To fabricate the cellulose-based three-dimensional network, CMC and TA were first dissolved in water and mixed slowly, and CPAM was then lowly dropped into the resulting solution at a rate of 5 mL/min under a stirring rate of 500 r/min. The formed precipitate was collected using a filter and vacuum freeze-dried to obtain the CMC–CPAM–TA adsorbent.
FTIR was performed to investigate the structural changes of CMC, CPAM, and TA before and after the self-assembly process. As shown in Fig. 2a, the –NH2/–NH–, and –OH groups of CMC–CPAM–TA appeared at 3200–3600 cm− 1.(Gupta M et al., 2018) The spectra of TA and CMC–CPAM–TA exhibited prominent absorption peaks at 1607 and 1533 cm− 1, which are the vibrational absorption peaks of the benzene ring skeleton. Compared with TA, a new absorption peak at 1660 cm− 1 appeared in the spectrum of CMC–CPAM–TA, possibly owing to strong chemical interactions between the chemical groups connected to the benzene ring in TA and other chemical groups. In addition, the stretching vibration peak of C = O in TA was observed at 1721 cm− 1 in the spectra of TA and CMC–CPAM–TA. Although CMC and CPAM contain C = O groups, no absorption peak attributable to this functional group was observed. Instead, an absorption peak appeared at 1663 cm− 1, which may be due to the presence of different chemical groups around the C = O groups in CMC, CPAM, and CPAM. Moreover, the corresponding absorption peak of CMC–CPAM–TA may overlap with the absorption peak of the benzene ring skeleton. These results indicate the occurrence of the self-assembly process of CMC and CPAM with the formation of strong chemical interactions, which are most likely hydrogen bonding and electrostatic interactions according to the structural analysis of CMC, CPAM, and TA.
A TGA was performed to investigate the thermal stability of the materials. Figure 2b shows that the thermal degradation of CMC, CPAM, and TA occurred at 295°C and 206°C, 400°C, and 280°C, and 310°C, respectively. When the temperature reached 800°C, the weight of CMC, CPAM, and TA decreased to 40.34%, 15.80%, and 27.73%, respectively. Meanwhile, the thermal degradation temperature of CMC–CPAM–TA was 250°C and its weight decreased to 38.80%, which is between that of CMC, CPAM, and TA and close to that of CMC. This indicates the presence of chemical bonds between CMC, CPAM, and TA.
To study the mechanism of the self-assembly into CMC–CPAM–TA, the zeta potential of the materials was analyzed before and after the self-assembly. As shown in Fig. 2c, the zeta potential of CMC was − 18 mV, indicating its negatively charged characteristic. In contrast, CPAM and TA were positively charged, with zeta potentials of 16.9 and 11.6 mV, respectively. This potential difference would provide a strong driving force for the self-assembly. The zeta potential of CMC–CPAM–TA was − 10.2 mV, which differs considerably from that of CMC, CPAM, and TA. This indicates the formation of electrostatic interactions between CMC, CPAM, and TA during the self-assembly process. Moreover, the negative feature of CMC–CPAM–TA implies that the negative carboxyl and phenoxy groups originating from CMC and TA tend to locate at the surface of CMC–CPAM–TA, which is beneficial for the removal of positive Cu(II) ions and RhB via coordination or electrostatic effects.
In addition, the UV–vis spectra shown in Fig. 2d illustrate that CMC and CPAM exhibited no obvious absorption peaks in the range of 200–800 nm. Meanwhile, TA showed an absorption peak at 276 nm corresponding to the catechol group.(Lee S J et al., 2023) After the self-assembly, this characteristic absorption peak in CMC–CPAM–TA was blue-shifted compared with that in TA, also demonstrating the self-assembly between CMC, CPAM, and TA. Taken together, the FTIR, zeta potential, and UV–vis results confirmed the successful self-assembly of CMC, CPAM, and TA via hydrogen bonds or electrostatic interactions among the –OH, –COOH, and –NH/–NH2 groups.
To further explore the micromorphology of CMC–CPAM–TA, a SEM analysis was conducted. Fig. S1 shows that CMC, CPAM, and TA exhibited bulk morphologies with relatively smooth or block surfaces. However, obvious interconnected nanoparticles can be observed in CMC–CPAM–TA (Fig. 3a). The average size of the nanoparticles was determined to be ca. 44 nm using Gaussian function fitting (Fig. 3b). Moreover, the interconnection between nanoparticles leads to a mesoporous/macroporous structure, which is conducive to the mass transfer and diffusion during the adsorption of Cu(II) ions and RhB. These observations indicate that the hydrogen bonds and electrostatic interactions between CMC, CPAM, and TA are the driving force for the supramolecular self-assembly.
