENZOPHENONE-4 13 користећи (BP4) као матрицу. Комбинацијом стабилности (механичке и хемијске), селективности и робусности молекулски обележених полимера са својствима BP4 извршено је успешно обележавање (фактор обележавања 1,05–2,60). Карактеризација добијених полимера је извршена применом инфрацрвене спектроскопије, елементалне анализе, кондуктометријских титрација и физисорпције азота на 77 К. Адсорпциони капацицети и селективност испитани су за 7 других органских UV филтера (бензофе- нон-3, бензофенон-8, хомосалат, бутилметоксидибензоилметан, етилхексил-салицилат, етилхексил-р-диметиламинобензоат и етилхексил-р-метоксицинамат), потврђујући велики адсорпциони капацитет и високу селективност за везивање BP4. Највећи адсорп- циони капацитет показао је ко-полимер 4-винилпиридина и дивинилбензена добијен коришћењем диметил-сулфоксида као порогена (1,108 mmol/g). Полимер са највећим капацицетом за везивање BP4 примењен је као сорбент за екстракцију чврстом фазом BP4 из водених раствора са ефикасношћу од 98,5 %. (Примљено 25. марта, ревидирано 6. маја, прихваћено 9. маја 2022) REFERENCES 1. B. Sellergren, Molecularly Imprinted Polymers: Man-Made Mimics of Antibodies and Their Applications in Analytical Chemistry, Elsevier, Amsterdam, 2001. (ISBN: 978- 0444547859) 2. Handbook of Molecularly Imprinted Polymers, C. Alvarez-Lorenzo, A. Concheiro (Eds.), Smithers Rapra, Shrewsbury, 2013 (ISBN: 978-1847359605) 3. Molecularly Imprinted Materials: Science and Technology, M. Yan, O. Ramström (Eds.), Marcel Dekker, New York, 2005 (ISBN: 978-0367578190) 4. Molecular Imprinting of Polymers, Landes Bioscience, S. Piletsky, A.P.F. Turner (Eds.), Georgetown, TX, 2006 (ISBN: 978-1587062193) 5. D. Sun, Z. Song, Y. Zhang, Y. Wang, M. Lv, H. Liu, L. Wang, W. Lu, J. Li, L. Chen, Front. Environ. Chem. 2 (2021) 703961 (https://doi.org/10.3389/fenvc.2021.703961) 6. F. Qiao, H. Sun, H. Yan, K. Row, Chromatographia 64 (2006) 625 (https://doi.org/10.1365/s10337-006-0097-2) 7. M. Sanagi, S. Salleh, W. Ibrahim, A. Naim, D. Hermawan, M. Miskam, I. Hussain, H. Aboul-Enein, J. Food Composit. Anal. 32 (2013) 155 (https://doi.org/10.1016/j.jfca.2013.09.001) 8. M. Lasakova, P. Jandera, J. Sep. Sci. 32 (2009) 799 (https://doi.org/10.1002/jssc.200800506) 9. F. Tamayo, E. Turiel, A. Martin-Esteban, J. Chromatogr., A 1152 (2007) 32 (https://doi.org/10.1016/j.chroma.2006.08.095) 10. Y. Huang, C. Zheng, Z. Liu, Curr. Org. Chem. 15 (2011) 1863 (https://doi.org/10.2174/138527211795656651) 11. W. Cheong, S. Yang, F. Ali, J. Sep. Sci. 36 (2013) 609 (https://doi.org/10.1002/jssc.201200784) 12. W. Cheong, F. Ali, J. Choi, J. Lee, K. Sung, Talanta 106 (2013) 45 (https://doi.org/10.1016/j.talanta.2012.11.049) 13. O. Baltrons, M. Lopez-Mesas, C. Palet, F. Derf, F. Porter-Koltalo, Anal. Methods 5 (2013) 6297 (https://doi.org/10.1039/C3AY41227J) 14. M. Malik, H. Shaikh, G. Mustafa, M. Bhanger, Sep. Purif. Rev. 48 (2019) 179 (https://doi.org/10.1080/15422119.2018.1457541) 15. T. Hishiya, H. Asunama, M. Komiyama, Polym. J. 35 (2003) 440 (https://doi.org/10.1295/polymj.35.440) 16. Y. Saylan, S. Akgonullu, H. Yavuz, S. Unal, Sensors 19 (2019) 1279 (https://dx.doi.org/10.3390%2Fs19061279) https://doi.org/10.3389/fenvc.2021.703961 https://doi.org/10.1365/s10337-006-0097-2 https://doi.org/10.1016/j.jfca.2013.09.001 https://doi.org/10.1002/jssc.200800506 https://doi.org/10.1016/j.chroma.2006.08.095 https://doi.org/10.2174/138527211795656651 https://doi.org/10.1002/jssc.201200784 https://doi.org/10.1016/j.talanta.2012.11.049 https://doi.org/10.1039/C3AY41227J https://doi.org/10.1080/15422119.2018.1457541 https://doi.org/10.1295/polymj.35.440 https://dx.doi.org/10.3390%2Fs19061279 14 PEŠIĆ, KRSTIĆ and VERBIĆ 17. K. Haupt, K. Mosbach, Chem. Rev. 100 (2000) 2495 (https://doi.org/10.1021/cr990099w) 18. N. Leibl, K. Haupt, C. Gonzato, L. Duma, Chemosensors 9 (2021) 123 (https://doi.org/10.3390/chemosensors9060123) 19. H. Lan, PhD Тhesis, Iowa state University, 2018 (https://dr.lib.iastate.edu/entities/publication/baf42fc0-5fec-465e-a78e-d8a0524c44f7) 20. D. Mathew, B. Thomas, K. Devaky, Artif. Cells Nanomed. Biotechnol. 47 (2019) 1149 (https://doi.org/10.1080/21691401.2019.1576703) 21. X. li, M. Zangiabadi, Y. Zhao, J. Am. Chem. Soc. 143 (2021) 5172 (https://doi.org/10.1021/jacs.1c01352) 22. S. Zaidi, Mater. Sci. Technol. 3 (2020) 2589 (https://doi.org/10.1016/j.mset.2019.10.012) 23. S. He, L. Zhang, S. Bai, H. Yang, Z. Cui, X. Zhang, Y. Li, Eur. Polym. J. 143 (2021) (https://doi.org/10.1016/j.eurpolymj.2020.110179) 24. R. Liu, A. Poma, Molecules 26 (2021) 3589 (https://doi.org/10.3390/molecules26123589) 25. M. Pesic, M. Todorov, G. Becskereki, G. Horvai, T. Verbic, B. Tóth, Talanta 217 (2020) 121075 (https://doi.org/10.1016/j.talanta.2020.121075) 26. M. Sibrian-Vazquez, D. Spivak, J. Am. Chem. Soc. 126 (2004) 7827 (https://doi.org/10.1021/ja038961b) 27. L. L. R. de Miranda, K. E. Harvey, A. Ahmed, S. C. Harvey, Environ. Monit. Assess. 193 (2021) 1 (https://doi.org/10.1007/s10661-021-09626-6) 28. Y. Huang, P. Wang, J. C. F. Law, Y. Zhao, Q. Wei, Y. Zhou, Y. Zhang, H. Shi, K. S. Y. Leung, Environ. Int. 143 (2020) 105961 (https://doi.org/10.1016/j.envint.2020.105961) 29. Y. Huang, J. C. F. Law, T. K. Lam, K. S. Y. Leung, Sci. Total. Environ. 755 (2021) 142486 (https://doi.org/10.1016/j.scitotenv.2020.142486) 30. Regulation (EC) No. 1223/2009 of the European Parliament and of the Council of 30 November 2009, Official Journal of the European Union (2009) L 342/59–L 342/209 (http://data.europa.eu/eli/reg/2009/1223/oj) 31. C. A. Downs, E. Kramarsky-Winter, J. E. Fauth, R. Segal, O. Bronstein, R. Jeger, Y. Lichtenfeld, C. M. Woodley. P. Pennington Y. Loya, Ecotoxicol. 23 (2014) 175 (https://doi.org/10.1007/s10646-013-1161-y) 32. C. Corinaldesi, F. Marcellini, E. Nepote, E. Damiani, R. Danovaro, Sci. Total Environ. 