Structural morphology and surface analysis of nano-cellulose beads
Figure 1 shows the surface morphology of nano-cellulose hydrogel beads with different carboxyl group content, Fig. 1 (a)-(d) are the SEM images at 20000 magnification of nano-cellulose with different carboxyl content, nanocellulose beads present a smooth and flat surface and a porous fiber network structure. The adsorption capacity of the adsorbent is related to the pore size, it can be seen that the cellulose nanofibers in the nanocellulose beads form a stable three-dimensional network porous structure from Fig. 1 (e) and Fig. 1 (f), and the pore size of the nanocellulose beads was is in the mesoporous range (several nanometers to several tens of nanometers). Relevant studies have shown that dye molecules hardly gather on the larger pore size of the adsorbent, and are mainly uniformly adsorbed on the small pores and mesoporous structure of the adsorbent (Qiao et al. 2021). This shows that the mesoporous structure of nanocellulose is beneficial to its application in the adsorption of small molecules (Gaduan et al. 2021).
Figure 2 (a) is the nitrogen isotherm desorption curve of different nanocellulose beads. Under the same relative pressure, the adsorption capacity of TOCNB1, TOCNB2, TOCNB3, and TOCNB4 showed an increasing trend. The specific surface areas of TOCNB1, TOCNB2, TOCNB3, and TOCNB4 obtained by using the BET calculation model are 173 m2/g, 246 m2/g, 299 m2/g, and 367 m2/g, respectively, which shows that the increase in carboxylate content is beneficial to obtaining beads with a nanocellulose network structure with a larger specific surface area. Figure 2 (b) shows the pore size distribution curve of different nanocellulose beads. BET results showed that the pore diameters of nanocellulose beads were all concentrated between 5 and 15 nm, which belonged to the mesoporous structure, and the pore size of nanocellulose beads is relatively uniform. BET test shows that the specific surface area of nanocellulose beads was related to the content of carboxyl groups in nanocellulose. The increase in the content of carboxylate on the nanocellulose will promote the formation of a porous network structure with more pores and a larger specific surface area (Mishnaevsky et al. 2019). It indicates that, as a functional material with a controllable specific surface area, nanocellulose beads have much application potential.
Chemical structure analysis of nano-cellulose beads
The FTIR spectra of TOCNB1, TOCNB2, TOCNB3 and TOCNB4 were shown in Fig. 3 (a). The peak at 3400 cm− 1 was ascribed to the -OH stretching vibration in cellulose macromolecule, the peak at 2904 cm− 1 and 1020 cm− 1 was ascribed to the C-H and C-O in cellulose macromolecule, respectively. The absorption peaks above were shown in all samples. The peak at 1614 cm− 1 and 1424 cm− 1 were represented for the asymmetric stretching and symmetrical stretching of -COO- in nanocellulose, respectively. The -COO- asymmetric stretching vibration peak of TOCN1, TOCN2, TOCN3 and TOCN4 at 1614 cm− 1 showed a gradual increase, it indicates that the carboxyl group content increases. With the increasing of carboxyl content, the absorption band of TOCN at 3400 cm− 1 shifts to a high wavenumber, which indicated that the hydrogen bonds between nanocellulose molecules were weakened with the increase of carboxyl group content. The mutual repulsion between the carboxylate ions will break the hydrogen bonds in the nanocellulose molecules, making the stretching vibration frequency of the hydroxyl groups on the nanocellulose molecules higher (Xue et al. 2017).
The surface charge of the adsorbent is an important indicator of nanocellulose as a cationic adsorbent, the Zeta potential of TOCNB1, TOCNB2, TOCNB3 and TOCNB4 were shown in Fig. 3 (b), the potential values are − 27 mV, -46 mV, -57 mV, and − 63 mV, respectively. This shows that as the content of carboxyl groups increases, the negative surface charge of nanocellulose increases. This result not only confirmed the successful preparation of TOCN with different carboxyl group content, but also confirmed that nanocellulose that more carboxyl groups is expected to exhibit excellent adsorption performance in the application of high-efficiency cationic adsorbents.