The transmission electron microscopy (TEM) image presented in Fig. 3c shows the interconnected nanoparticles and apparent mesopores/macropores in CMC–CPAM–TA. The corresponding EDS mapping demonstrated that C, N and O were uniformly distributed in CMC–CPAM–TA (Fig. 3d). To determine the surface area and porous structure of CMC–CPAM–TA, which can be correlated to the number of adsorption sites available for the removal of Cu(II) ions and RhB, nitrogen adsorption–desorption measurements were performed. According to Fig. 3e, the surface area of CMC–CPAM–TA was 37.02 m2/g. A clear hysteresis loop in the relative pressure (p/p0) range of 0.82–1.0 was detected in the nitrogen adsorption–desorption isotherm of CMC–CPAM–TA, indicating the presence of mesopores. Furthermore, when the p/p0 value approached 1.0, the nitrogen adsorption capacity drastically increased, indicating the presence of macropores in CMC–CPAM–TA. The mesoporous/macroporous structure of CMC–CPAM–TA was confirmed by the pore size distribution, which was in the range of 10–110 nm (Fig. 3f). This result is in agreement with the SEM and TEM observation (Fig. 3a and 3c). The large surface area and abundant mesoporous/macroporous supramolecular network structure of CMC–CPAM–TA can be expected to improve the adsorption of Cu(II) ions and RhB by exposing more adsorption sites (i.e., the catechol of TA and carboxyl groups of CMC) and enhancing the fast mass transport of Cu(II) ions and RhB from the solution to the adsorption sites.
Adsorption experiments using CMC–CPAM–TA
The pH of the solution affects the surface charge of CMC–CPAM–TA, which influences the adsorption capacity of CMC–CPAM–TA for Cu(II) ions and RhB. To evaluate the optimum pH for the removal of Cu(II) ions and RhB at 25℃, the pH of the initial adsorption solution was controlled at 4–9. As shown in Fig. 4a, the adsorption capacity of CMC–CPAM–TA for Cu(II) ions increases when the pH rises from 3 to 6, which can be attributed to the occurrence of protonation of the active carboxyl, phenoxy and amino groups at low pH values. The presence of high concentration of H+ at low pH values would cause a competition between the H+ ions and Cu(II) for the adsorption sites of CMC–CPAM–TA.(Teow Y H et al., 2018; Zhang P et al., 2021) When the pH increases, the surface sites of CMC–CPAM–TA are deprotonated, which would promote the coordination and electrostatic interaction between Cu(II) ions and the adsorption sites,(Gupta M et al., 2018) enhancing the adsorption capacity. At pH values of > 6, a precipitate originating from the reaction of Cu(II) and OH− is formed. For RhB, the adsorption capacity of CMC–CPAM–TA gradually increases with increasing pH, and the growth rate tends to stabilize at pH values of > 5. This is mainly due to the increase in the potential difference between CMC–CPAM–TA and RhB with increasing pH for pH < 5 (Fig. 4b), which increases the adsorption capacity for RhB by improving the electrostatic interaction between CMC–CPAM–TA and RhB.(Liu K et al., 2015) At a pH of > 5, the adsorption capacity changes only slightly because the potential difference between CMC and RhB tends to stabilize (Fig. 4b). Thus, the pH values of 6 and 5 were selected as the optimal pH for the adsorption of Cu(II) ions and RhB, respectively.
To explore the equilibration time for the maximum adsorption capacity and the adsorption kinetics of CMC–CPAM–TA, the adsorption capacity for Cu(II) ions and RhB was evaluated as a function of time. As depicted in Fig. 4c and 4d, the adsorption capacity increased gradually with time. When the adsorption time of CMC–CPAM–TA for Cu(II) ions reached 60 min, the growth rate of the adsorption capacity decreased and reached equilibrium at 180 min. In the case of RhB, the turning point of the growth rate of the adsorption capacity appeared at 90 min and the equilibrium was reached at 240 min. This result was not only due to the presence of abundant adsorption sites but also to the macropores/mesopores and large surface area of CMC–CPAM–TA, which facilitate the exposure of the adsorption sites and a rapid mass transfer of Cu(II) ions and RhB from the solution to the adsorption sites within 60 and 90 min, respectively. The adsorption capacities of CMC–CPAM–TA for Cu(II) ions and RhB reached equilibrium at 180 and 240 min, respectively. Furthermore, the adsorption kinetics of CMC–CPAM–TA were evaluated using the pseudo-first-order and pseudo-second-order rate models. Figure 4c and 4d and Table S1 show that the pseudo-second-order rate model described better the adsorption of Cu(II) ions and RhB over CMC–CPAM–TA compared with the pseudo-first-order rate model due to its higher correlation coefficient (R2), which suggests that the removal of Cu(II) ions and RhB over CMC–CPAM–TA probably occurred via a chemical adsorption process.
