Various TA amounts of TA@HAS and TA@HAS/PVA solutions were prepared. HAS, PVA, and TA were crosslinked using glutaraldehyde with no pre-treatment owing to the high reactivity of the hydroxyl groups of HAS, PVA, and the phenolic hydroxyl groups of TA. The unreacted TA was quantified by UV–vis spectroscopy26. The relationship between the TA dosage and crosslinking ratio of TA in TA@HAS and TA@HAS/PVA is plotted in Fig. 2a. The results confirmed the successful crosslinking of TA with HAS and PVA and revealed that the crosslinking ratio of TA in TA@HAS and TA@HAS/PVA decreased with increasing TA dosage. This implied that the expected adsorption capacity of the composite adsorbent increased with increasing TA dosage, whereas the TA utilization rate decreased. A TA dosage of 2.5 g was selected as the optimal amount for preparing TA@HAS/PVA fibers for use in the subsequent adsorption experiments.
The FT-IR spectra of HAS/PVA fibers, TA, and TA@HAS/PVA fibers were presented in Fig. 2b. Compared to the HAS/PVA fibers, the TA@HAS/PVA fibers displayed a characteristic peak indicating hydroxyl substitution of the benzene ring at 765 cm− 1. The characteristic TA peak at 870 cm− 1, which shifted to 860 cm− 1 in the TA@HAS/PVA fibers, was assigned to the out-of-plane bending vibration of aromatic C–H moieties. The TA peaks at 1720 and 1630 cm− 1, which shifted to 1730 and 1644 cm− 1 in the TA@HAS/PVA fibers, were assigned to aromatic ring skeleton vibrations. The peaks at 1000–1200 cm− 1 observed for the TA@HAS/PVA fibers were ascribed to the C–O bonds of the benzene rings and ether bridges, and the broadening of this band was related to the crosslinking of TA with HAS and PVA. Therefore, the FT-IR spectra demonstrated the successful crosslinking of TA in the TA@HAS/PVA fibers.
The surface and cross-sectional morphologies of the HAS/PVA and TA@HAS/PVA fibers were examined by SEM, as shown in Fig. 3. Porous structures with irregular shapes were observed in the cross-section of the HAS/PVA fibers (Fig. 3a and 3c). The diameter of the porous core was approximately 1–5 µm. After crosslinking with TA to produce the TA@HAS/PVA fibers, the cross-sectional structure appeared to become more compact with smaller pores (< 2 µm) (Fig. 3b and 3d). Several previous studies found that the 3D porous structures may be related to the structure of crosslinked HAS33,38. It was because the retrogradation of SS immersed in absolute ethanol endowed the materials with complex 3D porous structures. Furthermore, the observed porous structure was expected to be favorable for adsorption.
The mechanical properties of the HAS/PVA and TA@HAS/PVA fibers were measured by tensile testing in the wet state to mimic the behavior of the fibers under actual adsorption conditions in the aqueous solution. The results were presented in Fig. 2c. The tensile strength and tensile strain of the HAS/PVA fibers were 9.83 MPa and 160.10%, respectively. For the crosslinked TA@HAS/PVA fibers, the tensile strength increased to 11.31 MPa, while the tensile strain decreased to 159.56%. These results demonstrated that the TA@HAS/PVA fibers retained excellent mechanical properties after crosslinking. The results indicated that the fibers should maintain their structure under actual adsorption conditions and encouraged us to further explore their adsorption performance.
Adsorbents often suffer from biological fouling under aqueous solution conditions, which can severely affect their adsorption capacity. We thus evaluated the anti-biofouling activity of the TA@HAS/PVA fibers. Samples of 0.01 g TA@HAS/PVA fibers were incubated in 10 mL lysogeny broth containing 0.1 mL microorganisms (Chlorella vulgaris, Nannochloropsisoculata) and cultured at 37°C for 7 days. A control group was established using the same amount of liquid medium and microorganisms without TA@HAS/PVA fibers as used in the experimental group. Photographs showing the growth of the microorganisms were presented in Fig. 4a. The samples were subjected to cell counting to determine the number of microorganisms and evaluated the anti-biofouling activity of the TA@HAS/PVA fibers, as shown in Fig. 4b. The results revealed that the number of microorganisms in the experimental group was greatly reduced compared to the control group, indicating that the TA@HAS/PVA fibers exerted a strong inhibitory effect on all two microorganisms owing to the excellent antimicrobial activity of TA. These observations indicated that the TA@HAS/PVA fibers retained the antimicrobial activity of TA even after crosslinking and provided further supporting evidence of the presence of TA in the composite fibers.
