The efficient enrichment of U(VI) by graphene oxide-supported chitosan

Abstract

Graphene oxide-supported chitosan (GO-Ch) composites were synthesized using a covalent method for U(VI) adsorption and were characterized by scanning electron microscopy (SEM), Fourier transform infrared (FT-IR) spectroscopy, thermogravimetric analysis (TGA), differential thermal analysis (DTA) and extended X-ray absorption fine structure (EXAFS). The characteristic results indicated that Ch was successfully grafted onto GO. The adsorption of U(VI) on GO-Ch was investigated under different environmental conditions. The adsorption kinetics showed that the adsorption of U(VI) on GO-Ch followed the pseudo-second-order equation. The maximum adsorption capacity of U(VI) on GO-Ch at pH 4.0 and T = 303 K calculated from the Langmuir model was 225.78 mg g⁻¹. Thermodynamic parameters calculated from temperature-dependent adsorption isotherms suggested that U(VI) adsorption on GO-Ch was an endothermic and spontaneous process. The batch desorption indicated U(VI) cannot be completely desorbed from GO-Ch without intervention, suggesting the irreversible adsorption of U(VI) on GO-Ch. The analysis of FT-IR spectra suggested that the interaction mechanism of U(VI) on GO-Ch was mainly chemical adsorption by –NH₂ and –COOH groups. According to EXAFS analysis, the peaks at ∼2.9 Å can be satisfactorily fitted by the U–C/N shell, revealing the formation of inner-sphere surface complexes. It is demonstrated that the GO-Ch nanocomposite can be a promising material for the preconcentration and solidification of U(VI) from large volumes of aqueous solution.

---

1. Introduction

The release of uranium from mine tailings and the nuclear industry has been of concern due to its radioactivity and biological toxicity.^1–3 For the sake of public health and ecosystem stability, it is necessary to eliminate U(VI) from contaminated wastewater prior to releasing it into the natural environment. The adsorption method is considered to be a simple and economical method for the removal of U(VI) from wastewater.⁴ A great number of researchers have investigated the adsorption of U(VI) on minerals, organic macromolecules and organisms.^4–6 In these studies, the impact of environmental factors (e.g., pH, organic substances, ionic strength, and temperature) on the adsorption capacity of adsorbents have been extensively investigated. However, the relatively low adsorption efficacy limited their application to large volumes of aqueous solution. Thus, the development of new artificial adsorbents with high adsorption capacity is of particular importance for the safe treatment and the disposal of U(VI) from aqueous solution.

Graphene oxide (GO) has been evidenced to exhibit great removal of heavy metals,^7–9 radionuclides^6,10–12 and dye.^13–15 Ding et al.¹⁶ found the maximum adsorption of U(VI) on GO was 208.33 mg g⁻¹ at pH 3.0 and T = 303 K. Sun et al.¹² also demonstrated GO presented high adsorption capacity for Eu(III) (175.44 mg Eu(III) per gram of GO at T = 298 K). However, the irreversible aggregation of GO or the polydisperse in its thickness, lateral size, and shape can hinder the adsorption performance and reduce their adsorption capacity.¹⁷ Therefore, the modification of GO by introducing various functional groups has been reported to enhance its dispersity and efficacy.^12,18 The functionalization of GO with nanoparticles, organic compounds, biomaterials, and polymers was reviewed by Chen et al.¹⁹ As a natural product, chitosan (Ch) is identified to have strong affinity for heavy metals (i.e., Cu(II), Cr(VI), Cd(II), Ag(I)) and organic matter^20–23 due to large numbers of amino (–NH₂) and hydroxyl (–OH) groups. However, the poor mechanical property and the dissolution of Ch in strong acids greatly affect its experimental study and commercial application.²⁴ It is demonstrated that the modification of Ch by biomaterials not only prevents the dissolution in strong acids, but also increases the porosity of the polymers.²⁵ To the authors' knowledge, few studies on the adsorption of U(VI) on GO-supported chitosan (GO-Ch) composites are reported, and the interaction mechanism between U(VI) and GO-Ch is not well defined.

