International Journal of Biological Macromolecules 269 (2024) 131851Available online 29 April 20240141-8130/© 2024 Elsevier B.V. All rights reserved.An eco-friendly porous hydrogel adsorbent based on dextran/phosphate/ amino for efficient removal of Be(II) from aqueous solution Xu Zhaoa, Qingliang Wangb, Yige Sunc, Haoshuai Lib, Zhiwu Leib, Boyuan Zhengb, Hongyang Xiab, Yucheng Sub, Kham Muhammad Yaruq Alib, Hongqiang Wangb,c, Fang Hub,* aSchool of Nuclear Science and Technology, University of South China, Hengyang 421001, Hunan, China bSchool of Resource Environment and Safety Engineering, University of South China, Hengyang 421001, Hunan, China cCollege of Resources and Environment, Anhui Agricultural University, Hefei 230000, Anhui, China ARTICLE INFO Keywords: Dextran Selectivity Beryllium Adsorption ABSTRACT A novel environmentally-friendly porous hydrogel adsorbent (GHPN) is firstly designed and prepared using dextran, phosphate, and calcium hydroxide for the adsorption of Be(II). GHPN shows good adsorption selectivity for Be(II) (Kd =1.53 ×104 mL/g). According the adsorption kinetics and thermodynamics, the theoretical adsorption capacity of GHPN to Be(II) is 43.75 mg/g (35 ◦C, pH =6.5), indicating a spontaneous exothermic reaction. After being reused for 5 cycles, the adsorption and desorption efficiencies of Be(II) with GHPN are obtained to be more than 80 %, showing acceptable recycling performance. Both of the characterizations and theoretical calculations indicate that the phosphate group, hydroxyl group, and amino group own the affinity to form stable complexes with Be(II). Benefiting from the introduction of phosphate and amino, the adsorption effect of the hydrogel adsorbent on Be(II) can be greatly improved, and surface precipitation, complexation, and ligand exchange are the dominant mechanisms of beryllium adsorption. The results suggest that GHPN has great potential to be utilized as an eco-friendly and useful adsorbent of Be(II) from aqueous solution. 1.Introduction Beryllium is employed in aerospace and nuclear power plants due to its brilliant properties [1]. With the progress of human exploration in outer space and the use of beryllium alloy, beryllium metal is consumed in large quantities, intensifying the extraction of beryllium. During the beryllium ore mining process, beryllium inevitably migrates into the environment in various ways, especially in the aquatic environment. At least 8.58412.9 mg/kg of beryllium is imported into the water environment annually through human activities and atmospheric sedimentation. Soils in the southeastern coastal plain and Piedmont region of the United States contain up to 30.5 mg/kg [USGS, 2020]. Zhong et al. [2] deeply investigated the migration of beryllium in the environment. The results of studies revealed that about 202 mg/L of beryllium migrated to the aquatic environment. It should be emphasized that beryllium and its compounds have significant toxicity [3]. Long-term exposure to beryllium would cause people to suffer from diseases such as lung cancer, which is listed as a Class A carcinogen by the World Health Organization [4]. Beryllium is also very harmful to plants and when beryllium migrates from water to soil, it gradually prevents plants from absorbing nutrients until they die [5]. China stipulates that the discharge of beryllium in industrial wastewater should be less than 5 ×106 g/L. Therefore, in order to resolve beryllium pollution problem, it is necessary to efficiently eliminate beryllium from aqueous solution. Beryllium exists mainly as a cation in solutions [6]. However, the complex composition and low concentration of beryllium-containing aqueous solutions make beryllium adsorption challenging [7]. Among the reported methodologies, the adsorption approach has been extensively utilized due to its simple operation and low price [810]. Because of its effectiveness, several novel adsorbent systems have been developed to remove Be(II) from beryllium-containing aqueous solution [6]. It’s well known that hydrogels have high porosity, and the pores can be sufficiently contacted with pollutants, thereby improving the waste treatment effectiveness [11]. Recently, hydrogel has been considered as a green and efficient adsorbent and utilized to remove pollutants from aqueous solutions. For instance, Xin et al. [12] prepared phosphate chitosan (SPCCHC) for uranium adsorption, and the Qe of SPCCHC for U (VI) was reported to be 579.27 mg/g. The phosphate ligand in SPCCHC *Corresponding author. E-mail address: csuhufang@163.com (F. Hu). Contents lists available at ScienceDirect International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac https://doi.org/10.1016/j.ijbiomac.2024.131851 Received 25 January 2024; Received in revised form 5 April 2024; Accepted 23 April 2024

International Journal of Biological Macromolecules 269 (2024) 1318512increased the adsorption capacity and selectivity of U(VI). Wu et al. [13] prepared a novel iron-modified gel ball for heavy metal adsorption, and the gel pellet demonstrated an excellent adsorption effect for Cu(II) (Qe =55.96 mg/g), Cd(II) (Qe =86.28 mg/g), and Pb(II) (Qe =189.04 mg/g). However, to the best of the authors’ knowledge, few study has been devoted to the exploitation of hydrogels as adsorbents for the selective adsorption of Be(II) from aqueous solution. Dextran is often utilized to prepare gel materials because it has multiple hydroxyl groups, like other polysaccharide derivatives [12]. It was reported that the OH bond in dextran was also able to effectively form a stable complex with Be(II) [6]. To improve the mechanical strength of dextran and facilitate adsorption, dextran was usually made into porous composites such as hydrogels [14]. By preparing hydrogels, the mechanical strength and adsorption capacity of dextran could be substantially enhanced [15]. Previous studies indicated that O–H, amino and phosphoric acid groups had a synergistic effect to improve the selective adsorption capacity