XPS analyses were used to investigate the surface chemical compositions of DAWP, DAWP-PEI and DAWP-PEI-β-CD. As showed in Fig. 2, the C1s spectrum of DAWP could be quantitatively resolved into three bonds at 284.81, 286.62 and 288.12 eV corresponded to carbon-carbon bond (C-C), carbon-oxygen bond (C-OH) and carbon and oxygen double bond (C = O), respectively. Interestingly, once the PEI molecules were immobilized on DAWP via Schiff base reaction, two new peaks appeared at 287.15 eV (C = N) and 285.74 eV (C-N), proving that a mass of amino groups existed on the surface of DAWP-PEI (Liu et al., 2020). However, as the β-CD molecules were further introduced into DAWP-PEI through cross-linking reaction, the relative content of C-N on DAWP-PEI (14.20%) was lower than that of DAWP-PEI-β-CD (25.58%), and other characteristic peaks were still presented in DAWP-PEI-β-CD. These observations conformed that PEI and β-CD molecules had been successfully immobilized on DAWP via a facile two-step method.
FT-IR measurements were mostly used in order to confirm the formation of functional groups. Thus, different functional groups were detected in the IR spectra of β-CD, DAWP, DAWP-PEI and DAWP-PEI-β-CD. In Fig. 2d, the typical peaks of β-CD at 3352 cm− 1 and 1208 cm− 1 were generated by the -OH and -CH2 bending vibrations, whereas another band at 1059 cm− 1 was produced by inverse-symmetric glycosidic (Wang et al., 2014). The bands centered at 1688 cm− 1, 1427 cm− 1 and 1109 cm− 1 ascribed to -C = O, -CH2 and C-O-C stretching vibrations of DAWP, respectively (Liu et al., 2019). Notably, once the PEI molecules were introduced into oxidized waste paper, two regular peaks appeared at 1649 cm− 1 and 1422 cm− 1 corresponded to -C = N and -NH2, respectively (Al-Harahsheha et al., 2020). Interestingly, after the cross-linking reaction, the characteristic peaks of β-CD and PEI were also appeared at the IR spectra of DAWP-PEI-β-CD, such as primary amine (1455 cm− 1) and the b(1–4) skeleton vibrations (1033 cm− 1). These analysis results further proved that the β-CD and PEI were successfully introduced into the oxidized waste paper.
The chemical structure of all adsorbents was further assessed by solid state NMR measurements. In Fig. 3, one can see that the broad signals from 102 ppm to 59 ppm were attributed to the carbon atoms in β-CD, and the signals at δ = 63, 72, 84 and 102 ppm corresponding to C-d, C-a, C-e and C-b, respectively (Shen et al., 2015). Compared with β-CD, the characteristic peak of carbon atoms at δ = 72 ppm (C-2) and 65 ppm (C-4) were almost overlapped or disappeared in the NMR spectrum of DAWP, which may be related to the signal amplification of C-5 and C-3 after oxidation reaction. However, when DAWP was reacted with PEI, there were several new characteristic peaks appeared at NMR spectra of DAWP-PEI. For instance, the carbon and nitrogen double bond (C = N) as well as carbon and nitrogen single bond (C-N) were observed at δ = 143 ppm and 30 ppm, respectively. Interestingly, most of DAWP and PEI’s carbon atom signals were also detected on the NMR spectra of DAWP-PEI-β-CD after cross-linking reaction, and a different signal appeared at δ 169 ppm may ascribed to C = O. These results were in accordance with FT-IR and XPS results, which further manifested that the successful cross-linking between DAWP and β-CD/PEI.
Influences of pH values
As is well-known, solution pH not only can affect the relative distribution of pollutants species but also change the surface charge of adsorbents. Thus, the adsorption abilities of DAWP-PEI-β-CD towards Eu(III) and Au(III) were investigated in this work. As depicted in Fig. 3a, adsorption amounts of Eu(III) on DAWP-PEI-β-CD were up to 282 mg/g rapidly from 2.0 to 5.0, subsequently maintained the high level within pH the range of 6.0–9.0, which implied that solution pH played a vital role in elimination of contaminants. This phenomenon could be explained by the relative distribution of Eu(III) species and the surface charge of DAWP-PEI-β-CD. It was reported that Eu(III) ions were mainly presented as Eu3+ species at pH < 5.0 in water solution, and then Eu(OH)2+ and Eu2(OH)24+ species began to increase with pH further increasing (Li et al., 2020). Meanwhile, zeta potential analyzer determined the surface charge density of DAWP-PEI-β-CD was 3.29 (Fig. S1). Thus, according to Coulomb's law, the positive charge of Eu(III) species were more easily adsorbed on the negatively charged DAWP-PEI-β-CD through electrostatic attraction. However, the adsorption behavior of Au(III) was significantly different from that of Eu(III) at various pH values. In Fig. 3b, one can see that adsorbed amounts of Au(III) on DAWP-PEI-β-CD increased sharply as pH values were raised from 1.0 to 2.0, and then sharply decreased over the pH range 2.0–6.0, and achieving its best performance at pH 2.0. Such a pH-dependent adsorption can be explained from the point of view of the following aspects. At lower pH values, the Au(III) ions were mainly existed as AuCl4−, which were favorable to capture on the positively-charged DAWP-PEI-β-CD via electrostatic interactions. Notably, the chloride ion occupied a part of the active sites during adsorption process at pH < 2.0, which led to a lower adsorption amounts (Wang et al., 2015). However, under high pH conditions, the negative charge increased on the surface of DAWP-PEI-β-CD and the electrostatic repulsion would occur between the bio-adsorbents and adsorbates at high pH values.