The objectives of the current study are (1) to synthesize GO- Ch composites and to characterize the microscopic and macroscopic surface properties of GO-Ch composites by using scanning electron microscopy (SEM), thermal gravity analysis and differential thermal analysis (TGA/DTA); (2) to investigate the adsorption of U(VI) on GO-Ch under various environmental conditions (i.e., contact time, pH, ionic strength, initial U(VI) concentration and temperature) by batch techniques; and (3) to discuss the interaction mechanism between U(VI) and GO-Ch composites by using Fourier transformed infrared spectroscopy (FT-IR) and extended X-ray absorption fine structure (EXAFS) spectroscopy.

2. Experimental section

2.1. Materials

All chemicals used in the experiments were purchased in analytical purity and were used without any purification. All reagents were prepared with Milli-Q water. Ch (with a deacetylation degree of 90%) was purchased from Sinopharm Chemical Reagent Co., Ltd, Beijing. The glutaraldehyde (GLA 50 wt% in water, Sigma-Aldrich) was used as the cross-linking agent.

2.2. Preparation of GO-Ch composites

GO was synthesized by the modified Hummers method.²⁶ Briefly, graphite powder (2.0 g) and sodium nitrate (1.5 g) were added into concentrated sulfuric acid (0 °C) under stirring conditions. Then, potassium permanganate (9.0 g) was slowly added to the suspension under ice bath conditions. The mixtures were reacted at room temperature for 5 days under vigorous stirring conditions. After that, 230 mL of the diluted H₂SO₄ (5 wt%) was injected when the aforementioned suspension was heated to 98 °C. The residual MnO₄⁻ was removed by adding 100 mL of hydrogen peroxide (30 wt%) at 60 °C. The GO nanosheets were obtained by centrifuging it at 18 000 rpm for 60 min after ultrasonic treatment several times.

The GO-Ch composites were synthesized by chemical method. Typically, 2.0 g of Ch power was dissolved in a 100 mL of 4 wt% acetic acid solution under ultrasonic treatment (100 w) for 30 min. Follow that, 15 mL GLA and 0.5 g GO was added into the aforementioned suspension under vigorous stirring conditions at room temperature for 90 min. Then the mixture was adjusted to pH 9–10 and was stirred at 80 °C for 1 h. The precipitate was washed with ethanol and followed distilled water until the pH neutral.

2.3. Characterization of GO-Ch composites

The morphology of GO-Ch composites was conducted by SEM (JSM-6320F FE-SEM electron microscope), TGA/DTA (TGA Q5000 V3.15 Build 263TA Instruments) with heating rate of 10 K min⁻¹ and a flow rate of 75 mL min⁻¹. The potentiometric acid–base titrations were conducted under argon using a DL50 Automatic Titrator (Mettler Toledo) in 0.01 M NaClO₄ as a background electrolyte. FT-IR (Perkin-Elmer Spectrum 100, USA) in the range of 4000–400 cm⁻¹ was performed using KBr pellets. Uranium L_III-edge EXAFS of samples was conducted at the Shanghai Synchrotron Radiation Facility (SSRF, Shanghai, China). All of wet samples were measured in the fluorescence mode using Ge detector (Canberra Ultra-LE Ge) due to the low uranium concentration. EXAFS data fitting was performed using Athena and Artemis interfaces to IFFEFIT 7.0 software. The EXAFS threshold energies (E₀) were defined as 17174 eV for the U samples, indicating that uranium in these systems was U(VI) species. The incident X-ray intensity I⁰ is measured with an Ar-filled ionization chamber at ambient pressure. 25 scans were recorded for each individual sample. The Autobk algorithm was applied for background removal and the frequency cutoff parameter R_bkg was set to 1.20. In all cases, the models are fitted to the k³-weighted raw EXAFS spectra.