637 (2018) 1279 (https://doi.org/10.1016/j.scitotenv.2018.05.108) 33. C. A. Downs, E. Kramarsky-Winter, R. Segal, J. Fauth, S. Knutson, O. Bronstein, F. R. Ciner, R. Jeger, Y. Lichtenfeld, C. M. Woodley, P. Pennington, K. Cadenas, A. Kushmaro, Y. Loya, Arch. Environ. Contam. Toxicol. 70 (2016) 26 (https://doi.org/10.1007/s00244-015-0227-7) 34. O. Ben-Zvi, G. Eyal, Y. Loya, Hydrobiologia 759 (2015) 15 (https://doi.org/10.1007/s10750-014-2063-6) 35. B. Kasprzyk-Hordern, R. Dinsdale, A. Guwy, Water Research 43 (2009) 363 (https://doi.org/10.1016/j.watres.2008.10.047) 36. E. Paredes, S. Pérez, R. Rodil, J. B. Quintana, R. Beiras, Chemosphere 104 (2014) 44 (https://doi.org/10.1016/j.chemosphere.2013.10.053) 37. H. Sun, Y. Li, C. Huang, J. Peng, J. Yang, X. Sun, S. Zang, J. Chen, X. Zhang, J. Sep. Sci. 38 (2015) 3412 (https://doi.org/10.1002/jssc.201500419) 38. C. Ayadi, A. Anene, R. Kalfat, Y. Chevalier, S. Hbaieb, Colloids Surfaces A 567 (2019) 32 (https://doi.org/10.1016/j.colsurfa.2019.01.042) 39. Z. Dorko, A. Szakolczai, T. Verbic, G. Horvai, J. Sep. Sci. 38 (2015) 4240 (https://doi.org/10.1002/jssc.201500874) 40. Molecularly Imprinted Polymers in Biotechnology, B. Mattiasson, L. Ye (Eds.), Springer International Publishing, 2015 (ISBN: 978-3319207292). https://doi.org/10.1021/cr990099w https://doi.org/10.3390/chemosensors9060123 https://dr.lib.iastate.edu/entities/publication/baf42fc0-5fec-465e-a78e-d8a0524c44f7 https://doi.org/10.1080/21691401.2019.1576703 https://doi.org/10.1021/jacs.1c01352 https://doi.org/10.1016/j.mset.2019.10.012 https://doi.org/10.1016/j.eurpolymj.2020.110179 https://doi.org/10.3390/molecules26123589 https://doi.org/10.1016/j.talanta.2020.121075 https://doi.org/10.1021/ja038961b file:///C:/Users/Milos%20Pesic/Dropbox/Milos%20-%20PhD/Rad%20II%202022/ https://doi.org/10.1007/s10661-021-09626-6 https://d.docs.live.net/595bad50a54dda5b/02_JSCS_OnLine_First/priprema/11718/ https://doi.org/10.1016/j.envint.2020.105961 https://doi.org/10.1016/j.scitotenv.2020.142486 http://data.europa.eu/eli/reg/2009/1223/oj https://doi.org/10.1007/s10646-013-1161-y file:///C:/Users/Milos%20Pesic/Dropbox/Milos%20-%20PhD/Rad%20II%202022/ https://doi.org/10.1016/j.scitotenv.2018.05.108 file:///C:/Users/Milos%20Pesic/Dropbox/Milos%20-%20PhD/Rad%20II%202022/ file:///C:/Users/Milos%20Pesic/Dropbox/Milos%20-%20PhD/Rad%20II%202022/ https://doi.org/10.1007/s00244-015-0227-7 file:///C:/Users/Milos%20Pesic/Dropbox/Milos%20-%20PhD/Rad%20II%202022/ file:///C:/Users/Milos%20Pesic/Dropbox/Milos%20-%20PhD/Rad%20II%202022/ https://doi.org/10.1007/s10750-014-2063-6 https://doi.org/10.1016/j.watres.2008.10.047 file:///C:/Users/Milos%20Pesic/Dropbox/Milos%20-%20PhD/Rad%20II%202022/ file:///C:/Users/Milos%20Pesic/Dropbox/Milos%20-%20PhD/Rad%20II%202022/ https://doi.org/10.1016/j.chemosphere.2013.10.053 https://doi.org/10.1002/jssc.201500419 https://doi.org/10.1016/j.colsurfa.2019.01.042 https://doi.org/10.1002/jssc.201500874