The solution pH is an important factor influencing U(VI) adsorption because it affects the surface charge and metal-binding sites of the adsorbent in addition to the degree of ionization of the adsorbent. The influence of the initial solution pH on the U(VI) adsorption by the TA@HAS/PVA fibers was thus examined in the range of 2.0–9.0. As shown in Fig. 5a, the U(VI)adsorption was strongly affected by the solution pH, which was ascribed to the role of the pH in determining the state of the uranium ions and the surface charge of the adsorbent. The equilibrium adsorption capacity of the TA@HAS/PVA fibers rapidly increased as the pH was increased from 2 to 6 and thereafter promptly decreased. The maximum adsorption capacity was found to be 86 mg g− 1, which occurred at pH 6.
To evaluate the adsorption rate, the influence of the contact time on the adsorption capacity of the TA@HAS/PVA fibers was studied in the range of 10–2000 min for an initial U(VI) concentration of 200 mg L− 1. As plotted in Fig. 5b, adsorption occurred rapidly during the initial 500 min and gradually slowed down until equilibrium, which was reached after approximately 700 min of shaking. To further investigate the adsorption process, we fitted the experimental data using the pseudo-first-order and pseudo-second-order models described by Eqs. (4) and (5), respectively:
$$\begin{array}{c}\text{ln}\left({q}_{\text{e}}-{q}_{t}\right)=ln{q}_{\text{e}}-{k}_{\text{I}}t\end{array}$$
4
$$\begin{array}{c}\frac{t}{{q}_{t}}=\frac{1}{{k}_{\text{I}\text{I}}{q}_{\text{e}}^{2}}+\frac{t}{{q}_{\text{e}}}\end{array}$$
5
where qe (mg g− 1) and qt (mg g− 1) denote the U(VI) adsorption capacities at equilibrium and time t(min), respectively, and kI (min− 1) and kII (g mg− 1 min− 1) are the pseudo-first-order and pseudo-second-order adsorption rate constants, respectively.
The fitted parameters and correlation coefficients (R2) obtained for the two adsorption models were listed in Table S1. Based on the R2 values, the pseudo-second-order equation (R2 = 0.9299) provided a better fit for the measured adsorption data than the pseudo-first-order equation (R2 = 0.8243). Moreover, the qe value calculated from the pseudo-second-order fitting was close to the experimental qe values, confirming the validity of this model in relation to the U(VI) adsorption data. Over the studied range of contact time, the adsorption process could be divided into three stages (as shown in Fig. 5b). In the initial stage with the greatest slope, the high adsorption rate was attributable to the diffusion of U(VI) to the surface through the aqueous solution. In the second stage with a gentle slope, the slow adsorption rate corresponded to internal diffusion or pore diffusion in the TA@HAS/PVA fibers. In the third stage, the system had reached equilibrium.
The U(VI) adsorption performance of the TA@HAS/PVA fibers was further evaluated by determining the adsorption isotherms over an initial concentration range of 20–2000 mg L− 1 at pH 6. The variation of the equilibrium adsorption capacity with respect to the equilibrium concentration is plotted in Fig. 5c. The adsorption capacity increased with increasing initial U(VI) concentration. The adsorption process on the surface of an adsorbent is typically described using the Langmuir and Freundlich adsorption isotherm models, as described by Eqs. (6) and (7), respectively:
$$\begin{array}{c}\frac{{C}_{\text{e}}}{{q}_{\text{e}}}=\frac{1}{{K}_{\text{L}}{q}_{\text{m}}}+\frac{{C}_{\text{e}}}{{q}_{\text{m}}}\end{array}$$
6
$$\begin{array}{c}{q}_{\text{e}}={K}_{\text{F}}{C}_{\text{e}}^{\frac{1}{n}}\end{array}$$
7
where qe (mg g− 1) and qm (mg g− 1) are the equilibrium and maximum adsorption capacity of U(VI), respectively, Ce (mg L− 1) is the U(VI) concentration at equilibrium, KL (L mg− 1) is a constant related to the adsorption, KF (mg g− 1) is associated with the adsorption capacity, and 1/n describes the surface heterogeneity of the adsorbent (with values ranging between 0 and 1).
The fitted curves are shown in Fig. 5c and the fitted parameters and correlation coefficients (R2) are listed in Table S2. Based on the R2 values, the Langmuir model (R2 = 0.9492) described the adsorption process better than the Freundlich model (R2 = 0.8133), indicating that the adsorption of U(VI) onto the TA@HAS/PVA fibers occurred through a monolayer adsorption process.
