English

• Arsenic (As) and antimony (Sb) have been listed as priority pollutants by the U.S. Environmental Protection Agency due to their high toxicities and ubiquitous existence in the natural environment [1]. As and Sb are redox-sensitive elements and exist predominantly as trivalent oxyanions (As(Ⅲ) and Sb(Ⅲ)) and pentavalent oxyanions (As(Ⅴ) and Sb(Ⅴ)) in anoxic and oxic environments, respectively [2,3]. These metal(loid)s enter water sources through natural processes and anthropogenic activities, posing significant risks to human health and ecosystems. According to a recent global As risk assessment, about 94–220 million people worldwide are exposed to high levels of As in groundwater [1]. Long-term exposure to As can cause cancers of the skin, bladder, and lungs [4-6]. Sb exposure has been reported to be associated with respiratory and gastrointestinal diseases [7]. As and Sb commonly occur in the same geochemical environment and their co-contamination of soil, surface water, and groundwater has attracted increasing concern globally [8,9]. Thus, the simultaneous removal of As and Sb species is of great urgency and significance.

  To this end, various strategies have been developed, such as membrane separation [10,11], biological treatment [12], coagulation/filtration [13,14], and adsorption [15]. Among them, adsorption is one of the most extensively utilized technologies due to its cost-effectiveness and high removal efficiency [16,17]. Titanium dioxide (TiO₂) has been widely used for the removal of As and Sb due to its low toxicity, high stability, and strong adsorption capacity. Arsenic and antimony species can easily be adsorbed onto TiO₂ through bidentate or monodentate configurations [18,19]. However, the difficulty in recovering TiO₂ nanoparticles from post-treatment slurries hinders their practical large-scale application [20]. Pentavalent species of As and Sb are known to dominate in natural waters when exposed to sunlight and dissolved oxygen [21], while TiO₂ possesses lower adsorption capacities for As(Ⅴ) (35.2 mg/g) and Sb(Ⅴ) (8.6 mg/g) than for As(Ⅲ) (93.0 mg/g) and Sb(Ⅲ) (12 mg/g), respectively [18,22]. It is thus urgent and significant to tackle the recovery issue of TiO₂ nanoparticles and enhance their adsorption capacity for As(Ⅴ) and Sb(Ⅴ) in water.

  The recovery problem can be resolved by incorporating nanosized TiO₂ onto larger particulate substrates to form composite adsorbents at the millimeter scale [23]. The high stability of the substrate and the feasibility of functionalization are central to the design of the composite adsorbents. Polystyrene (PS) resin is a promising support of TiO₂ nanoparticles due to its high mechanical strength, large pore size, efficient mass transfer, and cost-effectiveness [24-26]. Moreover, thanks to its benzene rings and vinyl groups, PS can be readily modified with various functional groups, offering possibilities for various chemical reactions [27].

  The amine group (–NH₂) can be protonated to –NH₃⁺ and thus favor the adsorption of pentavalent oxyanions, such as H₂AsO₅⁻ and Sb(OH)₆⁻, due to electrostatic attractions [28]. Tetraethylenepentamine (TEPA) contains two –NH₂ and three –NH– groups, which readily form complexes with metals via the coordination bonds. The lone pair of the N atom engages in coordination bonding with the metal, thus establishing a robust linkage between the amine group and the metal (e.g., Ti, Cr) [29,30]. With the above strategy, loading TiO₂ onto TEPA-modified PS support can facilitate the easy recovery of TiO₂ after treatment, and simultaneously enhance the adsorption capacity toward pentavalent As and Sb species. The obtained composite should behave with excellent stability and selectivity in theory due to the excellent coordination ability of N–Ti and the specific affinity of TiO₂ for As and Sb. The functionalized PS resin has been generally synthesized via the chloromethylation method, and the use of carcinogenic chloromethyl ether is not environmentally friendly. Moreover, poly-substitution easily occurs to generate byproducts via this method, reducing the purity of the synthesized functionalized PS [31,32].

  In this study, we synthesized a novel composite adsorbent, nano TiO₂-loaded aminated polystyrene crosslinked with divinylbenzene (AmPSd-Ti), via the environmentally friendly Friedal-Crafts acylation and Mannich reaction. The mm-scale AmPSd-Ti was fully characterized by a series of complementary techniques. Its high adsorption capacities for As and Sb were confirmed by batch and fixed-bed column experiments using simulated and real natural waters. The results of adsorption kinetics, isotherms, X-ray photoelectron spectroscopy (XPS) analysis, and molecular dynamics simulations revealed that the nano TiO₂ was firmly anchored onto AmPSd through N–Ti bonding. The protonated –NH₂ group enhanced the adsorption of H₂AsO₅⁻ and Sb(OH)₆⁻ by electrostatic attraction. AmPSd-Ti exhibited high affinity to As and Sb via inner-sphere complexation. The comprehensive investigation by various techniques revealed that the amine group not only acts as an anchor, effectively binding TiO₂ to the resin via stable N–Ti coordination bonds but also significantly enhances the adsorption of pentavalent As and Sb in a wide pH range. This study addresses a significant improvement in understanding the mechanism of amine functionalization in the adsorption of arsenic and antimony on TiO₂, providing new insights into how the amine groups contribute to enhancing the adsorption efficiency.

  Styrene (C₈H₈, CP, with 10–15 mg/L 4-tert-butyl-catechol (C₁₀H₁₄O₂, TBC) as a stabilizer), divinylbenzene (C₁₀H₁₀, 80% mixture of isomers with 0.1% TBC as a stabilizer), and 2,2-azobis(isobutyronitrile) (C₈H₁₂N₄, AIBN, 98%) were purchased from Shanghai Macklin Chemical Co., Ltd., and further purified before use. Titanium oxysulfate-sulfuric acid hydrate (TiOSO₄·xH₂SO₄·xH₂O), tetraethylenepentamine (C₈H₂₄N₄, TEPA, AR, 95%), n-heptane (C₇H₁₆, AR, 98%) and aluminum chloride (AlCl₃, AR, 99%) were obtained from Aladdin Co. (Shanghai, China).