2.4. Adsorption and desorption Experiments

Batch adsorption experiments were conducted in the polycarbonate tubes with 0.75 g L⁻¹ GO-Ch and 1.7 × 10⁻⁴ mol L⁻¹ U(VI) in the presence of 0.01 mol L⁻¹ NaClO₄. The bulk suspensions of GO-Ch and NaClO₄ were pre-equilibrated for 24 h, and then U(VI) solution was spiked into the bulk suspension. The breakthrough experiment was conducted according to the method of previous paper.²⁷ Briefly, 0.5 g dried GO-Ch was packed into the glass column with an inner-radius of 0.4 cm, then the column was injected with constant water. Before U(VI) injection, the column was percolated with a 0.01 mol L⁻¹ NaClO₄ solution (pH 4.0) at a flow rate of 0.25 mL min⁻¹. The adsorbent volume has an average of 0.87 mL (V₀). The solution containing 3 × 10⁻³ mol L⁻¹ U(VI) (C₀) was injected into the GO-Ch column. The pH of suspension was adjusted from pH 2.5 to 10.0 by the drop wise addition of the negligible volume of 0.1 mol L⁻¹ HClO₄ or NaOH solutions. The isothermal adsorptions of U(VI) on GO and GO-Ch were investigated at pH 4.0 with the initial concentration in the range of 10–60 mg L⁻¹. The method of synthetic wastewater used in this study is similar to that of Yan et al.²⁸ The suspensions were shaken at T = 303 K for 24 h, which was demonstrated to be adequate for the suspension to obtain equilibrium by the preliminary experiments. The solid phase was then separated from the solution phase by the centrifugation it at 18 000 rpm for 30 min. The adsorption percentage of U(VI) (adsorption (%)) and adsorption capacity (Q_s, mg g⁻¹) can be expressed as eqn (1) and (2), respectively:

adsorption (%) = (C₀ − C_eq)/C₀ × 100% (1)

Q_s = V × (C₀ − C_eq)/m (2)

where C₀ (mg L⁻¹) and C_eq (mg L⁻¹) are initial concentration and concentration after adsorption, respectively. m (g) and V (mL) are the mass of adsorbents and the volume of the suspension, respectively.

The desorption experiments were conducted by lowering U(VI) concentration. After adsorption equilibrium, half volume of the supernatant was displaced by the equivalent volume of U(VI)-free background electrolyte solution (0.01 M NaClO₄) with the same pH values. The amount of desorbed U(VI) was measured in terms of the residual concentrations of U(VI) on GO-Ch after desorption. The concentration of U(VI) was measured by a kinetic phosphorescence analyzer (KPA-11, Richland, USA). All experimental data were the average of triplicate determinations and the relative errors were within ±5%.

3. Results and discussion

3.1. Characterization of GO-Ch

As shown by SEM image in Fig. 1A, the nanosheet-like structure of GO with large thickness, smooth surface, and wrinkled edge is observed. The GO-Ch (Fig. 1B) presents the aggregation and much rougher surface, revealing that Ch has been assembled on the surface of the GO layers with a high density. Additionally, the formation of some holes is also observed on the GO-Ch surface. The TGA curves of GO, Ch and GO-Ch are shown in Fig. 1C, in the first decomposition phase (0–100 °C), approximately 7%, 10% and 15% of weight loss are registered for GO, Ch and GO-Ch, respectively, which are assigned to the evaporation of adsorbed water. In the 150–300 °C region, the Ch and GO-Ch lose about 31% of their weight, whereas the GO loses about 50% of its weight. The weight loss of GO can be due to the decomposition of massive oxygenated functional groups. The GO-Ch exhibits significant weight loss at temperatures about 450 °C, whereas little weight loss is registered for the GO and Ch. The rate of weight loss of GO-Ch at 0–250 °C is significantly lower than GO and Ch, it is apparent that GO-Ch has the different decomposition patterns as compared to pure GO and Ch, indicating the Ch is successfully grafted on the surface GO. Information about the differences in the chemistry of the surfaces can be seen in the DTA curves (Fig. 1D). The first peak centered at about 80–90 °C for all samples is due to the evaporation of physically adsorbed water. For the GO sample, the second peak presented between 250 and 300 °C is related to the decomposition of massive oxygenated functional groups such as epoxy and carboxyl groups.²⁹ It is observed that Ch displays an exothermic peak around 250 °C, which can be attributed to the decomposition of the hydroxyl and amino groups of chitosan.³⁰ The broad peak at about 500–600 °C for Ch can be assigned to the further decomposition of chitosan residues. Two sharp exothermic peaks of GO-Ch at around 200 and 450 °C represent the degradation of GO and Ch, respectively. According to Fig. 1E, the pH_pzc of GO and GO-Ch is 4.12 and 5.45, respectively.