of dextran [16,17]. Based on this concept, a porous hydrogel adsorbent is designed, and creatively used for the adsorption of Be(II) from aqueous solution. In the present work, the porous hydrogel adsorbent (GHPN) with outstanding selective adsorption performance was prepared by using dextran, phosphate, and calcium hydroxide. The adsorption performance of GHPN for Be(II) was appropriately optimized by changing the composition and experimental conditions, such as the amount of phosphoric acid, the amount of GHPN, the effect of initial pH, temperature, and ion concentration. The adsorption kinetics were evaluated using pseudo-first-order (PFO), pseudo-second-order (PSO), and intraparticle diffusion (IPD) models, and the adsorption isotherms were examined with Langmuir, Freundlich, and Temkin models. In addition, the reusability of the GHPN was also methodically investigated by using adsorption-desorption cycle. Crystallinity, chemical structure, thermal, and surface topography of GHPN before and after adsorption were then evaluated through XRD, XPS, TG, and SEM-EDS methods. The adsorption mechanism of Be(II) by GHPN was explored by theoretical approach such as relativistic density functional theory (DFT) calculations. Finally, the adsorption mechanism of beryllium adsorption with GHPN was discussed in some detail. 2.Experimental materials and methods 2.1.Materials Dextran (C6nH10nO5n, purity: ~95 %, molecular weight: 10000), N, N-Methylenebisacrylamide Bis-acrylamide (C7H10N2O2, purity: ~99 %), Acrylamide (C3H5NO, purity: ~95 %), Dimethylammonium chloride (purity: ~98 %), Ammonium persulfate (H8N2O8S2, purity: ~98.5 %), and Tetramethylethylenediamine (C6H16N2, purity: ~99 %) were purchased from Shanghai McLean Biochemical Co., LTD. 2.2.Preparation of GHPN Typically, 15 mL of dextran solution (2 g/L) was first introduced into a 50 mL beaker at 25 ◦C. Then, under constant stirring at 250 rpm, 0.6 g of acrylamide, 1.0 g of diallyl dimethylammonium chloride, 1 mL of ammonium persulfate, 0.1 mL of tetramethylethylenediamine, and 10 mL of deionized water were continuously added until the solution was clarified. Then, 15 mL of methylene bisacrylamide (2 % v/v) was added to the beer glass and the mixture was stirred at 250 rpm for 3 min. The mixture was labeled as Mixture C (GH). After pouring 1 g of phosphoric acid and 1 g of Ca(OH)2, Mixture E was obtained and subsequently transferred to a conical flask, and polymerized in a constant temperature vibrating oscillator at 25 ◦C for 6 h. Then, the as-prepared hydrogel was immersed in a deionized solution for 72 h to remove any soluble polymers, unreacted monomers and initiators, and other impurities. The hydrogel (marked as Mixture F) was immersed in a mixture solution with 20 mL NH3⋅H2O (10 %) and 5 mL (8 %) NaOH solution for 10 h. After freeze-dried, GHPN was prepared for later use (Fig. 1). 2.3.Batch adsorption experiments In this paper, a single beryllium solution with concentrations of 15 mg/L was prepared and employed for batch experiments. The effect of WH3PO4:WDextran (wt:wt), amount of GHPN, and initial pH on the adsorption rate of Be(II) were investigated. The concentrations of beryllium in the solutions before and after adsorption were measured and the corresponding adsorption rate was calculated (see Eqs. (S1) and (S2)). 2.4.Recycling performance of GHPN To test the recycling performance of GHPN for beryllium, the GHPN after adsorption of 5- and 10-mg/L beryllium solutions was eluted with 25 mL NaOH (8 %) at 25 ◦C for 4 h. The desorption rate of GHPN was appropriately evaluated (see Eqs. (S3) and (S4)). The adsorbent after each desorption was washed twice with deionized water, and then used for the next beryllium adsorption-desorption cycle. 2.5.Adsorption selectivity experiments Previous research [18] indicated that uranium-beryllium ore usually contained elements such as Al, Zn, Mn, Na, Mg, U and Be. To explore the adsorption selectivity of GHPN, a simulated solution containing the mentioned elements with concentration of 30 mg/L was utilized. Besides, to further estimate the competitive adsorption effect of the coexisting ions, six binary solutions containing Be(II) with fixed concentration (15 mg/L) and another impurity ion with varied concentrations (1050 mg/L) were used for selective experiments. To calculate the adsorption selectivity of GHPN for beryllium, the dispersion coefficient Kd (mL/g) was used. (see Eq. (1)) [12]. Kd=(C0−Ce)×VCe×m×1000(1) where C0 (mg/L) and Ce (mg/L) represent the concentration of Be(II) before and after adsorption, respectively, V (L) represents the volume of the solution, and m (g) represents the dry weight of GHPN. 2.6.Adsorption kinetics and thermodynamics At 25 ◦C, the effect of time on the adsorption rate was studied and the data were fitted to different kinetic models (see Eqs. (S5)(S7)). The thermodynamics of GHPN on the adsorption of solutions with different initial concentrations of 1200 mg/L Be(II) were conducted at 15 ◦C, 25 ◦C, and 35 ◦C. Langmuir, Freundlich, and Temkin models (Eqs. (S8)– (S10)) were mainly used to nonlinear fit the results of the adsorption of Be(II). The Gibbs free energy (ΔG0), enthalpy (ΔH0), and entropy (ΔS0) (see Eqs. (S11)(S14)) at various concentrations were calculated to determine the adsorption type of Be(II) by GHPN. 2.7.Characterizations The specific surface area and pore size of GHPN was investigated by Brunauer, Emmett and Teller (BET, Micromeritics, ASAP2020, America). The surface morphology of GHPN was investigated by scanning electron microscopy (SEM) (Zeiss sigma 300, Germany), meanwhile, the element distribution was recorded by energy dispersive X-ray spectroscopy (EDX). Fourier transform infrared spectroscopy (FTIR) (PE FTIR Frontier, USA) was used to investigate the surface functional groups of GHPN with a wavelength range of 4000400 cm1. The zeta potentials of GHPN at different pH values were measured with a Zetasizer Nano zs90 instrument. The composition of the samples was determined by Xray diffraction (XRD, D8 advanced x-ray diffractometer). X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 1318513Thermogravimetric analysis (TGA) was carried out on a TG thermogravimetric analyzer (TG/DTA7300) at heating efficiency of 10 ◦Cmin1 between 25 ◦C and 800 ◦C in air. X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific KAlpha) was used to analyze the sample’s surface chemical state, and the binding energy of the spectrum is normalized by the C 1s peak at 284.8 eV. 2.8.DFT calculation details The dynamic process of conformation search was completed on xtb, Fig. 1.