The reusability of the TA@HAS/PVA fibers was assessed following desorption and separation. After each adsorption experiment at a U(VI) concentration of 200 mg L− 1, the fibers were thoroughly washed with 0.1 M HNO3 as a desorption reagent, followed by soaking and washing in pure water to remove the desorbed U(VI) and HNO3, immersion in anhydrous ethanol to remove residual water, and drying at 60°C. The regenerated TA@HAS/PVA fibers were then used in the next adsorption step for a total of ten cycles. The regeneration rate E (%) was calculated after each cycle, and the results were presented in Fig. 6a. The fibers retained > 90% of their initial adsorption capacity over three cycles and > 80% over ten cycles. The good mechanical properties of the TA@HAS/PVA fibers were advantageous for repeated use. These results demonstrated the feasibility of recycling and reusing the TA@HAS/PVA fibers to realize a highly efficient and low-cost method for U(VI) adsorption from aqueous solution. Moreover, as shown in Fig. 6b, the fibers underwent a noticeable color change from pale yellow to dark brown during adsorption, which was a strong indication for the process of adsorption. The major role in the process of U(VI) adsorption was shown in Fig. 6c. The hydroxyl groups on TA has the ability to capture U(VI) in solution23.
The TA@HAS/PVA fibers were analyzed by FT-IR spectroscopy before and after U(VI) adsorption and the resulting spectra were presented in Fig. 2b. Upon adsorption, the peaks at 1730 and 1644 cm− 1 corresponding to the aromatic ring skeleton vibrations shifted to 1630 and 1586 cm− 1, the aryl ether peaks at 1384 and 1025 cm− 1 shifted to 1350 and 1090 cm− 1, and the peak at 765 cm− 1 corresponding to a benzene ring vibration shifted to 778 cm− 1. These results confirmed the adsorption of U(VI) by the TA@HAS/PVA fibers. The chemical composition and adsorption mechanism of the U(VI)-loaded TA@HAS/PVA fibers were further examined by XPS analysis. As shown in Fig. 7a, compared to the TA@HAS/PVA fibers, after U(VI)adsorption, two distinct U 4f peaks appeared in the core-level spectra with binding energies corresponding to U 4f5/2 (392.91 eV) and U 4f7/2 (382.08 eV), confirming the presence of uranium in the U-TA@HAS/PVA fibers. To further examine the role of the phenolic hydroxyl groups in the U(VI) adsorption process, the fine spectra of the U 4f, C 1s, and O 1s core energy levels were recorded, as presented in Fig. 7b, 7c, and 7d, respectively. The U 4f5/2 binding energies of the U-TA@HAS/PVA fibers were observed at 393.07 and 391.08 eV, while the U 4f7/2 binding energies occurred at 382.26 and 380.26 eV, indicating the adsorption of U(VI) on the TA@HAS/PVA fibers. The C 1s spectrum of the TA@HAS/PVA fibers contained peaks at 284.81, 286.42, and 288.66 and 290.48eV, corresponding to C–OH, C–O–C, and C(O)O, respectively, while the O 1s spectrum displayed peaks at 531.78 and 534.50 eV, corresponding to C(O*)O and C–O, respectively. Upon U(VI) adsorption, the C–OH and C–O–C peaks in the C 1s spectrum shifted to 284.80 and 286.30 eV, while the C(O)O peak shifted to 288.04 eV. Furthermore, the C–O and C = O peaks in the O 1s spectrum shifted to 532.58 eV and an additional peak corresponding to U = O was observed at 530.88 eV. These results indicated that the phenolic hydroxyl groups were involved in chelation during the U(VI) adsorption process.
The influence of other species on U(VI) adsorption was examined for five potentially interfering metal ions, namely, Al3+, Ca2+, K+, Mg2+, and Na+. A solution containing each of these metal ions as their nitrates and U(VI) was prepared in which the concentration of each metal ion was 100 mg L− 1. The TA@HAS/PVA fibers were added to this solution followed by stirring at 30°C for 10 h. The fibers were then removed using tweezers, and the residual amounts of the metal ions in solution were measured by ICP-OES. As shown in Fig. 8, the TA@HAS/PVA fibers displayed the ability to adsorb not only U(VI) but also Al3+ and Na+. In contrast, the equilibrium adsorption capacities of the fibers regarding Ca2+, Mg2+, and K+ were low. Thus, the TA@HAS/PVA fibers exhibited high U(VI) adsorption capacity (16.51 mg g− 1) even in the presence of coexisting metal ions.