  AmPSd-Ti was prepared according to the following procedure: Firstly, polystyrene cross-linked with divinylbenzene (PSd) was prepared by the suspension polymerization of styrene at the interface of aqueous-organic phases. Divinylbenzene, AIBN, and n-heptane acted as the cross-linking agent, initiator, and porogenic agent, respectively. Secondly, aminated PSd (AmPSd) was achieved by Friedal-Crafts acylation and Mannich reactions. Carcinogenic chloromethyl ether has been used for conventional Friedal-Crafts acylation. Herein, we first used acetyl chloride to introduce the functional group at the styrene para position, which is more environmentally friendly due to its lower toxicity. TEPA functionalization was then achieved via the Mannich reaction. The introduction of two -NH₂ and three -NH- groups into the PSd backbone ensured the robust loading of TiO₂ through the N-Ti coordination bond. Thirdly, AmPSd-Ti was synthesized via the in-situ deposition of TiO₂ onto AmPSd, which was achieved through the hydrolysis of titanium oxysulfate and sulfuric acid hydrate. The procedure of the synthesis of AmPSd-Ti is shown in Scheme 1. The purification methods of styrene, divinylbenzene, AIBN, and the synthesis procedure of AmPSd-Ti are detailed in Text S1 (Supporting information).

  Scheme 1

  

  Scheme 1.  The fabrication procedure of AmPSd-Ti.

  DownLoad: Full-Size Img PowerPoint

  To study the adsorption capability of AmPSd-Ti via column experiments, two 50-L water samples were collected from representative As- and Sb-contaminated areas in China. Sample 1 (pH 7.7) characterized by a high As level (319 µg/L) was collected from the groundwater of Datong basin (39°26′15.33″ N, 112°56′23.87″ E), China. Sample 2 (pH 8.9) was sampled from the industrial wastewater containing 14.8 mg/L Sb at Xikuangshan (27°39′44.10″ N, 111°18′12.23″ E), China, the world's largest Sb deposit. The method of analysis of the environmental samples is detailed in Text S2 (Supporting information) and the concentrations of metal elements and anions are summarized in Table S1 (Supporting information).

  The morphology of the as-prepared materials was characterized by a scanning electron microscope (SEM, Apero S Hivac, Thermofisher, USA) equipped with an energy-dispersive X-ray spectroscopy (EDS, Oxford, UK) for the elemental analysis. The functional groups were examined by Fourier-transform infrared spectroscopy (FTIR, Thermo Nicolet iS50R, USA). The N₂ adsorption-desorption isotherms and specific surface area were determined by the Brunauer–Emmett–Teller (BET) method at 77 K via ASAP2460 analyzer (ASAP 2020HD88, Micromeritics, China). The X-ray diffraction (XRD) spectra were collected using an X'Pert PRO diffractometer (PANalytical, Netherlands) with Cu Kα radiation. The elemental composition and corresponding oxidation state of the adsorbent before and after adsorption were analyzed by XPS (Thermo Scientific K-Alpha, USA). All XPS spectra were calibrated using the C 1s signal located at 284.8 eV. The zeta potential was measured using Zetasizer Nano ZS (Malvern Instrument Ltd., UK). Thermogravimetric analysis (TGA) was performed via TG209F1 Libra (USA).

  To investigate the optimum dosage, the adsorption experiments were conducted in the presence of AmPSd-Ti at concentrations ranging from 0 g/L to 1.5 g/L and an initial pollutant concentration of 1 mg/L. Adsorption isotherms were obtained by adding 10 mg of AmPSd-Ti into 40 mL of As or Sb solution (1–100 mg/L) after a 24-h equilibration on a rotator. The adsorption kinetics experiment was performed with pollutant concentrations of 500 µg/L and 1 mg/L. The adsorption pH envelope experiment was conducted with pH 2–12 and an initial concentration of As and Sb at 10 mg/L. The competitive adsorption of common anions in the environment, including SO₄²⁻, NO₃⁻, and F⁻, was studied with a molar ratio of the anions to As or Sb of 1:1, 5:1, and 10:1. AmPSd-Ti was regenerated via acid treatment using a 0.5 mol/L HCl solution after adsorption. The ionic strength, initial pH, dosage, and temperature for all batch experiments were 0.05 mol/L NaCl, 6.9 ± 0.1, 0.25 g/L, and 25 ℃, respectively, if not specifically stated. The suspension samples were collected and filtered through a 0.22 µm polyethersulfone membrane filter (Supor, Pall, USA), and then were analyzed by a high-performance liquid chromatograph (HPLC) coupled with a hydride generation-atomic fluorescence spectrometer (HG-AFS, Haiguang, China). The isotherm results were fitted with Langmuir and Freundlich models (Text S3 in Supporting information). The kinetic results were fitted using pseudo-first-order, pseudo-second-order, intra-particle diffusion, and liquid film diffusion models, as detailed in the Text S4 (Supporting information).

  We packed 1.50/2.78 g AmPSd-Ti into a glass column with an inside diameter of 1 cm and an adjustable length of 3.3/6.4 cm, producing a 2.59/5.03 mL bed volume for sample 1 and sample 2, respectively. Two 0.22-µm membrane filters were placed at the end of each column (Fig. 1A). The environmental samples were continuously injected into the column by a peristaltic pump at a fixed flow rate of 1 mL/min in an up-flow fashion. The empty bed contact time (EBCT) was set to 2.59 and 5.02 min, respectively. Breakthrough curves were obtained by collecting 1 mL of the effluent solution at certain intervals using a fraction collector. In the c_t/c₀ ratio against the time plot, c₀ and c_t represent the influent concentration at time 0 and t, respectively. The breakthrough curves for As and Sb were fitted by Thomas and Adams-Bohart models (Text S5 in Supporting information) to assess the efficacy of AmPSd-Ti within fixed-bed column setups.

  Figure 1

  

  Figure 1.  (A) Scheme of the fixed-bed column experiments. (B) Optical photograph of multiple AmPSd-Ti spheres. (C, D) Cross-section SEM image in an overview of an individual sphere and the enlarged view of the selected area. (E) EDS line scan analysis of AmPSd-Ti with adsorbed Sb(Ⅴ) and elemental distribution of (F) C, (G) O, (H) N, (I) Ti, and (J) Sb. (K) FTIR spectra of synthesized PSd, AcPSd, AmPSd, and AmPSd-Ti. (L) High-resolution XPS spectra of synthesized TiO₂ and AmPSd-Ti. (M) XRD patterns of Synthesized TiO₂, AmPSd, AmPSd-Ti, and AmPSd-Ti after As/Sb adsorption.