(a) Preparation of hydrogel adsorbent precursor (GH). (b) Preparation of hydrogel adsorbent (GHPN). Fig. 2.BET analysis of GHPN X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 1318514the open source software developed by Grimme et al. [19,20]. After the dynamic process, Multiwfn software developed by Tian Lu et al. [21] was used for structural screening, and the ten configurations with the lowest energy were obtained. The calculations were performed by using relativistic density functional theory (DFT) with basis set m062x/6- 311++g(28d,p), as implemented in Gaussian 16 software package. And D3 correction was added to improve its accuracy for describing the weak interaction. The implicit solvent model IEFPCM was used in solvation model. The geometric configurations were optimized in the aqueous phase. After the structure is optimized, the frequency is calculated to verify the rationality of the structure. Based on the optimized structure, the binding energy is calculated using the Eq. (2). ΔE=EBerylliumcompound(EBe+ELigand)(2) where, ΔE (kcal/mol) represents the binding energy between Be(II) and the ligand. EBeryllium compound represents the zero-point energy of beryllium compound. EBe (kcal/mol) represents the zero-point energy of Be (II). ELigand (kcal/mol) represents the zero-point energy of the ligand. 3.Results and discussion 3.1.BET analysis The specific surface area and pore size of GHPN were analyzed by BET. The corresponding results are presented in Fig. 2. The results reveal that the specific surface area of GHPN is 78.7307 m2/g, the pore volume is 0.064786 cm3/g, and the average pore size is 4.7129 nm, indicating the existence of mesopores. In addition, it can be seen that the N2 curve of GHPN is a reversible Type II isotherms, and the gradual curvature at Point B indicates the overlap of monolayer coverage and the onset of multilayer adsorption [7]. 3.2.Batch experiments The influence of different weight ratio of phosphoric acid and dextran on Be(II) adsorption by GHPN were examined by varying WH3PO4WDextran (wt:wt) from 0.5 to 2.5. The results are shown in Fig. 3 (a). From Fig. 3(a), it is obvious that the adsorption capacity (Qe) and adsorption rate of GHPN are enhanced, when WH3PO4:WDextran is in the range of 0.5 to 1. The optimal adsorption rate of Be(II) (99 %) is obtained at WH3PO4WDextran =1, and the corresponding adsorption capacity are approx. 12 mg/g. After that, the adsorption rate of GHPN decrease with the increase of WH3PO4:WDextran. Therefore, WH3PO4:WDextran is selected as 1 for preparing GHPN. Fig. 3(b) shows the influence of different GHPN dosage on the adsorption of Be(II). When the dosage of GHPN is 0 to 1 g/L, the adsorption rate of Be(II) by GHPN increases, while the adsorption capacity of Be(II) by GHPN decreases. At the GHPN dosage of 1 g/L, the adsorption rate reaches 99 % and the adsorption capacity is 12.5 mg/g. When the GHPN dosage is higher than 1 g/L, the adsorption rate of Be (II) tends to be stable. However, it is observed the adsorption capacity of Be(II) GHPN gradually decreases. This phenomenon can be attributed to the limited amount of Be(II) available in the solution, and 1 g/L of adsorbent is able to effectively remove Be(II) from aqueous solutions [22]. Therefore, the amount of GHPN is set as 1 g/L. pH value can impact the adsorption process by altering the speciation of beryllium in the solution as well as the properties of GHPN. When the initial pH is 0 to 5.5, the main beryllium species in the solution is Be2+. When the initial pH is between 5.5 and 10, beryllium will produce Be(OH)2 precipitate and exist in the form of Be(OH)+[6]. When the initial pH is greater than 10, Be(OH)2 precipitate dissolves as Be(OH)n2-n in the solution [7]. Several studies have highlighted the influences of solution initial pH on adsorption processes. In Fig. 3(c), the adsorption capacity of Be(II) increases from 6 mg/g at pH of 2.5 to 12 mg/g at pH of 5.5. Meanwhile, the adsorption rate of Be(II) reaches increases from 50 % to 99 %. When the pH is higher than 5.5, both of the adsorption Fig. 3.The influence of WH3PO4:WDextran (wt:wt) (a), dosage of GHPN (b), initial pH (c) on adsorption of Be(II), the Zeta potential (d) and final pH (e) at different initial pH, adsorption-desorption cycle with initial concentrations of 5 mg/L (f) and 10 mg/L (g) Be(II), selective adsorption of Be(II) from simulated solution (h) and binary solutions(i). X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 1318515capacity and adsorption rate of Be(II) by GHPN exhibits a decreasing trend. This phenomenon may be related to the zeta potential of GHPN. Fig. 3(d) shows the Zeta potential values at different initial pH. It can be seen that the pHzpc of GHPN is 5.2. When the pH is greater than pHzpc, the GHPN surface is positively charged. When the pH is lower than pHzpc, the surface of GHPN is positively charged, and the electrostatic repulsion with Be2+reduces the adsorption rate of GHPN [11]. When the pH increases, the deprotonation of the GHPN surface decreases, the surface becomes negatively charged, and the phosphate and amino groups are more likely to form stable complexes with Be2+[17]. When the initial pH is higher than 5.5, the amount of Be(OH)+in the solution increases, which impairs the adsorption capacity of GHPN. The final pH remains at around 7 when the initial pH is 5.5 (Fig. 3(e)). 