Fig. 1 Characterization of GO and GO-Ch composites. A and B: SEM images of GO and GO-Ch composites; C and D: TGA and DTA analysis; E: potentiometric acid-base titrations, respectively.

The FTIR spectra of GO, Ch, GO-Ch before and after adsorption are showed in Fig. 2. It is observed that GO nanosheets show characteristic bands at 1057.76 (epoxy groups C–O), 1620.87 (C=C), 1727.91 (C=O bonds from carboxyl groups) and the broad band at 2200–3800 cm⁻¹ (C–OH stretching from adsorbed water).¹² The FTIR spectrum of Ch exhibits the stretching vibration mode of –OH group at 3200–3400 cm⁻¹, –C=O stretching vibration of –NHCO (amide I) at 1647.88 cm⁻¹, the –NH– deformation of –NH₂ at 1597.74 cm⁻¹, and the amino group (−NH₂) at 1420.32 cm⁻¹. It should be noted that the peak of GO-Ch at 1727.91 cm⁻¹ (carboxyl groups) disappears as compared with GO, moreover the peak density of –NH₂ groups at 1420.32 cm⁻¹ on GO-Ch decreases while the increase of characteristic peaks of –NHCO– and –NH– (at 1647.88 and 1597.74 cm⁻¹) are observed. This is evidence that the –NH₂ groups on the macromolecular chain of Ch can react with carboxyl groups of GO under cross-linking agent conditions. The GO-Ch composites are synthesized by the interaction between carboxyl groups of GO and amino groups of Ch.

Fig. 2 FTIR spectra of the controls including GO, Ch and GO-Ch, and U(VI)-adsorbed samples including GO and GO-Ch at different pH values.

3.2. Adsorption kinetics

Fig. 3A shows the effect of contact time on U(VI) adsorption on GO-Ch and GO. The adsorption rates of U(VI) sharply increase during the initial 30 min, then increase slowly to reach equilibrium at 2 h, which is consistent with previous studies.^11,13 The initially rapid adsorption may be attributed to the relatively larger available sites on the adsorbent. It is obvious that GO-Ch exhibits greater adsorption capacity for U(VI) than GO. Approximately 75% and 65% of U(VI) are adsorbed on GO-Ch and GO at 24 h, respectively. The pseudo-first-order and pseudo-second-order kinetic equations are applied to describe the experimental data. Their linear forms are given in eqn (3)³¹ and (4),¹² respectively:

ln(q_e − q_t) = ln q_e − k_f × t (3)

t/q_t = 1/(k_s × q_e²) + t/q_e (4)

where q_e and q_t (mg g⁻¹) are the amount of U(VI) adsorbed at equilibrium and at time t, respectively. k_f and k_s are the pseudo-first order and pseudo-second order kinetic rate constant, respectively. As shown in Fig. 3A, the adsorption of U(VI) on GO-Ch and GO can be satisfactorily fitted by pseudo-second-order kinetic equation (R² > 0.99), which indicates that the rate-limiting step of U(VI) adsorption on GO-Ch and GO are chemisorptions rather than interparticle diffusion.³¹ Table 1 shows the equilibrium time of U(VI) adsorption on different adsorbents, GO-Ch has a shorter time to achieve equilibrium.