Adsorption performance analysis of nano-cellulose beads
To investigate the adsorption capacity of negatively charged nanocellulose beads for cationic dyes, methylene blue (MB) adsorption analysis was performed and the results were shown in Fig. 4. The initial concentration of MB is 5 mmol L− 1, as shown in Fig. 4 (a), different nano-cellulose beads showed excellent adsorption capacity for cationic dye MB, which were 506 mg g− 1, 543 mg g− 1, 583 mg g− 1, 639 mg g− 1 respectively. With the increase of carboxyl group content, the adsorption capacity of nano-cellulose beads to MB was also enhanced. The concentration of dye was also one of the main factors affecting adsorption. Investigating the adsorbent removal rate of dyes at different concentrations was helpful to understand the water purification capacity of the adsorbent and its application fields. MB solutions of different concentrations (0.01 ~ 5 mmol L− 1) were prepared to explore the dye removal rate of nanocellulose beads. As shown in Fig. 4 (b), at the low concentration (0.01 ~ 0.5 mmol L− 1) TOCNB owned an excellent removal effect on dyes, and the removal rate of nanocellulose beads with different carboxyl group content had no obvious difference, all of which were above 90%. With the increasing of dye concentration, the removal rate of dye increases with the increase of TOCNB carboxyl content. Among them, TOCNB4 still maintained a removal rate of 80% at concentration of 5 mmol L− 1. The high adsorption efficiency of TOCNB4 was still maintained at high concentrations, mainly due to the electrostatic interaction between carboxyl groups and dye molecules, and the aggregation of dyes at high concentrations (Qiao et al. 2021).
Because TOCNB was based on a cationic adsorbent under the action of static electricity, and the action of static electricity was significantly affected by the pH change of the solution. As shown in Fig. 4 (c), With the increase of the pH from 3 to 9 in the MB solution, the adsorption capacity of TOCNB on MB increased significantly. Among them, TOCNB1 increased from 237 mg g− 1 to 538 mg g− 1, TOCNB2 increased from 298 mg g− 1 to 614 mg g− 1, TOCNB3 increased from 360 mg g− 1 to 633 mg g− 1, and TOCNB4 increased from 395 mg g− 1 increased to 682 mg g− 1. In an acidic solution, the hydrogen ions in the solution and the dye molecules competed with each other, and the adsorption capacity of -COO−will deteriorate due to protonation. The increased alkalinity in the solution gradually weakened the adsorption competition from ions, which was conducive to enhancing the interaction between carboxylate ions and dye molecules, so that the adsorption capacity of TOCNB was significantly enhanced. This indicates that alkalinity can enhance the adsorption performance of cationic adsorbent (Luo et al. 2019). Figure 4 (d) shows that with the increase of concentration of NaCl, the adsorption capacity of nanocellulose beads to dyes decreases. Relevant research has showed when there is electrostatic interaction between the dye and the adsorbent, the adsorption capacity of the dye will decrease with the increase of the electrolyte concentration (Hong et al. 2018).
In the adsorption kinetics, the adsorption rate was an important indicator of the adsorption performance. As shown in, the adsorption capacity of nanocellulose beads rose rapidly in short time, especially in the first 20 minutes. This rapid adsorption effect was attributed to the electrostatic attraction between the negatively charged TOCNB and the positively charged MB dye. Subsequently, TOCNB reached adsorption equilibrium in 40–50 minutes. It can be obviously observed in the Fig. 5 (a) that with the content of carboxyl groups increased, the adsorption rate of TOCNB increased significantly. It was worth noting that TOCNB4 owned abundant binding sites due to more carboxyl group content and larger specific surface area, so it took more time to reach adsorption equilibrium. It indicated that the carboxyl content and specific surface area of TOCNB were positively correlated with the dye adsorption capacity. This figure shows all TOCNB achieve adsorption equilibrium in 40–50 min. In order to further explore