3.3.Recycling performance of GHPN Fig. 3(f) and (g) present the recycling performance of GHPN at two different initial concentrations of beryllium. According to Fig. 3(f), both of the adsorption rate and desorption rate of Be(II) by GHPN remain almost stable for 5 mg/L Be(II) solution in 5 adsorption-desorption cycles. In Fig. 3(g), the adsorption rate and desorption rate of Be(II) by GHPN for 10 mg/L Be(II) solution decreased to 80 % in the first three adsorption processes, and then stabilized at around 80 %. Therefore, the adsorption and desorption performances of GHPN on 5 mg/L and 10 mg/L Be(II) solutions are basically stable after five adsorptiondesorption cycles, which indicates that GHPN has acceptable recycling performance. 3.4.Selective adsorption studies Fig. 3(h) illustrates the selective adsorption of Be(II) by GHPN from the simulated solution, which containing Be(II), U(VI), Zn(II), Mg(II), Na (I), Mn(II), and Al(III). It is obvious that the maximum adsorption rate of Be(II) is up to 99 %, which is much higher than other impurity ions, i.e., U(VI) (20 %), Zn(II) (4 %), Mg(II) (10 %), Na(I) (6 %), Mn(II) (7 %), and Al(III) (12 %). To further explore the influence of the coexisting impurity ions on selective adsorption of Be(II) by GHPN, binary solutions with different coexisting impurity ions concentration (1050 mg/L) were used to further examine the adsorption selectivity of GHPN. The results are presented in Fig. 3(i). It can be seen that there is no obvious impact on the adsorption of Be(II) by GHPN even though the concentration of the coexisting impurity ions increase in the experimental range. The adsorption rates of Be(II) by GHPN from the binary solutions vary from 97 % to 99 %, which is close to the adsorption rate of Be(II) from single beryllium solution under same adsorption conditions (98.5 %). To quantitatively evaluate the influence of the coexisting impurity ions on the adsorption selectivity of GHPN, the dispersion coefficient Kd were calculated based on the data from Fig. 3(i). Fig. 4 shows the influence of U(VI), Zn(II), Mg(II), Na(I), Mn(II), and Al(III) with different concentrations on the Kd of Be(II) by GHPN. As for Al(III), it is found that the maximum Kd of Be(II) is 1.63 ×104 mL/g, and the minimum Kd is 1.44 ×104 mL/g (Fig. 4(a)). Kd of Be(II) by GHPN exhibits an increasing trend when the concentration of Al(III) increases. The reason for this probably is due to co-precipitation. Moreover, from Fig. 4(b)(f), it can be inferred that U(VI), Zn(II), Mg(II), Na(I), Mn(II), and Al(III) have no significant influence on the adsorption selectivity of GHPN in the studied concentration range. Basically, it can be concluded that GHPN exhibits strong adsorption selectivity for Be(II), and the coexisting ions hardly affect the adsorption performance of Be(II) by GHPN. 3.5.Adsorption kinetics Fig. 5(a), (b) and (c) display the kinetics fitting curves with the pseudo-first order, pseudo-second order, and intra-particle diffusion models. Fig. 5(a) shows that the effect of adsorption time on adsorption of Be(II) by GHPN. The adsorption reaches equilibrium at about 120 min. At early stage, the adsorption capacity of Be(II) by GHPN increases rapidly due to the existence of more active sites on the GHPN surface, and a mass of active sites are easier to be occupied by Be(II) and its compounds [23]. Table 1 presents the relevant kinetic parameters for the adsorption of Be(II) by GHPN with different kinetic models. By comparing the fitting coefficient R2 values (R12 =0.989, R22 =0.990) of the pseudo-first-order model, pseudo-second-order model, The results shows that the adsorption of Be(II) by GHPN is more consistent with the pseudo-second-order model, suggesting that the beryllium adsorption is probably governed by chemisorption [24]. In order to gain more information about the rate-controlling steps affecting the kinetics of adsorption, intra-particle diffusion model was applied to analyze the kinetic data [23]. It can be seen that the rate constant (kI1, kI2, and kI3) and the coefficient of determination (RI12, RI22, and RI32) for first phase for the adsorption of Be(II) on GHPN is greater than the second phase and third phase. According to linear regressive analysis if the plot of Qe versus t1/2 is linear and pass through the origin then intraparticle diffusion is sole limiting step. However, values of “C” for the adsorption of Be(II) by GHPN depicted in Table 1 are non zero. Thus, it can be predicted that the adsorption process may be of complex nature consisting of both surface adsorption and intraparticle diffusion [23]. It is suspected that the initial sharper stage is the external surface adsorption or the instantaneous adsorption of Be(II) by GPHN. The second stage is the gradual adsorption stage where intra-particle diffusion is rate-limiting [25]; and the third stage is the final equilibrium stage. The multi-linearity of the curves implies that both surface adsorption and intra-particle diffusion occur simultaneously [26]. Adsorption thermodynamics is able to effectively explain the energy change of the adsorption process [30]. Fig. 5(d), (e) and (f) show the adsorption isotherms for GHPN. The adsorption capacity of Be(II) increases with the rise of initial concentration Ce, which is probably because high starting concentrations may generate a motivation against mass transfer resistance. When Ce becomes higher than 100 mg/L, the adsorption capacity of Be(II) tends to be stable. This is mainly attributed to the fact that Be(II) gradually occupies the adsorption sites of GHPN at the early stage of adsorption, and the effective adsorption sites of GHPN can easily become saturated in the solutions with high concentrations [31]. Fig. 4.Influence of the coexisting impurity ions with different concentrations on the adsorption selectivity of GHPN. X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 1318516Table 2 presents the fitting isotherm parameters for the adsorption of Be(II) by GHPN with