Fig. 3 A: Adsorption kinetics of U(VI) on GO and GO-Ch (A) pH = 4.0, T = 303 K, C₀ = 1.7 × 10⁻⁴ mol L⁻¹, I = 0.01 mol L⁻¹ NaClO₄. The inserted curves: the fitting of pseudo-second order kinetic model; B: breakthrough curve of U(VI) on GO-Ch, pH = 4.0, T = 303 K, C₀ = 3 × 10⁻³ mol L⁻¹, I = 0.01 mol L⁻¹ NaClO₄.

Table 1 Comparison of the U(VI) sorption capacity of GO-Ch with other sorbents

Material	Experimental conditions	Qmax (mg g^−1)	Equilibration time (h)	Reference
Nanoporous alumina	pH 6.8, T = 298 K	11.6	4	37
Multi-walled carbon nanotubes	pH 5.0, T = 298 K	26.18	6	38
Akaganéite	pH 6.0, T = 303 K	90.44	8	39
GO	pH 4.0, T = 303 K	208.33	2	16
GO-Ch	pH 4.0, T = 303 K	225.78	2	This work
PANI@GO	pH 3.0, T = 298 K	242.52	2	12

---

The breakthrough curve of U(VI) on the GO-Ch is depicted in Fig. 3B. The breakthrough is shown as the ratio of the outlet to the inlet concentrations (C/C₀) as a function of the ratio of eluted pore to adsorbent volume (V/V₀). As shown from Fig. 3B, the C/C₀ of influent samples was zero before the pore volume was 50. Even the pore volume of influent reached 170, the C/C₀ of influent samples was less than 0.93. These results indicate that the GO-Ch composites have a high kinetic adsorption capacity.

3.3. Effect of pH and ionic strength

The effect of pH on the adsorption of U(VI) on GO and GO-Ch is shown in Fig. 4(A and B). As shown in Fig. 4A, U(VI) adsorption on GO increases with increasing pH from 3.0 to 5.0, then maintains this high level at pH 5.0-7.0. The decreased adsorption is observed at pH > 7.0, which is agreement with our previous work.¹⁶ The deprotonation of GO increases with increasing pH, especially above the pH_pzc = 4.12, which favors the electrostatic attraction between negative charged GO and positive charged U(VI). Ding et al.¹⁶ has determined that the increasingly U(VI) hydrolyzed species (e.g., UO₂OH⁺, (UO₂)₃(OH)₅⁺, (UO₂)₄(OH)₇⁺ and UO₂(OH)₂ species) are formed at pH > 4.0. However, at pH > 7.0, the negatively charged U(VI) species (i.e., UO₂(OH)₃⁻ and (UO₂)₃(OH)₇⁻ species) increase, which probably explains the suppressed adsorption at pH > 7.0 because of the electrostatic repulsion between negative U(VI) species and negative charged GO. It should be noted that the decreased adsorption at pH > 7.0 is not observed for GO-Ch (Fig. 4B) due to the different functional groups contributed to U(VI) adsorption. It is worth noting that the pHpzc of GO-Ch (about 5.45) is higher than GO (Fig. 1E) due to the introduction of –NH₂ groups. Sun et al.¹² has demonstrated that the enriched amino and amido groups on PANI@GO have a high affinity for U(VI) at alkaline pH because of the formation of strong complexes with cations. Therefore, residual –NH₂ groups can also react with the negative charged U(VI), thus, keeping the adsorption in plateau at pH > 7.0. It is also worth noting that U(VI) adsorption on GO-Ch is higher than that on GO at variable pH due to the presence of oxygen and nitrogen-containing functional groups of GO-Ch composites.

Fig. 4 The effect of pH and ionic strength on U(VI) adsorption onto GO and GO-Ch, T = 303 K, C₀ = 1.7 × 10⁻⁴ mol L⁻¹.