Adsorption isotherms and thermodynamic studies
The adsorption isotherms are an important factor in understanding the adsorption efficiency and studying the adsorption mechanism. Figure 5 displays the relationship between the residual Eu(III) and Au(III) concentrations and the equilibrium adsorption capacities on DAWP-PEI-β-CD at room temperature. Evidently, along with the initial concentration increasing, the adsorption amounts of Eu(III) and Au(III) on DAWP-PEI-β-CD rapidly increased at the beginning and then tended to adsorption equilibrium. Interestingly, the equilibrium adsorption capacities of Au(III) were both higher than that of Eu(III) on DAWP-PEI and DAWP-PEI-β-CD, which was evident that the greater affinity to Au(III) than Eu(III) in the adsorption process. To determine the saturated capture amounts and sorption type, the Langmuir and Freundlich equations were applied to fit the experimental data, and the detailed fitting results were provided in Fig. 5 and Table S1. As expected, the results of calculation and simulation showed that the adsorption behavior of Eu(III) and Au(III) were fitted well by the Langmuir model, which indicated that uptake of the two targeted contaminants onto DAWP-PEI-β-CD surfaces was localized in a monolayer (Yang et al., 2019). Besides, the adsorption thermodynamics of Eu(III) and Au(III) on DAWP-PEI-β-CD were also measured using the Gibbs equations. As listed in Table S2, the spontaneous and endothermic adsorption processes were revealed by the negative ΔG° and positive ΔHo values, whereas the negative ΔSo reflected a decrease in the randomness during the adsorption process. Notably, the maximum uptake amounts of DAWP-PEI-β-CD for single Eu(III) and Au(III) were 241.3 and 424.2 mg mg/g, which were far higher than that of the reported available materials (Table S3). Such a high adsorption amount could be ascribed to the special chemical structure and the large cavities, which were advantageous for accommodating more Eu(III) and Au(III).
Adsorption kinetics not only focuses on the relationship between adsorption capacity and adsorption time, but also evaluate the rate of the interaction process. Figure 6 illustrates the time-dependent experiments of Eu(III) and Au(III) adsorbed onto DAWP-PEI-β-CD. As excepted, the concentration of targeted contaminants rapidly declined during the initial phase, and then tend to a constant value with the extension of incubation time. The rapid adsorption in the beginning may be due to the greater concentration gradient and more available sites for adsorption. Subsequently, the active sites were gradually depleted, resulting in the final adsorption equilibrium. Note that the Au(III) exhibited faster kinetics compared with Eu(III) at optimal water environmental condition, which suggested that Eu(III) ions took a relatively longer time to diffuse into the interlayer of DAWP-PEI-β-CD. To figure out the adsorption rate and rate determining step, different kinetic models were applied to simulate the kinetic data, and the corresponding fitting curves and parameters were displayed in Fig. 6 and Table S4. According to the fitted results, on can see that the PFO equation either overestimated or underestimated the kinetics behavior of Eu(III) and Au(III) on DAWP-PEI-β-CD, while the PSO model could well simulate the kinetic process from the beginning to the end, and the calculated limits of quantification were close to the experimental limits of quantification, which implied that uptake of the two targeted pollutants mainly depended on chemical adsorption (Cheng et al., 2018, Wang et al., 2019). To further identify the main rate-controlling step, the kinetic data were further analyzed by the Intraparticle diffusion equation, and the plots of qt vs. t1/2 were displayed in Fig. 6c, d. The simulation results exhibited that the lines do not pass through the origin and tend to be multilinear, which manifested that the rate controlling steps did not only include intraparticle diffusion, and other reactions according to the definition of the equation (Qiu et al., 2020, Hu et al., 2021).