Langmuir, Freundlich, and Temkin thermodynamic models. By comparing the three thermodynamic fitting coefficient R2 values (RL2 =0.973, RF2 =0.991, RT2 =0.946), it is obvious that Freundlich model can describe the adsorption more effectively, and the adsorption process is multilayer adsorption. When 1/nF is less than 0.5, Be(II) is readily adsorbed by GHPN [32]. In addition, the fitting degree of the Temkin model is also high, indicating that GHPN has a high intermolecular force when adsorbing beryllium [24]. The bt value, which is less than 1, indicates that the adsorption reaction of Be(II) on GHPN is endothermic in the studied concentration range [32] (Table 3). The thermodynamic fitting parameters of the GHPN/Be(II) adsorption system are given in Table 4. The ΔG0 is −14.72 kJ/mol, which shows that the adsorption process of the adsorption of Be(II) by GHPN is Fig. 5.Adsorption kinetics of Be(II) by GHPN (a–c). Adsorption isotherms of Be(II) by GHPN (d–f). Table 1 Kinetic parameters for the beryllium adsorption by GHPN with different kinetic models. T Pseudo-first order Pseudo-second order Intra-particle diffusion model Qe1 (mg/g) k1 R12 Qe2 (mg/g) k2 R22 CI1 KI1 RI12 CI2 KI2 RI22 CI3 kI3 RI32 25 ◦C 11.29 0.064 0.989 12.07 0.008 0.990 0.432 1.884 0.982 7.57 0.349 0.946 11.15 0.017 0.721 Table 2 Isotherm parameters for the beryllium adsorption by GHPN with different thermodynamic models. T Langmuir Freundlich Temkin KL Qm RL2 KF 1/nF RF2 at bt RT2 15 ◦C 0.1190 28.75 0.973 14.29 0.177 0.978 67.68 0.001394144 0.927 25 ◦C 0.0042 38.21 0.975 15.99 0.185 0.980 78.87 0.001577686 0.927 35 ◦C 0.0050 43.75 0.668 16.97 0.195 0.991 57.69 0.001808091 0.946 Table 3 Thermodynamic parameters for the adsorption of Be(II) by GHPN. T (◦C) ln Kc ΔG0 (kJ/mol) ΔH0 (kJ/mol) ΔS0 [J/(mol⋅K)] 15 5.72 −13.72 0.507 44.61 25 5.93 14.72 0.523 46.13 35 6.05 −15.50 0.561 47.25 X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 1318517feasible and spontaneous at 25 ◦C [33]. The ΔH0 value (0.523 kJ/mol) is positive in the studied temperature range at 25 ◦C, which rationally indicates that the adsorption of Be(II) by GHPN is an endothermic process [34]. The ΔS0 values (46.13 J/(mol⋅K)) at 25 ◦C are positive, indicating that the randomness of the solid–liquid interface grows during the beryllium adsorption process [35]. In summary, the adsorption of Be(II) by GHPN is spontaneous and endothermic. The Langmuir model predict that the maximum adsorption capacity Qe of GHPN would reach up to 43.75 mg/g. Compared with some relevant studies listed in Table 4, the adsorption capacity of Be(II) by GHPN is closed to the adsorbents Fe-AC and Polystyrene-based chelating adsorbent, and higher than other known adsorbents. Table 4 Comparison of different adsorbents. Sample Fe-AC Al-AC Modified chitosan resin Polystyrene-based chelating adsorbent Aerobic granule GHPN pH 6.0 6.0 1.0 8.0 7.5–8 5.5 qm (mg/g) 45.685 32.86 44.96 22.5 14 43.75 Adsorption rate % 99 99 – 98 82 99 Reference [7] [27] [28] [29] [22] This study Fig. 6.SEM analysis (a–f) and EDS analysis (g–h). X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 13185183.6.Characterizations The surface changes of GHPN before and after adsorption can be visually observed using a scanning electron microscope (SEM). Fig. 6 shows the SEM images and energy dispersive spectrum (EDS) of GHPN before and after adsorption. According to Fig. 6(a)(c), it can be seen that the GHPN is a porous structure, and its surface exhibits mainly small and medium-sized holes. From Fig. 6(d)(f), many flocs can be found on the surface of GHPN after adsorption (named as Used-GHPN). Based on the results presented in Fig. 6(g) and (h), a large amount of oxygen element has accumulated in Used-GHPN, which can be inferred that many oxides are produced during the reaction. Because the beryllium itself is below the EDS detection line, the beryllium change could not be detected. The compositions of GHPN before and after adsorption were analyzed by X-ray diffraction (XRD). The corresponding results are presented in Fig. 7(a). The plotted results reveal that GHPN after adsorption produces multiple peaks at 2θ =11.5◦, 13.6◦, 15.2◦, 17.3◦, 21.5◦, and 23.7◦. The two peaks are pertinent to beryllium ammonia phosphate and beryllium hydroxide [7]. Thus, it is suspected the chelation and precipitation might be the main interactions during the beryllium adsorption process. Fourier transform infrared spectroscopy (FT-IR) was also implemented to analyze the group changes of GHPN before and after adsorption (in Fig. 7(b)). After adsorption, the N–H stretching vibration peak of GHPN at 1668.5 cm1 changes to 1618.3 cm1 [40]. The NH stretching vibration peak at 2900 cm1 is reduced and transferred [41,42], indicating that NH is involved in the adsorption process and forms a complex with Be(II) [43]. The intensity of O–H bond stretching vibration of GHPN after adsorption is increased at 1411.3 cm−1 [37], implying the formation of hydroxide. Both of the P–O bond vibration at 1155 cm−1 [44] and P-O-C bond vibration at 1066.3 cm−1 [31] GHPN Fig. 8.XRD analysis of GHPN and Used-GHPN (a), FT-IR analysis of GHPN and Used-GHPN (b), TG analysis of GHPN (c) and Used-GHPN (d). X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 1318519after adsorption are transferred, representing that a new substance might be formed. The bond vibration of GHPN at 885.6 cm1 is attributed to the P-OH bond [45]. After adsorption, the P-OH bond vibration is shifted to 776 cm1, indicating that the Be(II) in solution and phosphatebased surface active site on the GPHN might also form a surface complex through complexation. Figs. 7(c) and 8(d) illustrate the thermogravimetric (TG) analysis of GHPN before and after adsorption, respectively. Fig. 8(c) shows that the first loss peak of GHPN locates at around 100 ◦C, and the loss value is estimated to be 10.16 %, which is attributed to the free water adsorption in GHPN [46]. The second loss peak appears at 200–300 ◦C, and the amount of loss is obtained as 