  DownLoad: Full-Size Img PowerPoint

  As detailed in the Text S6 (Supporting information), the TiO₂ nanocluster (Ti₂₈O₅₆) was trimmed from bulk anatase. The TEPA-functionalized PSd was simplified to TEPA-functionalized benzene (AmBenzene, C₁₇H₃₄O₁N₅). The geometrical structures of Ti₂₈O₅₆ and C₁₇H₃₄O₁N₅ were optimized using the B3LYP hybrid function with 3–21G(d) and 6–31G(d) basis sets, respectively. Dominant species of As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ) at pH 6.9 were As(OH)₃, H₂AsO₅⁻, Sb(OH)₃, and Sb(OH)₆⁻, and were geometrically optimized with def2TZVP basis set. The optimization was performed with the Gaussian 09 package, and the optimized structures were then used to calculate their restrained electrostatic potential (RESP) charge distributions. Based on the Universal force field (UFF) [33], the force field parameters of TiO₂, AmPSd, and pollutants were developed by Multiwfn and Sobtop [34]. All molecular dynamics simulations were performed using the GROMACS 2023 package [35]. Energy minimization was first performed using the steepest descent method, followed by a 100-ps NVT simulation at 300 K and a 100-ps NPT simulation at 300 K and 1 bar, to equilibrate the solvated system. A 10-ns production run was then performed at 300 K and 1 bar The cutoff values for both the short-range van der Waals (vdW) and electrostatic interactions were 1.2 nm. The LINCS algorithm was employed to contain the bonds and hydrogen atoms [36]. The results of the molecular dynamics simulation were visualized by Visual Molecular Dynamics program [37].

  The morphology and structure of the synthesized TiO₂ and AmPSd-Ti were characterized by complementary techniques. Fig. S1 (Supporting information) shows the microscopic morphology of synthesized TiO₂ as clusters formed by nanoparticles. SEM-EDS analyses revealed the uniform spherical morphology and porous structure of AmPSd-Ti with a diameter of ~1.2 mm (Figs. 1B–D). The elemental EDS mapping images showed the uniform distribution of C, O, N, and Ti elements throughout the whole sphere (Figs. 1E–J), indicating the successful incorporation of TiO₂ nanoparticles into the aminated PS resin framework.

  The amine functionalization of the PS framework was confirmed by FTIR spectra (Fig. 1K). The FTIR spectra of PSd, acetylated PSd (AcPSd), AmPSd, and AmPSd-Ti showed two bands at 2920 and 2848 cm⁻¹, attributed to the stretching vibration of the C–H bond in methylene groups (–CH₂–) and methine groups (–CH–) [38]. The peaks at 757 and 697 cm⁻¹ originated from the stretching vibrations of C–H in the benzene ring [39]. These results verified the structure of the PS resin framework. The FTIR spectrum of AcPSd showed two peaks at 1680 and 1357 cm⁻¹, corresponding to the carbonyl (C═O) and methyl (–CH₃) groups, respectively, introduced by Friedal-Crafts acylation [40]. An additional peak was observed at 830 cm⁻¹ in the FTIR spectrum of AcPSd, attributed to the bending vibrations of C–H of the 1,4-disubstituted benzene ring, indicating that the carbonyl group substituted the H atom in the para position of the polystyrene [41]. For the AmPSd, the appearance of two weak peaks at 1182 and 1569 cm⁻¹, corresponds to the C–N stretching vibrations in secondary amines and N–H stretching vibrations [42,43], suggesting the successful amination of PS framework via Mannich reaction using TEPA.

  The incorporation of nano TiO₂ into the PS resin framework was further confirmed by FTIR, XPS, XRD, and thermogravimetric analysis. The characteristic peaks of O–Ti–O at 441 cm⁻¹ and N–Ti at 636 cm⁻¹ were observed in the FTIR spectrum of AmPSd-Ti [44]. Only AmPSd-Ti showed the peak at 636 cm⁻¹ among the spectra of the three samples (Fig. S2 in Supporting information), indicating the successful incorporation of nano TiO₂ into the AmPSd via forming N–Ti coordination bond, consistent with the previous works [45-48]. This result was further confirmed by the XPS analysis. Due to the weaker electronegativity of N than O, the electron density of Ti in N–Ti is higher than that in Ti–O. For the Ti 2p XPS spectra, the peaks assigned to Ti 2p_1/2 and Ti 2p_3/2 of AmPSd-Ti shifted to a lower binding energy by 0.5–0.6 eV compared with those of synthesized TiO₂ (Fig. 1L), corresponding to the higher electron density of Ti and thus confirming the formation of N–Ti bonds [49]. The XRD pattern of the synthesized TiO₂ showed diffraction peaks at 25.31°, 37.80°, 48.04°, 55.06°, and 62.69° (Fig. 1M), well indexed to the anatase (JCPDS #84-1285). The diffraction peaks of both anatase TiO₂ and AmPSd were observed in the XRD pattern of AmPSd-Ti, indicating the successful coupling of the two components. Based on the thermogravimetric analysis of AmPSd-Ti, AmPSd completely decomposed at 775.2 ℃ with 25.9 wt% of the sample remaining, indicating the considerable content of TiO₂ in AmPSd-Ti (Fig. S3 and Table S2 in Supporting information). AmPSd-Ti was thermally stable as its decomposition temperature was 207.8 ℃.

  The N₂ adsorption measurements were conducted to determine the Brunauer–Emmett–Teller surface area (S_BET) and pore structure of the adsorbents. AmPSd-Ti showed a Type-IV N₂ adsorption isotherm featured with an H3-type hysteresis loop, indicating the mesoporous structure of AmPSd-Ti (Fig. S4A in Supporting information) [50]. The pore size distribution curves of AmPSd-Ti derived from the Barrett–Joyner–Halenda method showed an obvious peak at 20–50 nm (Fig. S4B in Supporting information). AmPSd-Ti was measured to possess an S_BET of 44.40 m²/g and a pore size of 30 nm (Table S3 in Supporting information).

  The adsorption capacities of AmPSd-Ti were evaluated via batch experiments. AmPSd-Ti exhibited removal rates of 95.32%, 94.86%, 97.59%, and 96.83% for As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ) in the contaminated water within 24 h, respectively, with an initial concentration of 1 mg/L and an adsorbent dosage of 0.25 g/L (Fig. S5 in Supporting information). The adsorption isotherms were obtained under the same conditions (Fig. 2). The adsorption of As and Sb by AmPSd-Ti fitted better with the Freundlich isotherm model (0.985 ≤ R² ≤ 0.991) than the Langmuir one (0.956 ≤ R² ≤ 0.986), as summarized in Table S4 (Supporting information), implying the adsorption process is multilayered and non-homogeneous [2,51]. The adsorbed molecules could engage in intermolecular interactions, facilitating multilayer adsorption on the surface [52,53]. In addition, the n values (ranging from 1.764 to 2.660) were greater than 1, and the k_F values were in the range of 1–10, indicating that the adsorption process was thermodynamically favorable and tended to occur at lower concentrations [54,55]. The maximum adsorption capacity (q_max) of AmPSd-Ti reached 73.85, 153.29, 86.80, and 123.71 mg/g for As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ), respectively (Fig. 2).