the adsorption kinetics of TOCNB with fast adsorption efficiency, pseudo-first-order kinetic model, pseudo-second-order kinetic model and intra-particle diffusion model were used to fit the adsorption data of TOCNB with time. The pseudo-first-order kinetic model was ln(qe-qt) = lnqe-k1t, the pseudo-second-order kinetic model was t/qt = 1/k2q2e + t/qe, and the intra-particle diffusion model was qt = kit0.5 +C. According to the corresponding fitted curve, the slope, intercept and corresponding dynamic parameter values were shown in Table 1. The curves of pseudo-first-order dynamic model were shown in Fig. 5 (b) and (c). The correlation coefficient (R2) of TOCNB1, TOCN2, TOCN3 and TOCN4 were 0.85114,0.71711,0.87944 and 0.78287, respectively. The R2 of curves that pseudo-second-order dynamic model of TOCNB1, TOCN2, TOCN3 and TOCN4 were 0.99965,0.99962,0.99967,0.99958. This showed that the pseudo-second-order kinetic model can better describe the adsorption mechanism of TOCNB. Saturated adsorption capacity (qmax) of TOCNB1, TOCN2, TOCN3 and TOCN4 calculated by pseudo-second-order kinetic model was closer to the experimental value. The pseudo-second-order kinetic model was based on the chemisorption hypothesis. Therefore, the main driving force for the adsorption of dyes by nanocellulose beads in the process of this experiment should be chemical interaction. The adsorption rate is limited by the availability of carboxyl groups. which were attributed to the high adsorption rate constant of TOCNB. Figure 5 (d) simulated the intra-particle diffusion model of TOCNB, the correlation coefficients of TOCNB1, TOCN2, TOCN3 and TOCN4 were 0.57389, 0.54391, 0.56586 and 0.58258, respectively. This indicated that particle diffusion within the particles was not the main mechanism of the TOCNB adsorption process. Increasing the diffusion time of dyes was not the only way to improve adsorption capacity. The adsorption process was controlled by both physical and chemical interactions. The adsorption process may also was affected by other influences, such as electromagnetic, ion exchange, hydrogen bonding, etc., which also affected the adsorption efficiency (Xu, Ouyang and Yang 2021).
Table 1
Adsorption kinetic parameters of TOCNB adsorbent.
| | Pseudo-first-order ln(qe-qt) = lnqe-k1t | Pseudo-second-order t/qt = 1/k2qe2 + t/qe | Intraparticle diffusion qt = k3t0.5 +C |
| qe-exp (mg g-1) | qe-cal (mg g-1) k1 (min-1) | qe-cal (mg g-1) k2 (min-1) | k1 (min-1) C |
TOCNB1 | 505.76 | 121.62 0.02177 | 518.13 0.00039 | 568.61 294.32 |
TOCNB2 | 542.79 | 92.01 0.02049 | 552.49 0.00044 | 607.14 333.63 |
TOCNB3 | 582.75 | 142.94 0.02949 | 595.24 0.00038 | 653.87 354.79 |
TOCNB4 | 638.89 | 131.49 0.02177 | 653.59 0.00033 | 717.54 378.91 |
Adsorption isotherms of nanocellulose beads
The adsorption process between various adsorbents and small dye molecules can well described by adsorption isotherm. Therefore, the adsorption isotherm model of nano-cellulose adsorbent was fitted. As shown in Fig. 6 (a), the nano cellulose bead adsorbent increased with the increase of adsorption equilibrium concentration, which was positively correlated with the initial concentration of dye. The increased of dye concentration increased the driving force of dye mass transfer from water to solid nano-cellulose adsorbent, so as to provide more binding sites for MB and TOCNB to contact each other. Three isothermal models were selected as follows, Freundlich isothermal model: lnqe = lnKF + bFlnCe, Langmuir isothermal model: Ce/qe = Ce/qm + 1/qmKL, Temkin isothermal model: qe = BlnKT + BlnCe, to explore the adsorption theory of TOCNB. The fitting results were shown in Fig. 6 and Table 2. Langmuir isothermal model had the highest correlation coefficients: 0.99785, 0.99894, 0.99752 and 0.9984, which shown that TOCNB was more in line with this model. Langmuir isothermal model was based on the assumption that adsorption occurs at a specific homogeneous phase point and all homogeneous phase points are equal. Therefore, adsorption of MB based on TOCNB occurs at a specific homogeneous phase point. The adsorption process of TOCNB to MB is monolayer and uniform (Hokkanen et al. 2013).