17.36 %, indicating the ammonia adsorption in GHPN [47]. The third endothermic peak at 310–450 ◦C represents the performance of GHPN dehydroxylation [48]. In Fig. 8(d), more endothermic peaks are observed in the TG curves of Used-GHPN than that of GHPN. The intensity of the second endothermic peak is enhanced, and the loss value is estimated to be 29 %, indicating that GHPN after adsorption contains more amino groups than GHPN. The endothermic peak at the temperature range of 600800 ◦C shows the conversion process of pyrophosphate ion P2O74[47]. During the reaction process, the valence state of the element will change, so the reaction of the absorption process can be deduced. Fig. 8 demonstrates the XPS analysis of GHPN before and after adsorption. The results reveal that GHPN after adsorption has more oxygen elements, indicating that more oxides are produced during the adsorption process, which is consistent to the SEM and EDS results. The beryllium element is detected on Used-GHPN, which implies that beryllium is adsorbed on GHPN. From the C1s of GHPN before and after adsorption, it can be seen that GHPN has more CC bonds, and the C-O-C and O-C=O bonds on GHPN after adsorption increase, indicating that C-O-C and O-C=O are more likely to bind to beryllium [36]. The N1s of GHPN before and after adsorption show that the C-N=O bond reduces and the strength of -NH3+enhances in the adsorption reaction, indicating that the N element changes from C-N=O to -NH3+[37]. The O1s of GHPN before and after adsorption show that the C–O bond is formed on GHPN after adsorption and the CO bond is strengthened, indicating that the oxygen element is accumulated on the surface of GHPN [38]. From P2p of GHPN before and after adsorption, it can be seen that the energy value of the P element has altered, indicating that new P-containing materials have been produced [39]. Combined with N1s of GHPN before and after adsorption, it can be inferred that beryllium ammonium phosphate and beryllium hydroxide might be formed [6]. By combined the results from SEM, EDS, TG, XRD and XPS analysis, it is inferred that Be(II) might bind with the phosphate group and amino group as well as hydroxyl group from GHPN. However, the bonding information for the interactions Be(II) with GHPN still remains unknown. Detailed microscopic analysis can further afford atomic-level insights on the metal-ligand bonding for the adsorption of Be(II). 3.7.Exploration of beryllium adsorption mechanisms Herein, quantum chemical calculations were used for preliminary exploration of the interactions of Be(II) with the phosphate group and amino group as well as hydroxyl group from GHPN. The structure of Be (II) cation in aqueous solution have been studied by experimental and theoretical methods. [Be(H2O)4]2+were considered as the most favorable species, the structure of which was optimized in the aqueous solution (Fig. 9(a)). The hydroxyl group and amino group from GHPN were separately optimized (Fig. 9(b) and (c)). Besides, the ethyl group was used to replace carbon chains in GHPN. The deprotonated phosphate group was also optimized and presented in Fig. 9(d). Fig. 10(a)–(e) are optimized structures of possible beryllium complexes. These combinations are therefore thermodynamically stable [19]. The above results indicate that the binding energies for the BeFig. 7.XPS analysis of GHPN and Used-GHPN. X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 13185110amino-phosphate complex (225.64 kcal/mol) and Be-hydroxylphosphate-complex complex (−216.40 kcal/mol) are lower than Beamino complex (121.13 kcal/mol), Be-amino-hydroxyl complex (−109.25 kcal/mol) and Be-hydroxyl complex (−96.61 kcal/mol), indicating higher stability of Be-amino-phosphate complex and Behydroxyl-phosphate-complex complex. The higher stability is due to the bonding affinity of Be(II) between the phosphate group, amino group and hydroxyl group of GHPN. Hence, it can be concluded that the introduction of phosphate and amino will not only increase the number of active sites on the surface of dextran but also help in the formation of stable beryllium complex. 4.Conclusion In summary, a novel eco-friendly dextran/phosphate/amino porous composite hydrogel (GHPN) was designed and synthesized for the first time to efficiently remove Be(II) from aqueous solutions. GHPN exhibited significant specific surface area and mesopores. Kinetic and thermodynamic investigations revealed that the kinetics followed the pseudo-second-order model and the Freundlich as isotherm model seemed more suitable for the adsorption of Be(II). Bedsides, the adsorption of Be(II) by GHPN was spontaneous and endothermic. The maximum Qe of GHPN for Be(II) could reach 43.75 mg/g at 35 ◦C and pH =6.5. GHPN would still be able to maintain constant adsorption and desorption rate (~80 %) after 5 cycles. Moreover, GHPN had robust adsorption selectivity for Be(II). According to the performed analyses, the introduction of PO43−and NH4+groups in GHPN led to the enhancement of the adsorption capacity of Be(II). Combined various characterizations and DFT calculations, it can be concluded that the phosphate group, hydroxyl group, and amino group own the affinity to form stable complexes with Be(II). Benefiting from the introduction of phosphate and amino, the adsorption effect of the hydrogel adsorbent on Be(II) can be greatly improved, and surface precipitation, complexation, and ligand exchange are the dominant adsorption mechanisms of Be(II) by GHPN. Therefore, the above studies suggest that GHPN is a potential candidate with high selectivity and Qe for Be(II) adsorption. CRediT authorship contribution statement Xu Zhao: Writing – review & editing, Writing – original draft, Data curation. Qingliang Wang: Funding acquisition. Yige Sun: Formal analysis. Haoshuai Li: Data curation. Zhiwu Lei: Data curation. Boyuan Zheng: Data curation. Hongyang Xia: Data curation. Yucheng Su: Resources. Kham Muhammad Yaruq Ali: Data curation. Hongqiang Wang: Data curation. Fang Hu: Writing – review & editing, Writing – original draft. 