  Figure 2

  

  Figure 2.  Adsorption isotherms of AmPSd-Ti on (A) As and (B) Sb, and fitted Freundlich adsorption isotherm model lines. Dosage: 0.25 g/L, initial concentration: 1–100 mg/L, initial pH: 6.9 ± 0.1, contact time: 24 h, ionic strength: 0.05 mol/L.

  DownLoad: Full-Size Img PowerPoint

  As the co-contamination of As and Sb is ubiquitous worldwide [9,56,57], it is desired to obtain adsorbents for simultaneous removal and AmPSd-Ti is such an adsorbent. AmPSd-Ti exhibited considerably higher adsorption capacities for As or Sb than many other metal-based and polymer-based composites reported previously (Table S5 in Supporting information). In addition, superior to other materials, AmPSd-Ti could be used for the treatment of both As and Sb-contaminated water. Trivalent and pentavalent anions of As and Sb showed different features for adsorption treatment. Sb(Ⅴ) species have higher mobilities and are more reluctant to be adsorbed on iron oxides and titanium oxides than Sb(Ⅲ) [58]. The affinity of As(Ⅴ) on the TiO₂ surface is similar to that of As(Ⅲ) [59], and decreases when the pH value increases [18]. In the natural environment, As(Ⅴ) is more prevail in the presence of sunlight and dissolved oxygen [21], and Sb(Ⅴ) is the most commonly detected species on site [60]. For example, the concentrations of As(Ⅴ) and Sb(Ⅴ) in the wetlands of Idaho reached up to 21,902 and 1503 µg/L, respectively, which were higher than those of As(Ⅲ) (350.5 µg/L) and Sb(Ⅲ) (5.54 µg/L) [61]. It was noteworthy that AmPSd-Ti showed enhanced adsorption capacities for As(Ⅴ) and Sb(Ⅴ) compared with other reported materials. Therefore, the as-prepared AmPSd-Ti shows promise in the synergistic removal of As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ) from contaminated waters.

  The adsorption kinetics of As and Sb by AmPSd-Ti were investigated. The adsorbed species increased with time and reached equilibrium within 24 h (Figs. 3A, C, E, G). The adsorption of all four species followed the pseudo-second-order model (R² > 0.963, Tables S6 and S7 in Supporting information) with rate constants (k_s) of 2.78, 2.99, 8.54, and 4.21 (×10⁻³ min⁻¹) for As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ), respectively, implying that the chemisorption dominated in the adsorption [62].

  Figure 3

  

  Figure 3.  (A, C, E, G) Fitted pseudo-first-order, pseudo-second-order adsorption kinetic curves and (B, D, F, H) liquid film diffusion model of arsenic and antimony adsorption on AmPSd-Ti. Dosage: 0.25 g/L, initial concentration: the red and blue legends represent 0.5 and 1 mg/L, respectively, initial pH: 6.9 ± 0.1, contact time: 0–24 h, ionic strength: 0.05 mol/L.

  DownLoad: Full-Size Img PowerPoint

  To anchor the rate-limiting step of the adsorption process, we fitted the kinetics data using intra-particle and liquid film diffusion models, as detailed in the Text S4 (Supporting information). As shown in Fig. S6 (Supporting information), the intra-particle diffusion model, plotting q_t vs. t^1/2, failed to depict the adsorption process especially when approaching equilibrium because the successful modeling requires the fitted equation to have a zero intercept [63]. On the contrary, the liquid film diffusion model fitted the data well, suggesting that the rate-limiting step is liquid film diffusion (Figs. 3B, D, F, H). The calculated liquid film diffusion rate constant k_lfd for As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ) are 0.0043, 0.0044, 0.0069, and 0.0054 min⁻¹. The well-developed porous structure of AmPSd-Ti facilitated the diffusion of As and Sb molecules into the pores [64].

  To explore the adsorption mechanisms, the XPS spectra of AmPSd-Ti were obtained before and after As and Sb adsorption. The O 1s, Ti 2p, N 1s, and C 1s peaks were resolved in the survey spectrum of raw AmPSd-Ti (Fig. S7 in Supporting information), while As 3d (44.62 eV) and Sb 3d (540.69 eV) peaks were additionally observed in the spectrum of the used AmPSd-Ti. The high-resolution Ti 2p XPS spectrum of AmPSd-Ti shows two peaks at 464.95 and 459.20 eV, corresponding to the 2p_1/2 and Ti 2p_3/2 orbitals of Ti⁴⁺ (Fig. 4A) [44,65]. After As/Sb adsorption, their binding energies decreased slightly (Fig. 4A and Table S8 in Supporting information), which was attributed to the increased electron density around Ti due to the substitution of partial –OH groups on TiO₂ by As(Ⅲ/Ⅴ) and/or Sb(Ⅲ/Ⅴ). The O 1s region resolved three components including lattice oxygen (O²⁻ in TiO₂), metal-bonded hydroxyl (–OH) group, and surface adsorbed H₂O [53,66] in Fig. 4B. The –OH percentage decreased from 50.66% to 33.29%, 29.71%, 34.12%, and 35.93% after As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ) adsorption, respectively, originating from the exchange of present ions and the –OH groups on TiO₂ surface. The above results confirmed the formation of inner-sphere complexes between TiO₂ and As(Ⅲ/Ⅴ) and/or Sb(Ⅲ/Ⅴ) [67,68]. Consisting of the previously reported adsorption structure of As/Sb on pure nano-sized TiO₂ [19].

  Figure 4

  

  Figure 4.  High-resolution scan of (A) Ti 2p, (B) O 1s, and (C) N 1s core level of AmPSd-Ti before and after As(Ⅲ/Ⅴ) and Sb(Ⅲ/Ⅴ) adsorption.

  DownLoad: Full-Size Img PowerPoint

  To confirm the role of TiO₂ in the adsorption of As/Sb, we performed a molecular dynamics simulation of the adsorbate-TiO₂ systems. As shown in Fig. S8 (Supporting information), As and Sb species exhibited a high affinity for the TiO₂ surface. The total interaction energies (TIE), including short-range Coulombic interaction energy (CIE) and short-range Lennard–Jones energy (LJE), were calculated. As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ) showed total interaction energies of −1686.81, −818.17, −837.48, and −617.74 kJ/mol, respectively. These results demonstrated the first step of the adsorption process that forms inner-sphere complexes on the surface of nano-sized TiO₂.