The effect of ionic strength on U(VI) adsorption on GO and GO-Ch are also shown in Fig. 4A and B. Ion strength effect is performed by using NaClO₄ since ClO₄⁻ has little capacity to coordinate with metal ions.³² Therefore, NaClO₄ would not affect U(VI) species with varied pH and would contribute to the mechanistic description of U(VI) adsorption on GO-Ch. The increasing ionic strength has little effect on the adsorption of U(VI) on GO and GO-Ch throughout a wide pH range, which indicates that inner-sphere surface complexation dominates U(VI) adsorption on GO and GO-Ch.^11,12 The inner-sphere surface complexes are formed by the complexation of adsorbate with amphoteric groups by chemical bonds, which are less ionic strength-dependent.^33,34 The oxygen/nitrogen-containing functional groups contribute to the inner-sphere adsorption of U(VI) and GO or GO-Ch by chemical bonding.

3.4. Adsorption isotherms and thermodynamics

Adsorption isotherms of U(VI) on GO and GO-Ch are displayed in Fig. 5A and B. The adsorption of U(VI) on GO and GO-Ch increases with increasing solution concentration. It can be observed that the adsorption amount of U(VI) on GO-Ch is higher than GO, indicating that GO-Ch composites present greater adsorption capacity for U(VI) compared to GO. The adsorption isotherms are fitted by Langmuir, Freundlich³⁵ and D-R models,³⁶ their linear formations can be given in eqn (5)–(7), respectively:

C_eq/Q_s = 1/(b × Q_max) + C_eq/Q_max (5)

log Q_s = log K_F +1/n log C_eq (6)

ln Q_s = ln Q_max − βξ² (7)

where Q_max (mg g⁻¹) is the maximum adsorption capacity of adsorbent at complete monolayer coverage. b (L mg⁻¹) is a Langmuir constant, 1/n is the heterogeneity of the adsorption sites, K_F represents equilibrium coefficient, β is the activity coefficient related to the mean adsorption energy (mol² kJ⁻²), and ξ is the Polanyi potential, which is equal to eqn (8):

ξ = RT ln(1 + 1/C_eq) (8)

where R is ideal gas constant (8.314 J (mol⁻¹ K⁻¹)), and T is the absolute temperature in Kelvin (K).

Fig. 5 A and B: adsorption isotherm of U(VI) onto GO and GO-Ch, respectively; C and D: adsorption isotherm of U(VI) onto GO and GO-Ch in simulating wastewater; E and F: desorption of U(VI) from GO and GO-Ch, respectively; pH = 4.0, C₀ = 1.7 × 10⁻⁴ mol L⁻¹, I = 0.01 mol L⁻¹ NaClO₄.

The corresponding parameters of Langmuir, Freundlich and D–R models are tabulated in Table 2. The fitting results reveal that the adsorption of U(VI) on GO and GO-Ch can be well described by Langmuir model (R² > 0.99), which indicates U(VI) adsorption on GO and GO-Ch are monolayer adsorption. The maximum adsorption capacities of GO and GO-Ch calculated from Langmuir model at pH 4.0 and T = 303 K are 190.10 and 225.78 mg g⁻¹ for U(VI), respectively. GO-Ch exhibits superior performance for U(VI) removal as compared with other materials (Table 1) reported in previous researches. The effect of temperature on the adsorption of U(VI) on the two adsorbents is also shown in Fig. 5A and B. For GO and GO-Ch, the adsorption increases with increasing temperature from 303 K to 323 K, suggesting U(VI) adsorption on GO and GO-Ch are promoted at higher temperature. A thermodynamic analysis is realized based on the equilibrium data for isotherms at three different temperatures. The Gibbs free energy change (ΔG⁰, kJ mol⁻¹) of the adsorption process is expressed by the van't Hoff equation (eqn (9)):