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. Data availability All data generated or analyzed during this study are included in this published article [and its Supplementary information files]. Acknowledgement We thank the following funding agencies for supporting this work: Research on characteristic properties of typical radioactive solid waste and radiation protection regulation technology and operation management mechanism (2019YFC1907701), as well as the National Natural Science Foundation of China (No. 52204363). The authors would like to express their gratitude to EditSprings (https://www.editsprings.cn) for Fig. 9.Optimized structures of Be(H2O)42+(a), Amino group of GHPN (b), Hydroxyl group of GHPN (c), Phosphate group (d) in the aqueous phase. Fig. 10.Optimized structures of possible beryllium complexes (Be-amino complex (a), Be-amino-hydroxyl complex (b), Be-hydroxyl complex (c), Be-aminophosphate complex (d), Be-hydroxyl-phosphate-complex (e)) in the aqueous phase. X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 13185111the expert linguistic services provided. Also, the authors appreciate the use of Gaussian 16 software package provided by High Performance Computing Center of Central South University and also thank Dr. Hong Liu for performing the DFT calculations and analyses. Besides, we gratefully acknowledge helpful comments and suggestions from anonymous reviews. Funding Ministry of Science and Technology of People’s Republic of China (2019YFC1907701). Appendix A.Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.ijbiomac.2024.131851. References [1]Nikolaus Nagel, Beryllium and copper-beryllium alloys, ChemBioEng Rev. 5 (1) (2018) 30–33, https://doi.org/10.1002/cben.201700016. [2]Shan Zhong, et al., Toxic metals and the risks of sludge from the treatment of wastewater from beryllium smelting, Chemosphere 326 (2023) 138439, https:// doi.org/10.1016/j.chemosphere.2023.138439. [3]Andreas Trabesinger, Peaceful berkelium, Nat. Chem. 9 (9) (2017) 924, https:// doi.org/10.1038/nchem.2845. [4]Evgenii Drobyshev, et al., New insight in beryllium toxicity excluding exposure to beryllium-containing dust: accumulation patterns, target organs, and elimination, Arch. Toxicol. 93 (2019) 859869, https://doi.org/10.1007/s00204-019-02432-7. [5]Mohsin Tanveer, Lei Wang, Potential targets to reduce beryllium toxicity in plants: a review, Plant Physiol. Biochem. 139 (2019) 691696, https://doi.org/10.1016/j. plaphy.2019.04.022. [6]Xu Zhao, et al., Preparation of porous calcium carbonate biochar and its beryllium adsorption performance, J. Environ. Chem. Eng. 11 (3) (2023) 110102, https://doi. org/10.1016/j.jece.2023.110102[7]X. Zhao, Y. Su, Z. Lei, H. Wang, E. Hu, F. Hu, X. Hao, Adsorptive removal of beryllium by Fe-modified activated carbon prepared from lotus leaf, Environ. Sci. Pollut. Res. 30 (7) (2023) 1834018353, https://doi.org/10.1007/s11356-022- 23415-9. [8]P. Gu, S. Liu, X. Cheng, S. Zhang, C. Wu, T. Wen, X. Wang, Recent strategies, progress, and prospects of two-dimensional metal carbides (MXenes) materials in wastewater purification: a review, Sci. Total Environ. 912 (2023) 169533, https:// doi.org/10.1016/j.scitotenv.2023.169533. [9]X. Liu, Y. Li, Z. Chen, H. Yang, S. Wang, Z. Tang, X. Wang, Recent progress of covalent organic frameworks membranes: design, ynthesis, and application in water treatment, Eco-Environ. Health 2 (2023) 117–130, https://doi.org/10.1016/ j.eehl.2023.07.001. [10]Y. Zhao, H. Gu, Y. Zhou, C. Wen, X. Liu, S. Wang, Z. Chen, H. Yang, X. Wang, COFbased membranes for liquid phase separation: preparation, mechanism and perspective, J. Environ. Sci. 141 (2024) 63–89, https://doi.org/10.1016/j. jes.2023.06.037. [11]Xiaoyu Jiang, et al., Efficient adsorption of uranium from aqueous solutions by microalgae based aerogel, Microporous Mesoporous Mater. 305 (2020) 110383, https://doi.org/10.1016/j.micromeso.2020.110383. [12]Qi Xin, et al., Enhanced performance in uranium extraction by the synergistic effect of functional groups on chitosan-based adsorbent, Carbohydr. Polym. 300 (2023) 120270, https://doi.org/10.1016/j.carbpol.2022.120270. [13]Zhengguo Wu, et al., Novel magnetic polysaccharide/graphene oxide@ Fe3O4 gel beads for adsorbing heavy metal ions, Carbohydr. Polym. 216 (2019) 119128, https://doi.org/10.1016/j.carbpol.2019.04.020. [14]Wenjie Liu, et al., Efficiency and mechanism of adsorption of low-concentration uranium from water by a new chitosan/aluminum sludge composite aerogel, Dalton Trans. 49 (10) (2020) 32093221, https://doi.org/10.1039/C9DT04670D. [15]X. Qi, L. Wu, T. Su, J. Zhang, W. Dong, Polysaccharide-based cationic hydrogels for dye adsorption, Colloids Surf. B: Biointerfaces 170 (2018) 364372, https://doi. org/10.1016/j.colsurfb.2018.06.036. [16]A. De Sousa, Indirect complexometric determination of beryllium, Talanta 22 (1011) (1975) 910911, https://doi.org/10.1016/0039-9140(75)80192-5[17]K.R. Krishnamoorthy, R.K. Iyer, Homogeneous precipitation of beryllium by means of trichloroacetic acid hydrolysis and determination as phosphate, Anal. Chim. Acta 47 (2) (1969) 333338, https://doi.org/10.1016/S0003-2670(01)95685-5. [18]Peiwen Wang, et al., Selective extraction of uranium from uraniumberyllium ore by acid leaching, J. Radioanal. Nucl. Chem. 322 (2019) 597604, https://doi.org/ 10.1007/s10967-019-06689-1. [19]J. Kohn, S. Spicher, M. Bursch, S. Grimme, Quickstart guide to model structures and interactions of artificial molecular muscleswith efficient computational methods, Chem. Commun. 58 (2021) 258261, https://doi.org/10.1039/ D1CC05759F. [20]P. Pracht, F. Bohle, S. Grimme, Automated exploration of the low-energy chemical space with fast quantum chemical methods, Phys. Chem. Chem. Phys. 22 (2020) 71697192, https://doi.org/10.1039/C9CP06869D. [21]T. Lu, F.W. Chen, Multiwfn: a multifunctional wavefunction analyzer, Comput. Chem. 33 (2012) 580592, https://doi.org/10.1002/jcc.22885[22]Fang Sun, et al., Biosorption behavior and mechanism of beryllium from aqueous solution by aerobic granule, Chem. Eng. J. 172 (23) (2011) 783791, https://doi. org/10.1016/j.cej.2011.06.062[23]Peizhen Zhang, et al., Characteristics, adsorption behaviors, Cu(II) adsorption mechanisms by cow manure biochar derived at various pyrolysis temperatures, Bioresour. Technol. 