  The high-resolution N 1s XPS spectrum of AmPSd-Ti revealed three deconvolution peaks at 402.35, 400.45, and 399.42 eV, which were assigned to C–N, N⋯H–O, and –NH₂/N–H species, respectively [64,69] in Fig. 4C. The N⋯H–O species could be attributed to the formation of hydrogen bonds between the amine groups of TEPA and the oxyanions of As/Sb. Moreover, the N⋯H–O percentage increased from 7.03% to 26.82%, 24.33%, 28.96%, and 40.04% after As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ) adsorption, respectively. These results demonstrated that the formed hydrogen bonds between TEPA and As/Sb species played a crucial role in the adsorption process.

  As AmPSd-Ti contains the –NH₂ group, the surface charges of AmPSd-Ti can be affected by pH values, which makes the adsorption capacities of As and Sb AmPSd-Ti pH-dependent. As shown in Fig. 5A, AmPSd-Ti exhibited high adsorption capacities for As and Sb within a wide range of pH values from 2.0 to 10.0, outperforming many other adsorbents with a drastically decreased performance for As(Ⅴ) and Sb(Ⅴ) under alkaline condition [69-71]. The adsorption property of AmPSd-Ti decreased when the pH value was over 10.0. To reveal the underlying mechanism, the zeta potentials of the as-prepared samples were measured. The point of zero charge (PZC) of AmPSd-Ti was determined to be 10.4, at which the surface charge of AmPSd-Ti was neutral (Fig. 5B). When pH values were below 10.4, AmPSd-Ti was positively charged and tended to adsorb negatively charged As/Sb oxyanions due to the electrostatic attractions. After the adsorption of As/Sb, the zeta potentials of AmPSd-Ti became negative and decreased with increasing pH values, confirming the formation of negatively charged inner-sphere complexes [19,72]. The PZC of AmPSd-Ti (10.4) was much higher than that of TiO₂ (6.6) [19], benefiting from the amine groups of TEPA which endowed AmPSd-Ti with a positively charged surface via protonation in a wider pH range [30]. Therefore, the amine functionalization contributed to the enhancement of As(Ⅴ) and Sb(Ⅴ) adsorption on AmPSd-Ti. To further verify the role of amine groups in the adsorption process, we performed molecular dynamics simulations of the adsorbate-AmBenzene systems. As illustrated in Figs. 5C–F, hydrogen bonds were readily formed between As/Sb species and amine groups, according to the geometric criterion that intermolecular hydrogen bonds form if the donor-acceptor distance is less than 3.5 Å and the donor-acceptor angle is greater than 120°, and vice versa [73]. This result implied that the physical adsorption of the AmPSd-Ti framework contributed to the removal of As and Sb, in line with the XPS analysis. The H₂AsO₅⁻ and Sb(OH)₆⁻ anions were subjected to the Coulomb attraction of protonated TEPA, resulting in higher short-range Coulombic interaction energies than their trivalent counterparts. The total interaction energies of H₂AsO₅⁻ and Sb(OH)₆⁻ were −602.67 and −366.87 kJ/mol, respectively, higher than those of As(OH)₃ (−62.44 kJ/mol) and Sb(OH)₃ (−118.59 kJ/mol), contributing to the enhancement of physical adsorption. In addition, enhanced physical adsorption was beneficial for concentrating the target pollutants in the AmPSd-Ti framework, which facilitated chemisorption on the surface of TiO₂ nanoparticles.

  Figure 5

  

  Figure 5.  (A) Adsorption capacities of As(Ⅲ/Ⅴ) and Sb(Ⅲ/Ⅴ) on AmPSd-Ti at different initial pH (Dosage: 0.25 g/L, initial concentration: 1 mg/L, initial pH: 2.0–12.0, contact time: 24 h, ionic strength: 0.05 mol/L). (B) Zeta potentials of AmPSd-Ti at different pH before and after As(Ⅲ/Ⅴ) and Sb(Ⅲ/Ⅴ) adsorption. (C–F) Molecular dynamics simulation of the adsorbate-AmPSd systems.

  DownLoad: Full-Size Img PowerPoint

  The stability of AmPSd-Ti was assessed at different pH values. The release of Ti from the composite at pH 2–12 was below 65 µg/L (less than 0.1% of the weight of Ti in AmPSd-Ti) (Fig. S9 in Supporting information). Competitive adsorption experiments were performed to assess the selectivity of AmPSd-Ti. NO₃⁻ and SO₄²⁻ were reported to coexist with Sb in the environment and potentially have a competitive adsorption [19]. As shown in Fig. 6, the coexisting anions, including NO₃⁻ and SO₄²⁻, showed no significant influence on the removal of the four As/Sb species over a wide range of coexisting-to-target ion ratios from 1:1 to 10:1. The performance of AmPSd-Ti decreased slightly when the concentration of F⁻ increased from 1 mg/L to 10 mg/L, but the removal rate remained above 84.19%, 86.34%, 83.06%, and 67.74% for As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ), respectively. After five cycles of regeneration with 0.5 mol/L HCl acid, the adsorption capacities of AmPSd-Ti for As(Ⅲ), As(Ⅴ), Sb(Ⅲ), and Sb(Ⅴ) remained over 86.32%, 84.04%, 88.41%, and 94.25% (Fig. S10 in Supporting information). These results proved the outstanding stability, selectivity, and reusability of the amine-functionalized polystyrene resin-titania spheres.

  Figure 6

  

  Figure 6.  The adsorption capacities of As (A, B) and Sb (C, D) onto AmPSd-Ti in the presence of various coexisting anions (Dosage: 0.25 g/L, initial concentration: 1 mg/L, initial pH: 6.9± 0.1, contact time: 24 h, ionic strength: 0.05 mol/L). The red dashed line represents the adsorption capacity of AmPSd-Ti for As(Ⅲ/Ⅴ) and Sb(Ⅲ/Ⅴ) without the addition of coexisting anions.

  DownLoad: Full-Size Img PowerPoint

  To evaluate the capability of AmPSd-Ti for remediating As and Sb in real scenarios, fixed-bed column tests were performed using two representative environmental samples. The groundwater (sample 1, Fig. S11A in Supporting information) and industrial wastewater (sample 2, Fig. S11B in Supporting information) contained high levels of As (319 µg/L) and Sb (14.8 mg/L), respectively, which are similar to typical As/Sb contaminated waters [60].