Considering the practical application of the targeted pollutants removal from industrial wastewater, the effect of competing ions of adsorbents were also important. Thus, the selective binding ability of DAWP-PEI-β-CD towards Eu(III) and Au(III) was investigated in a mixed system, in which the concentration of metal ions were all controlled at 50 mg/L. In Fig. S2, one can see that the adsorption capacity of Au(III) on DAWP-PEI-β-CD was calculated as 209 mg/g in single system, and that was reduced by 20.1% in mixed system. This phenomenon occurred due to most of co-existing metal ions existed as their cationic or neutral forms, while Au(III) ions were mainly existed as AuCl4− in acidic solution. Thus, the negatively charged Au(III) species could occupy more adsorption sites on the surface of DAWP-PEI-β-CD, which led to a high Au(III) adsorption amounts. Besides, it should be pointed out that the molecular structure of Eu(III) were similar to that of Au(III) to some extent, but the removal amounts of Eu(III) was still lower than that of Au(III) at this adsorption conditions. This may be due to the electrostatic repulsion occurred between Eu(III) ions during adsorption process. This result manifested that DAWP-PEI-β-CD exhibited high selectivity for Au(III) against other co-existing pollutants in the multi-solute system.
To find out the detailed elimination mechanism, the as-prepared bio-adsorbent’s surface properties and specific elemental constitutes were explored by SEM, XPS and XRD after binding with Eu(III) and Au(III). In Fig. 7, one can see that the main components of DAWP-PEI-β-CD were C, H, N and O, whereas the element Eu and Au were spread over the whole surface of DAWP-PEI-β-CD-Eu and DAWP-PEI-β-CD-Au, suggesting that DAWP-PEI-β-CD could efficient enrichment of Eu(III) and Au(III) from water environment. Interestingly, in Fig. 8a, in addition to C1s and O1s, the spent DAWP-PEI-β-CD’s surface also appeared characteristic peaks of Eu3d and Au4f, which was consistent with the results analyzed in EDS, further proved that the successful immobilization of the two targeted pollutants onto DAWP-PEI-β-CD.
To resolve the chemical properties and states of Eu and Au atoms on the DAWP-PEI-β-CD, the high-resolution spectra of Eu3d and Au4f were further resolved by deconvolution method. In Fig. 8b, one could see that two main peaks located at 1134.49 eV (Eu3d3/2) and 1164.16 eV (Eu3d5/2) in the high-resolution Eu3d spectrum corresponded to the Eu(III) signals, which implied that only Eu(III) ions existed on the surface of DAWP-PEI-β-CD, and the surface complexes reaction dominated the processes of Eu(III) uptake on DAWP-PEI-β-CD (Huang et al., 2018). However, In Fig. 8c, the Au4f peak was mainly consisted of Au4f7/2 and Au4f5/2, each of which could be further divided into Au(III) and Au(0). Specifically, the peaks at 83.36 eV (Au4f7/2) and 87.08 eV (Au4f5/2) corresponded to Au(III), whereas the signatures at 84.18 eV (Au4f7/2) and 89.05 eV (Au4f5/2) were ascribed to Au(0), which demonstrated that a large amount of Au(0) existed onto the surface of spent bio-adsorbents (Pestov et al., 2015). This result was further proved by the XRD analysis. In Fig. 8e, one can see that the powder XRD patterns of DAWP-PEI-β-CD exhibited two major peaks at 2θ = 12.3o and 18.1o, which may be ascribed to the characteristic peak of β-CD. However, after adsorption of Au(III), there were four the high intensity peaks appeared at the XRD patterns of DAWP-PEI-β-CD-Au, and the reflections of elemental gold were observed at 2θ = 38.19o, 44.38o, 64.71o and 77.52o with the corresponding planes of (111), (200), (220) and (311). Noticeably, the specific area percentage of C = O enhanced from 12.09 % to 15.78%, and the content of Au(0) reached up to 54.65% after adsorption of Au(III) (see Table S5). These phenomena illustrated that part of surface adsorbed Au(III) has been reduced to Au(0) during the Au(III) capture processes. Besides, in Fig. 8f, the N1s peaks shifted from 398.86 eV to 399.09 eV after Au(III) adsorption, while it was shifted from 398.86 eV to 399.52 eV after Eu(III) adsorption, which illuminated that the nitrogen containing functional groups played roles not only as a cross-linker but also as adsorption sites for Eu(III) and Au(III). On the basis of characterization analysis as well as adsorption experiments, the mechanism of removing the targeted contaminant by DAWP-PEI-β-CD was schematically illustrated in Fig. 9. The mechanism of Eu(III) removal by DAWP-PEI-β-CD was mainly through the surface coordination and host-guest inclusion interaction. Specifically, on the one hand, the ample amino groups of DAWP-PEI-β-CD that could form complex compounds with Eu(III). On the other hand, the cavities of β-CD could make contributions to the host-guest inclusion complexes with Eu(III). However, the adsorption mechanism of Au(III) on DAWP-PEI-β-CD was mainly attributed to the electrostatic attraction, chelation, host-guest inclusion, and redox reaction. Once the fresh DAWP-PEI-β-CD was exposed to Au(III) solution, the targeted contaminants were rapidly enriched onto DAWP-PEI-β-CD surface via electrostatic interaction and followed the reduction of a part of Au(III) to Au(0) with the help of reductive functional groups. Finally, most of Au(III) and Au(0) were immobilized on the surface of DAWP-PEI-β-CD by the host-guest inclusion interaction.