331 (2021) 125013, https://doi.org/10.1016/j. biortech.2021.125013. [24]Xin Zhang, et al., Enhanced adsorption of tetracycline by an iron and manganese oxides loaded biochar: kinetics, mechanism and column adsorption, Bioresour. Technol. 320 (2021) 124264, https://doi.org/10.1016/j.biortech.2020.124264. [25]Deepali Rahangdale, Anupama Kumar, Chitosan as a substrate for simultaneous surface imprinting of salicylic acid and cadmium, Carbohydr. Polym. 202 (2018) 334–344, https://doi.org/10.1016/j.carbpol.2018.08.129. [26]W. Liu, J. Zhang, C. Zhang, Y. Wang, Y. Li, Adsorptive removal of Cr (VI) by Femodified activated carbon prepared from Trapa natans husk, Chem. Eng. J. 162 (2) (2010) 677–684, https://doi.org/10.1016/j.cej.2010.06.020. [27]Xu Zhao, et al., Modification of activated carbon from agricultural waste lotus leaf and its adsorption mechanism of beryllium, Korean J. Chem. Eng. 40 (1) (2023) 255–266, https://doi.org/10.1007/s11814-022-1251-8. [28]Abd El-Magied, O. Mahmoud, et al., Biosorption of beryllium from aqueous solutions onto modified chitosan resin: equilibrium, kinetic and thermodynamic study, J. Dispers. Sci. Technol. 39 (11) (2018) 15971605, https://doi.org/ 10.1080/01932691.2018.1452757. [29]N.N. Basargin, O.V. Miroshnichenko, Beryllium (II) sorption from aqueous solutions by polystyrene-based chelating polymer sorbents, Russ. J. Inorg. Chem. 57 (5) (2012) 758762, https://doi.org/10.1134/S0036023612050026. [30]Osvaldo Pezoti, et al., NaOH-activated carbon of high surface area produced from guava seeds as a high-efficiency adsorbent for amoxicillin removal: kinetic, isotherm and thermodynamic studies, Chem. Eng. J. 288 (2016) 778788, https:// doi.org/10.1016/j.cej.2015.12.042. [31]Qi Xin, et al., Mechanism for the seleikctive adsorption of uranium from seawater using carboxymethyl-enhanced polysaccharide-based amidoxime adsorbent, Carbohydr. Polym. 324 (2024) 121576, https://doi.org/10.1016/j. carbpol.2023.121576[32]Shahla Mirzaeei, et al., Solubility of minoxidil in binary mixture of ethanol+water at various temperatures, Phys. Chem. Liq. 57 (6) (2019) 788799, https://doi.org/ 10.1080/00319104.2018.1528596[33]Zhijun Ren, et al., Study on adsorption of ammonia nitrogen by iron-loaded activated carbon from low temperature wastewater, Chemosphere 262 (2021) 127895, https://doi.org/10.1016/j.chemosphere.2020.127895. [34]Yu Liu, Ya-Juan Liu, Biosorption isotherms, kinetics and thermodynamics, Sep. Purif. Technol. 61 (3) (2008) 229242, https://doi.org/10.1016/j. seppur.2007.10.002. [35]N.N. Basargin, O.V. Miroshnichenko, Beryllium (II) sorption from aqueous solutions by polystyrene-based chelating polymer sorbents, Russ. J. Inorg. Chem. 57 (5) (2012) 758–762, https://doi.org/10.1134/S0036023612050026. [36]Zhiwei Huang, et al., Interaction mechanism of uranium (VI) with threedimensional graphene oxide-chitosan composite: insights from batch experiments, IR, XPS, and EXAFS spectroscopy, Chem. Eng. J. 328 (2017) 1066–1074, https:// doi.org/10.1016/j.cej.2017.07.067. [37]Yawen Cai, et al., Rational synthesis of novel phosphorylated chitosancarboxymethyl cellulose composite for highly effective decontamination of U (VI), ACS Sustain. Chem. Eng. 7 (5) (2019) 5393–5403, https://doi.org/10.1021/ acssuschemeng.8b06416. [38]Runze Zhou, et al., Selective adsorption of uranium from aqueous solution by mesoporous polymer-carbon composites containing phosphate groups, Environ. Nanotechnol. Monit. Manag. 15 (2021) 100441, https://doi.org/10.1016/j. enmm.2021.100441. [39]Zhibin Zhang, et al., Synthesis of ultralight phosphorylated carbon aerogel for efficient removal of U (VI): batch and fixed-bed column studies, Chem. Eng. J. 370 (2019) 13761387, https://doi.org/10.1016/j.cej.2019.04.012. [40]Enas A. Imam, et al., Synthesis of α-aminophosphonate functionalized chitosan sorbents: effect of methyl vs phenyl group on uranium sorption, Chem. Eng. J. 352 (2018) 10221034, https://doi.org/10.1016/j.cej.2018.06.003. [41]Kaipeng Wang, Qi Liu, Adsorption of phosphorylated chitosan on mineral surfaces, Colloids Surf. A Physicochem. Eng. Asp. 436 (2013) 656663, https://doi.org/ 10.1016/j.colsurfa.2013.07.030. [42]Kaipeng Wang, Qi Liu, Chemical structure analyses of phosphorylated chitosan, Carbohydr. Res. 386 (2014) 4856, https://doi.org/10.1016/j.carres.2013.12.021[43]Izabela Polowczyk, Anna Bastrzyk, Marta Fiedot, Protein-mediated precipitation of calcium carbonate, Materials 9 (11) (2016) 944, https://doi.org/10.3390/ ma9110944[44]A.H. Touny, M.M. Saleh, Fabrication of biphasic calcium phosphates nanowhiskers by reflux approach, Ceram. Int. 44 (14) (2018) 1654316547, https://doi.org/ 10.1016/j.ceramint.2018.06.075. [45]Alessandro C. Martins, et al., Removal of tetracycline by NaOH-activated carbon produced from macadamia nut shells: kinetic and equilibrium studies, Chem. Eng. J. 260 (2015) 291299, https://doi.org/10.1016/j.cej.2014.09.017. X. Zhao et al.

International Journal of Biological Macromolecules 269 (2024) 13185112[46]Nongnuch Laohavisuti, et al., Simple recycling of biowaste eggshells to various calcium phosphates for specific industries, Sci. Rep. 11 (1) (2021) 15143, https:// doi.org/10.1038/s41598-021-94643-1. [47]M. Jamil, et al., Characterization of β-tricalcium phosphate-clay mineral composite obtained by sintering powder of apatitic calcium phosphate and montmorillonite, Surf. Interfaces 17 (2019) 100380, https://doi.org/10.1016/j.surfin.2019.100380[48]Dariusz Sternik, et al., A thermal, sorptive and spectral study of HDTMA-bentonite loaded with uranyl phosphate, J. Therm. Anal. Calorim. 129 (2017) 12771289, https://doi.org/10.1007/s10973-017-6384-3. X. Zhao et al.