  EPA's maximum contaminant levels of As and Sb for drinking water are 10 and 6 µg/L, respectively [74], and the industrial discharge limit of Sb is 300 µg/L [75]. The breakthrough of sample 1 (the concentration of As reached 10 µg/L) occurred at 6600 BV, at which point the capacity of AmPSd-Ti was determined to be 1.38 mg/g (Fig. 7A). The World Health Organization has estimated that the daily water demands for adult females and males are 2.2 and 2.9 L, respectively. Sample 1 can be treated with 1 g AmPSd-Ti to obtain 4.30 L clean water in hours, meeting the daily drinking water needs for 2.0 adult females or 1.5 males. As for sample 2, the breakthrough (the concentration of Sb reached 300 µg/L) occurred at 1260 BV, and an adsorption capacity of 6.65 mg/g was achieved by AmPSd-Ti (Fig. 7B). This finding indicated that each gram of AmPSd-Ti could remediate 0.45 L of Sb-contaminated industrial wastewater to an acceptable level for discharge. To achieve large-scale groundwater purification, we filled the column with 10 mL (5.8 g) of AmPSd-Ti and treated sample 1. The device (Fig. 7C) operated continuously for 19.0 days, producing 27.33 L of drinking water compliant with EPA standards (As concentration <10 µg/L), which suggested that AmPSd-Ti is an ideal packing material for drinking water filtration devices [76].

  Figure 7

  

  Figure 7.  The calculated treatment capacity of AmPSd-Ti for achieving compliance with the established standard limits for (A) As and (B) Sb. (C) Fixed-bed column experimental setup.

  DownLoad: Full-Size Img PowerPoint

  The removal capacities toward As and Sb of AmPSd-Ti are among the highest in the previously reported works. For example, a carbon block module impregnated with amorphous titanium (hydro)xide has a bed volume of 10,000 and an adsorption capacity of 0.6 mg/g for As-contaminated water under similar conditions [77]. A magnetic hydrochar adsorbent derived from mushroom residue was reported to be capable of purifying 1.31 L wastewater containing 5.62 mg/L Sb(Ⅴ) [78]. The breakthrough curves were well fitted by the Thomas model (0.98 and 0.97 R² for sample 1 and sample 2, respectively, Table S9 in Supporting information), implying that the adsorption process was controlled by the mass transfer process between phases. The effluent of samples 1 and 2 reached equilibrium at 18,600 and 5900 BV, respectively. These findings provide critical information for developing AmPSd-Ti adsorbates and technologies for the highly efficient removal of As and Sb in complex matrices.

  In summary, a novel composite adsorbent, nano TiO₂-loaded aminated polystyrene crosslinked with divinylbenzene (AmPSd-Ti), via the environmentally friendly Friedal-Crafts acylation and the Mannich reaction, was fabricated. The mm-scale AmPSd-Ti was characterized by complementary techniques. Its high adsorption capacities for As and Sb were confirmed by batch and fixed-bed column experiments using simulated and real environmental waters. The results of adsorption kinetics, isotherms, XPS analysis, and molecular dynamics simulations revealed that the aminated PSd served as a stable substrate for nano TiO₂ through N-Ti bonds and the protonated amine group enhanced the adsorption of pentavalent oxyanions, H₂AsO₅⁻ and Sb(OH)₆⁻, by electrostatic attraction and hydrogen bonding. Nano-sized TiO₂ stabilized in the polystyrene framework exhibited high affinity for As and Sb via inner-sphere complexation. Our study provides insights into the design of high-performance organic-inorganic composite adsorbents for the removal of heavy-metal oxyanion pollutants in real-world scenarios. Although AmPSd-Ti exhibits high selectivity in adsorbing As and Sb in the presence of Cl⁻, NO₃⁻, SO₄²⁻, and F⁻, the competing effects of other compounds in complex matrices, such as organic matter, and the underlying mechanisms remain unexplored. Our future work will explore strategies such as surface modification or optimization of adsorption conditions to enhance the selectivity for target contaminants in complex matrices.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Xiao Yang: Writing – original draft, Formal analysis, Data curation, Conceptualization. Wenjing Liu: Conceptualization, Formal analysis, Methodology, Writing – original draft. Jiarui Kong: Visualization. Xiangcheng Shan: Writing – original draft, Methodology, Formal analysis, Conceptualization. Qiupei Lei: Investigation. Zhipeng Yin: Writing – original draft, Investigation, Formal analysis, Data curation. Runzeng Liu: Writing – review & editing. Min Zhang: Writing – review & editing. Qingzhe Zhang: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Yongguang Yin: Writing – review & editing. Chuanyong Jing: Conceptualization. Yong Cai: Writing – review & editing, Supervision.

  Acknowledgments

  We acknowledge the financial support of the National Natural Science Foundation of China (No. 42230706), the Outstanding Youth Science Fund (Overseas) of Shandong Provincial Natural Science Foundation (No. 2022HWYQ-015), the Taishan Scholars Project Special Fund (No. tsqn202211039), and Qilu Youth Talent Program of Shandong University (No. 61440082163171). We appreciate the ICP-MS measurements assisted by Xue Li from Analytical Testing Center, School of Environmental Science and Engineering, Shandong University.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110856.

  
  
• 1. [1]

     J. Podgorski, M. Berg, Science 368 (2020) 845-850. doi: 10.1126/science.aba1510

  2. [2]

     L. Yan, T. Chan, C. Jing, Environ. Sci. Technol. 56 (2022) 3138-3146. doi: 10.1021/acs.est.1c07801

  3. [3]

     X. Cai, L.K. ThomasArrigo, X. Fang, et al., Environ. Sci. Technol. 55 (2021) 1319-1328. doi: 10.1021/acs.est.0c05329

  4. [4]

     Y. Meng, J.N. Wang, C. Cheng, et al., Chin. Chem. Lett. 23 (2012) 863-866.

  5. [5]

     Q. Zhang, X. Tan, T. Yu, Chin. Chem. Lett. 34 (2023) 107748.

  6. [6]

     T. Davydiuk, J. Tao, X. Lu, et al., Environ. Health 1 (2023) 236-248. doi: 10.1021/envhealth.3c00090

  7. [7]

     Y. Zhang, C. Ding, D. Gong, et al., Environ. Technol. Innov. 24 (2021) 102026.

  8. [8]

     Y. Li, M. Zhang, R. Xu, et al., Environ. Int. 153 (2021) 106522.

  9. [9]

     M.A. Dovick, R.S. Arkle, T.R. Kulp, et al., Environ. Sci. Technol. 54 (2020) 7983-7991. doi: 10.1021/acs.est.0c00558

  10. [10]

      K. Wan, T. Fang, W. Zhang, et al., Chem. Eng. J. 465 (2023) 143018.

  11. [11]

      M. Yan, Y. Xi, N. Jiang, et al., J. Membr. Sci. 703 (2024) 122847.

  12. [12]