Next, the effect of the initial concentrations of Cu(II) ions and RhB on the adsorption capacity was investigated. The corresponding adsorption isotherms are shown in Fig. 4e and 4f. As illustrated in Fig. 4f, the adsorption capacity of CMC–CPAM–TA increased with increasing initial concentrations of Cu(II) ions and RhB. In particular, when the RhB concentration exceeded 500 mg/L, the adsorption capacity of CMC–CPAM–TA reached stabilization, implying that the number of available adsorption sites in CMC–CPAM–TA became the limiting factor for the adsorption capacity. To further determine the maximum adsorption capacities of CMC–CPAM–TA for Cu(II) ions and RhB, the Langmuir and Freundlich adsorption isotherm models were used. The fitting parameters are displayed in Table S2, and the fitting curves are shown in Fig. 4e and 4f. The Langmuir model for Cu(II) and RhB gave R2 values of 0.9996 and 0.9808, respectively, which were higher than those obtained using the Freundlich model, demonstrating that the former model described the adsorption behavior of CMC–CPAM–TA for Cu(II) ions and RhB better than the latter. These results indicate that CMC–CPAM–TA adsorbs Cu(II) ions and RhB according to a monolayer adsorption mode. Moreover, the n values were greater than 1, indicating a high adsorption intensity and illustrating a favorable adsorption of Cu(II) ions and RhB over CMC–CPAM–TA.(Chen Q et al., 2019) According to the Langmuir model, the maximum adsorption capacity (qm) values of CMC–CPAM–TA for Cu(II) ions and RhB were calculated to be 669.8 and 202.1 mg/g, respectively, which are 1.3–23.6 and 1.1–72.2 times larger than those of most reported cellulose-based adsorbents (Fig. 4g and 4h and Tables S3 and S4).
According to the SEM images, the interconnected nanoparticle morphology of CMC–CPAM–TA was preserved after adsorption of Cu(II) ions and RhB (Fig. 5a and 5b). The EDS mapping obtained after adsorption of Cu(II) ions revealed the presence of Cu(II) ions and a decrease in the C and O content in CMC–CPAM–TA (Fig. 5c and 5d and Table S5), which indicates that a large amount of Cu(II) ions were adsorbed on CMC–CPAM–TA. After RhB adsorption, the EDS mapping showed an increase in the C content and a decrease in the O content in CMC–CPAM–TA, which is consistent with the mass ratio of C and O in RhB (Table S5), indicating that a large amount of RhB was adsorbed on CMC–CPAM–TA.
Next, an XPS analysis was performed to investigate the interaction between the adsorption sites and Cu(II) ions and RhB. According to Fig. 6a, compared with pristine CMC–CPAM–TA, an obvious Cu2p peak appeared in the XPS Cu2p spectrum after adsorption of Cu(II) ions, suggesting the efficient adsorption capability of CMC–CPAM–TA for Cu(II) ions. As RhB contains no other elements except C, O, and N, no new peak appeared after adsorption of RhB.
Before adsorption, the O1s XPS spectrum of CMC–CPAM–TA showed two peaks at 532.5 and 531.0 eV, which can be assigned to C = O and C–O bonds,(Nan Nan Xia Z H H, Jian Qiao Su, Fangong Kong, 2023) respectively (Fig. 6b). However, upon adsorption of Cu(II) ions, the C–O peak shifted from 531.0 to 531.25 eV (Fig. 6d), which indicates that the C–O bond of catechol in TA coordinates with Cu(II). Accordingly, after adsorption, no remarkable change in the peak at 399.25 eV due to the C–N group in CPAM was observed in the N1s XPS spectrum (Fig. 6c and 6e). These findings indicate that the adsorption over CMC–CPAM–TA proceeds via the formation of Cu–O coordination interactions, supporting the role of the catechol groups in the removal of Cu(II) ions. Meanwhile, the N1s XPS spectrum of RhB shows two peaks at 401.25 and 399.35 eV, among which the former, which corresponds to protonated N, shifted to 402.35 eV after adsorption (Fig. 6g and 6h). Similarly, the O1s XPS spectrum of RhB shows a peak at 531.75 eV and another at 533.55 eV, which is due to –C = O and shifted to 532.35 eV after adsorption (Fig. 6f and 6i). These results suggest that the adsorption of RhB on CMC–CPAM–TA occurs via electrostatic interactions between N+ and –C = O in RhB and –C = O and N+ in CMC–CPAM–TA.
To evaluate the regeneration, stability, and reusability of CMC–CPAM–TA, desorption experiments were conducted using distilled water as the desorption solvent. After the adsorption process, the Cu-loaded and RhB-loaded CMC–CPAM–TA materials were respectively immersed in distilled water for 30 min and the pH of the solution was adjusted to 3. Next, the spent CMC–CPAM–TA was freeze-dried and reused for the next adsorption cycle. The adsorption performance of CMC–CPAM–TA for Cu(II) ions and RhB was evaluated in 200 mL of a Cu(II) solution and a RhB solution (200 mg/L) for 60 min at pH values of 6 and 5, respectively, at 25°C. As depicted in Fig. 7a and 7d, CMC–CPAM–TA still preserved 79% and 63% of the initial adsorption efficiency for Cu(II) ions and RhB, respectively, after five cycles. The SEM image recorded after regeneration showed that CMC–CPAM–TA still presented the interconnected nanoparticle morphology of the pristine sample (Fig. 7b and 7e). Furthermore, the corresponding EDS mapping showed that the signal of the Cu element decreased considerably when the CMC–CPAM–TA adsorbent was regenerated (Fig. 7c and 7f). These results demonstrate the good regeneration performance of CMC–CPAM–TA, which endows it with huge potential for large-scale wastewater treatment.