ΔG⁰ = −RT ln(K⁰). (9)

Table 2 Equilibrium parameters for the adsorption of U(VI) onto GO and GO-Ch composites at 303, 313, and 323 K

	T (K)	Langmuir equation			Freundlich equation			D–R equation
		Qmax (mg g^−1)	b (L mg^−1)	R^2	kF (mg^1−n L^n g^−1)	1/n	R^2	Qmax (mg g^−1)	B (mg^2 kJ^−2)	R^2
GO	303	190.10	0.213	0.989	2.22	3.679	0.829	136.83	28.98	0.872
	313	205.70	0.394	0.996	5.75	2.403	0.994	135.33	12.24	0.991
	323	222.53	0.435	0.997	40.55	0.908	0.946	165.89	3.22	0.941
GO-Ch	303	225.78	0.129	0.974	81.28	0.458	0.970	165.14	1.69	0.847
	313	231.72	0.130	0.956	87.10	0.446	0.928	174.94	1.16	0.870
	323	240.72	0.151	0.940	103.51	0.414	0.934	188.21	0.85	0.904

---

The sorption equilibrium constant ln K⁰ can be calculated by plotting ln K_d versus C_e and extrapolating C_e to zero. The adsorption coefficient (K_d) can be calculated by eqn (10):

K_d = Q_s/C_e (10)

where Q_s (mg L⁻¹) is the amount adsorbed on the solid at equilibrium.

The change in entropy [ΔS⁰, kJ (mol⁻¹ K⁻¹)] and the enthalpy of adsorption (ΔH⁰, kJ mol⁻¹) at a constant temperature T (K) can be calculated from the eqn (11):

ΔG⁰ = ΔH⁰ − TΔS⁰. (11)

Based on eqn (9)–(11), one can write

ln K⁰ = ΔS⁰/R − ΔH⁰/RT. (12)

The values of ΔH⁰ and ΔS⁰ are calculated from a plot of ln(K_d) versus 1/T. The negative values of ΔG⁰ (−11.49 kJ mol⁻¹ at 303 K) testify that U(VI) adsorption on GO-Ch is spontaneous process. The value of ΔG⁰ of U(VI) on GO-Ch is lower than that on GO, suggesting that U(VI) is more prone to be adsorbed on GO-Ch. The positive values of ΔH⁰ (25.20 kJ mol⁻¹) suggest the endothermic nature of the adsorption process, which are consistent with the increased adsorption capacities with the rise of temperature. The positive values of ΔS⁰ (117.14 J mol⁻¹ K) express the increased randomness at the solid–liquid interface, which also reveals U(VI) adsorption on GO-Ch and GO are a spontaneous process. Thermodynamic parameters calculated from temperature-dependent adsorption isotherms suggests that U(VI) adsorption on GO-Ch is an endothermic and spontaneous process.

The adsorption isotherms of U(VI) in simulated wastewater are shown in Fig. 5C and D. In simulated wastewater, the maximum adsorption of U(VI) on GO and GO-Ch at 303 K and pH 4.0 are ∼150 and 180 mg g⁻¹, respectively. Although both adsorbents exhibit decreased adsorption capacity in simulated wastewater, the superior performance of GO-Ch for U(VI) removal is still observed compared with GO. The poor U(VI) removal in synthetic groundwater was also evidenced by previous studies.^40,41 The synthetic wastewater contains a range of cationic and anionic chemical components: K⁺, Mg²⁺, Na⁺, Ca²⁺, CO₃²⁻, HCO₃⁻ and NO₃⁻ etc., which cause the competing reactions. It was reported that carbonate can form complexes with U(VI) (carbonato-U(VI) complexes) which is more stable in solutions, thus decreasing U(VI) adsorption.⁴²

3.5. Desorption isotherms and regeneration

The desorption isotherms (desorption induced by replacing radionuclide supernatant) of U(VI) from GO and GO-Ch are shown in Fig. 5E and F. As shown in Fig. 5E and F, the separation of the lines is observed between desorption and adsorption curves, indicating the irreversible desorption hysteresis.⁴³ The irreversible hysteresis indicates that desorption can be not completely achieved without intervention under same pH value.