      X. Li, J. Fan, F. Zhu, et al., J. Hazard. Mater. 471 (2024) 134302.

  13. [13]

      M. Mitrakas, Z. Mantha, N. Tzollas, et al., Water 10 (2018) 1328-1339. doi: 10.3390/w10101328

  14. [14]

      Y. Liu, C. Li, Z. Lou, et al., Sep. Purif. Technol. 275 (2021) 119037.

  15. [15]

      X. Zhu, W. Jiang, W. Cui, et al., Chin. Chem. Lett. 30 (2019) 1133-1136.

  16. [16]

      A.P. Panda, P. Rout, K.K. Jena, et al., J. Mater. Chem. A 7 (2019) 9933-9947. doi: 10.1039/c9ta00428a

  17. [17]

      S. Fajal, W. Mandal, S. Mollick, et al., Angew. Chem. Int. Ed. 61 (2022) e202203385.

  18. [18]

      S. Hu, Q. Shi, C. Jing, Environ. Sci. Technol. 49 (2015) 9707-9713. doi: 10.1021/acs.est.5b01520

  19. [19]

      L. Yan, J. Song, T. Chan, et al., Environ. Sci. Technol. 51 (2017) 6335-6341. doi: 10.1021/acs.est.7b00807

  20. [20]

      T.T. Lim, P.S. Yap, M. Srinivasan, et al., Crit. Rev. Environ. Sci. Technol. 41 (2011) 1173-1230. doi: 10.1080/10643380903488664

  21. [21]

      M.E. Pena, G.P. Korfiatis, M. Patel, et al., Water Res. 39 (2005) 2327-2337.

  22. [22]

      T. Zhao, Z. Tang, X. Zhao, et al., Environ. Sci. 6 (2019) 834-850. doi: 10.1039/c8en00869h

  23. [23]

      S.M. Miller, J.B. Zimmerman, Water Res. 44 (2010) 5722-5729.

  24. [24]

      Y.X. Ma, D. Xing, C.P. Lu, et al., Polym. Compos. 39 (2018) 2232-2240. doi: 10.1002/pc.24187

  25. [25]

      M. Wang, M. Fu, J. Li, et al., Chin. Chem. Lett. 35 (2024) 108442.

  26. [26]

      O.S. Chowdhury, P.J. Schmidt, W.B. Anderson, et al., Environ. Health 2 (2024) 441-452. doi: 10.1021/envhealth.3c00174

  27. [27]

      S. Wang, N.S. Al-Hasni, Z. Liu, et al., Environ. Health 2 (2024) 688-701.

  28. [28]

      L. Zhang, S. Lou, X. Hao, et al., Sep. Purif. Technol. 303 (2022) 122253.

  29. [29]

      S. Liu, A. Motta, M. Delferro, et al., J. Am. Chem. Soc. 135 (2013) 8830-8833. doi: 10.1021/ja4039505

  30. [30]

      K.A. Gebru, C. Das, Chemosphere 191 (2018) 673-684.

  31. [31]

      E. Bayer, X.N. Liu, U. Tallarek, et al., Polym. Bull. 37 (1996) 565-572.

  32. [32]

      Q. Teng, S. Ma, M. Ni, et al., Water. Sci. Technol. 82 (2020) 440-453.

  33. [33]

      A.K. Rappe, C.J. Casewit, K.S. Colwell, et al., J. Am. Chem. Soc. 114 (1992) 10024-10035. doi: 10.1021/ja00051a040

  34. [34]

      T. Lu, F. Chen, J. Comput. Chem. 33 (2012) 580-592. doi: 10.1002/jcc.22885

  35. [35]

      M.J. Abraham, T. Murtola, R. Schulz, et al., SoftwareX 1-2 (2015) 19-25.

  36. [36]

      B. Hess, H. Bekker, H.J.C. Berendsen, et al., J. Comput. Chem. 18 (1997) 1463-1472.

  37. [37]

      W. Humphrey, A. Dalke, K. Schulten, J. Mol. Graph. 14 (1996) 33-38.

  38. [38]

      Q. Shi, Y. Mi, S. Wang, et al., Fuel 358 (2024) 130250.

  39. [39]

      A. Kawasaki, J. Furukawa, T. Tsuruta, et al., Polymer 1 (1960) 315-329.

  40. [40]

      H. Ning, Y. Fan, H. Liu, et al., J. Chromatogr. A 1671 (2022) 462994.

  41. [41]

      X.N. Liu, E. Bayer, G. Xue, Spectrosc. Lett. 30 (1997) 289-295. doi: 10.1080/00387019708006988

  42. [42]

      P. Liu, Y.B. Hu, X.Y. Li, et al., Angew. Chem,. Int. Ed. 61 (2022) e202208587.

  43. [43]

      Y. Zhao, X. Shan, Q. An, et al., Chem. Eng. J. 398 (2020) 125561.

  44. [44]

      A. Modwi, N.M. Basith, M.G. Ghoniem, et al., Mater. Sci. Eng. B 289 (2023) 116191.

  45. [45]

      L. Ma, L.-J. Yu, J. Liu, et al., Energy Storage Mater. 44 (2022) 180-189. doi: 10.3390/separations9070180

  46. [46]

      V. Etacheri, M.K. Seery, S.J. Hinder, et al., Chem. Mater. 22 (2010) 3843-3853. doi: 10.1021/cm903260f

  47. [47]

      X. Zeng, L. Huang, C. Wang, et al., ACS Appl. Mater. Interfaces 8 (2016) 20274-20282. doi: 10.1021/acsami.6b05746

  48. [48]

      N.T. Nolan, D.W. Synnott, M.K. Seery, et al., J. Hazard. Mater. 211-212 (2012) 88-94.

  49. [49]

      Y. Liu, M. Ren, X. Zhang, et al., Appl. Surf. Sci. 614 (2023) 156118.

  50. [50]

      S. Wang, Z. Xiao, S. Zhai, et al., J. Mater. Chem. A 7 (2019) 17345-17356. doi: 10.1039/c9ta05145g

  51. [51]

      Y. Wang, J. Liang, S. Liu, et al., Langmuir 40 (2024) 9265-9279. doi: 10.1021/acs.langmuir.4c00702

  52. [52]

      Y. Cao, Q. Guo, M. Liang, et al., Appl. Surf. Sci. 196 (2020) 105766.

  53. [53]

      B. Liu, Z. Liu, H. Wu, et al., Sci. Total Environ. 742 (2020) 140508.

  54. [54]

      X. Shan, L. Yang, Y. Zhao, et al., J. Colloid Interface Sci. 606 (2022) 736-747.

  55. [55]

      M.J. Ahmed, S.K. Theydan, J. Anal. Appl. Pyrolysis 99 (2013) 101-109.