Table 2
Adsorption isotherm parameters for the adsorption of TOCNB adsorbent.
| | Freundlich lnqe = lnKF + bFlnCe | Langmuir Ce/qe = Ce/qm + 1/qmKL | Temkin qe = BlnKT + BlnCe |
| qe-exp (mg g-1) | kF (mg g-1) bF | qm (mg g-1) kL (L mg-1) | kT (L mg-1) B (KJ− 2mol− 2) |
TOCNB1 | 505.76 | 11.547 0.65662 | 613.49 117.76 | 0.3078 89.52052 |
TOCNB2 | 542.79 | 13.768 0.63942 | 671.14 123.27 | 0.3496 92.78642 |
TOCNB3 | 582.75 | 15.052 0.64402 | 746.27 123.60 | 0.3526 101.02872 |
TOCNB4 | 638.89 | 16.688 0.64514 | 925.93 148.96 | 0.3670 109.65589 |
Effect of temperature on adsorption properties of nano-cellulose beads
Generally, the temperature has a certain effect on the adsorption performance. The thermodynamic behavior of the adsorbent was studied by using the change of the adsorption capacity of the adsorbent at different temperatures. Thermodynamic parameters included free energy change, enthalpy change and entropy change. These parameters were used to evaluate the adsorption behavior of TOCNB. As shown in Fig. 7 (a), the adsorption capacity of TOCNB for MB increased with the increase of temperature. The van Hoff equation: ln(qe/Ce) = ΔS/R-ΔH/RT, qe (mg g− 1) was the equilibrium adsorption capacity of TOCNB, Ce (mg L− 1) was the equilibrium concentration of dyes, T (K) was the specified temperature, and R (8.314 J mol− 1k− 1) was the ideal gas constant. The curve obtained by fitting was shown in Fig. 7 (b), The correlation coefficients of TOCNB1, TOCNB2, TOCNB3, and TOCNB4 were 0.95095, 0.9999, 0.99335, and 0.99932, respectively. Free energy change, enthalpy change and entropy change were calculated. As shown in Table 3, the adsorption data of TOCNB in temperature was well fitted by the van Hoof equation, the enthalpy change and entropy change of TOCNB were both positive, and the free energy change was negative, indicated that the adsorption process of TOCNB to MB was endothermic, also spontaneous. The adsorption process shown that the adsorption capacity increased with the increase of temperature, and higher temperature was beneficial to the progress of adsorption. This was consistent with our experimental results.
Table 3
Thermodynamic parameters for the adsorption of TOCNB adsorbent.
| ΔH (kJ mol-1) | ΔS (J mol-1 k-1) | ΔG (kJ mol-1) 298K, 308K, 318K |
TOCNB1 | 5.53 | 17.24 | -5.13, -5.30, -5.48 |
TOCNB2 | 5.50 | 18.93 | -5.64, -5.82, -6.01 |
TOCNB3 | 4.75 | 18.40 | -5.48, -5.66, -5.85 |
TOCNB4 | 6.65 | 28.04 | -8.35, -8.63, -8.91 |
Analysis of Desorption Performance of Nanocellulose Beads
In the practical application of adsorbents, desorption capacity and regeneration capacity were the key factors to evaluate the performance of adsorbents, these reflected the ability of the adsorbent to be reused. Different solutions such as sodium hydroxide solution (1 mol L− 1), deionized water, ethanol and hydrochloric acid solution (1 mol L− 1) were used as eluents to study the desorption performance of TOCNB adsorbent. The desorption efficiency obtained by desorbing TOCNB adsorbed MB was shown in Fig. 8 (a). The desorption efficiencies after desorption with HCl and ethanol were as high as 93% and 94%, with deionized water and sodium hydroxide solution were 45% and 34%, which meant that TOCNB can be desorbed by HCl or ethanol solution to obtain a cationic adsorbent that can be regenerated for multiple times. HCl was used as eluent to wash TOCNB4, the removal rate of 1 mmol L− 1 MB solution for five cycles was shown in Fig. 8 (b). Obviously, TOCNB4 had a high removal rate of 83% for MB after five cycles of TOCNB4. This verified that TOCNB was a high-efficiency adsorbent with good regenerable cycle performance and reusability.