  56. [56]

      S.G. Johnston, W.W. Bennett, N. Doriean, et al., Sci. Total Environ. 710 (2020) 136354.

  57. [57]

      G. Okkenhaug, Y.G. Zhu, J. He, et al., Environ. Sci. Technol. 46 (2012) 3155-3162. doi: 10.1021/es2022472

  58. [58]

      M. He, N. Wang, X. Long, et al., J. Environ. Sci. 75 (2019) 14-39.

  59. [59]

      L. Yan, S. Hu, C. Jing, J. Environ. Sci. 49 (2016) 74-85.

  60. [60]

      M. Filella, N. Belzile, Y.-W. Chen, Earth-Sci. Rev. 57 (2002) 125-176.

  61. [61]

      M.A. Dovick, T.R. Kulp, R.S. Arkle, et al., Environ. Chem. 13 (2016) 149-159.

  62. [62]

      H. Wang, C. Qiao, C. Chen, et al., Chin. Chem. Lett. 36 (2025) 110244. doi: 10.1016/j.cclet.2024.110244

  63. [63]

      G.M. Walker, L. Hansen, J.A. Hanna, et al., Water Res. 37 (2003) 2081-2089.

  64. [64]

      Z. Wang, X. Bi, X. He, et al., Water Res. 228 (2023) 119290.

  65. [65]

      D. Mitoraj, H. Kisch, Angew. Chem. Int. Ed. 47 (2008) 9975-9978. doi: 10.1002/anie.200800304

  66. [66]

      H. Guo, Q. Fei, M. Lian, et al., Adv. Mater. 35 (2023) e2301418.

  67. [67]

      L. Ma, S.X. Tu, Environ. Chem. Lett. 9 (2011) 465-472. doi: 10.1007/s10311-010-0303-1

  68. [68]

      X. Song, L. Nong, Q. Zhang, et al., J. Environ. Chem. Eng. 11 (2023) 110874.

  69. [69]

      K. Liu, Z. Huang, J. Dai, et al., Chem. Eng. J. 382 (2020) 122775.

  70. [70]

      L. Xiaojing, T. Wang, M. He, Environ. Tech. 44 (2022) 1-33. doi: 10.21656/1000-0887.420388

  71. [71]

      L. Zhang, R. Li, L. Zhou, et al., Environ. Technol. Innov. 24 (2021) 102001.

  72. [72]

      X. Long, T. Wang, M. He, Environ. Technol. 44 (2023) 2913-2923. doi: 10.1080/09593330.2022.2048084

  73. [73]

      J. Zheng, L. Li, S. Chen, et al., Langmuir 20 (2004) 8931-8938.

  74. [74]

      National Primary Drinking Water StandardsCriteria and Standards Division, U.S. Environmental Protection Agency, Washington DC, 2009.

  75. [75]

      Standardization Administration of ChinaEmission Standards of Pollutants for Stannum, Antimony, and Mercury Industries, Science Press of the Chinese Academy of Sciences, GB 30770-2014, Beijing, 2014.

  76. [76]

      J. Chen, L. Hu, Q. Chen, et al., Surf. Interfaces 46 (2024) 104184.

  77. [77]

      A. Farsad, M. Marcos-Hernandez, S. Sinha, et al., Environ. Sci. Technol. 57 (2023) 20410–20420. doi: 10.1021/acs.est.3c06586

  78. [78]

      R. Duan, L. Tian, Y. Liu, et al., Sep. Purif. Technol. 343 (2024) 127187.

• 

  Scheme 1  The fabrication procedure of AmPSd-Ti.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  (A) Scheme of the fixed-bed column experiments. (B) Optical photograph of multiple AmPSd-Ti spheres. (C, D) Cross-section SEM image in an overview of an individual sphere and the enlarged view of the selected area. (E) EDS line scan analysis of AmPSd-Ti with adsorbed Sb(Ⅴ) and elemental distribution of (F) C, (G) O, (H) N, (I) Ti, and (J) Sb. (K) FTIR spectra of synthesized PSd, AcPSd, AmPSd, and AmPSd-Ti. (L) High-resolution XPS spectra of synthesized TiO₂ and AmPSd-Ti. (M) XRD patterns of Synthesized TiO₂, AmPSd, AmPSd-Ti, and AmPSd-Ti after As/Sb adsorption.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Adsorption isotherms of AmPSd-Ti on (A) As and (B) Sb, and fitted Freundlich adsorption isotherm model lines. Dosage: 0.25 g/L, initial concentration: 1–100 mg/L, initial pH: 6.9 ± 0.1, contact time: 24 h, ionic strength: 0.05 mol/L.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (A, C, E, G) Fitted pseudo-first-order, pseudo-second-order adsorption kinetic curves and (B, D, F, H) liquid film diffusion model of arsenic and antimony adsorption on AmPSd-Ti. Dosage: 0.25 g/L, initial concentration: the red and blue legends represent 0.5 and 1 mg/L, respectively, initial pH: 6.9 ± 0.1, contact time: 0–24 h, ionic strength: 0.05 mol/L.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  High-resolution scan of (A) Ti 2p, (B) O 1s, and (C) N 1s core level of AmPSd-Ti before and after As(Ⅲ/Ⅴ) and Sb(Ⅲ/Ⅴ) adsorption.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (A) Adsorption capacities of As(Ⅲ/Ⅴ) and Sb(Ⅲ/Ⅴ) on AmPSd-Ti at different initial pH (Dosage: 0.25 g/L, initial concentration: 1 mg/L, initial pH: 2.0–12.0, contact time: 24 h, ionic strength: 0.05 mol/L). (B) Zeta potentials of AmPSd-Ti at different pH before and after As(Ⅲ/Ⅴ) and Sb(Ⅲ/Ⅴ) adsorption. (C–F) Molecular dynamics simulation of the adsorbate-AmPSd systems.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  The adsorption capacities of As (A, B) and Sb (C, D) onto AmPSd-Ti in the presence of various coexisting anions (Dosage: 0.25 g/L, initial concentration: 1 mg/L, initial pH: 6.9± 0.1, contact time: 24 h, ionic strength: 0.05 mol/L). The red dashed line represents the adsorption capacity of AmPSd-Ti for As(Ⅲ/Ⅴ) and Sb(Ⅲ/Ⅴ) without the addition of coexisting anions.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  The calculated treatment capacity of AmPSd-Ti for achieving compliance with the established standard limits for (A) As and (B) Sb. (C) Fixed-bed column experimental setup.

  下载: 全尺寸图片 幻灯片
•