From the viewpoints of low economic cost and environmental renewable, it is needful to take advantage of renewable material. On this account, the recycle use of GO-Ch is checked to assess its application potential in the actual environment for purifying of U(VI)-bearing outlet water. Acid wash could be a simple and feasible way for the recycling of U(VI)-loaded GO-Ch. Our results show that 0.5 mol L⁻¹ HCl solution can completely desorbed U(VI) from GO-Ch. The results (Fig. 6) reveal that the adsorption ability of GO-Ch slightly reduces from 220 mg g⁻¹ to 200 mg g⁻¹ after 8 recycles. the relationship between number of regeneration cycles and the after 8 cycles. The outstanding power of regeneration and the excellent performance of repeated use indicated that GO-Ch could endure long term use of wastewater treatment.

Fig. 6 Recycling of GO-Ch in the removal of U(VI), pH = 4.0, T = 303 K, C₀ = 1.7 × 10⁻⁴ mol L⁻¹, I = 0.01 mol L⁻¹ NaClO₄.

3.6. Interaction mechanism

The interaction mechanism between U(VI) and GO-Ch is demonstrated by FTIR analysis. As shown in Fig. 2, the spectrum of U-adsorbed GO presents a decrease in the peak at 1727.91 cm⁻¹ (C=O bonds from carboxyl groups) versus GO, indicating the carboxyl groups on GO is responsible for the adsorption of U(VI). The decrease of –NHCO– peaks (at ∼1630 and ∼1566 cm⁻¹) and –NH₂ peaks (at ∼1420 cm⁻¹) are also observed for U-adsorbed GO-Ch at pH 4.0 and 8.0 as compared with GO-Ch, indicating that the nitrogen containing functional groups are responsible for U(VI) adsorption on GO-Ch. The analysis of FITR reveals that the adsorption of U(VI) on GO and GO-Ch depend on different functional groups, and the higher adsorption on GO-Ch is attributed to the introduction of amine groups.

Fig. 7 shows the uranium L_III-edge EXAFS spectra Fourier transforms (FT) of references (UO₂²⁺) and adsorption samples at different pH conditions. The fitted results are summarized in Table 3. The similar spectra of GO and GO-Ch samples at pH 4.0 and 8.0 are observed. The FT peaks centered at ∼1.4 and 1.9 Å can be satisfactorily fitted by 2 axial O atoms (at 1.8 Å) and 4–5 equatorial O atoms (at ∼2.5 Å), respectively, which was consistent with the structural results obtained previously for the fully hydrated uranyl ion.⁴³ The FT peaks at 2.9 Å can be fitted by U–C/N shell,^44,45 revealing the formation of stable inner-sphere complexes of UO₂²⁺ with GO and GO-Ch. This is consistent with the results of FT-IR. The FT feature of GO and GO-Ch (pH 8.0) at 3.5 Å can be fitted by U–U shell at 3.94 Å very well (Table 3), indicating surface precipitation of U(VI) onto the GO and GO-Ch at high pH conditions.⁴⁶ This result can mainly explain the high adsorption of U(VI) on GO and GO-Ch between pH 6–8. Therefore, the results from EXAFS analysis indicate that the adsorption of U(VI) at pH 4.0 and 8.0 could be attributed to the inner-sphere surface complexation and surface co-precipitation, respectively.

Fig. 7 U L_III-edge Fourier transforms (FT) EXAFS data, T = 303 K, I = 0.01 mol L⁻¹ NaClO₄.

Table 3 EXAFS parameters of U(VI) L_III-edge, T = 303 K, I = 0.01 mol L⁻¹ NaClO₄

Samples	Shell	R^a (Å)	CN^b	σ^2^c (Å^2)
a R is the bond distance.b CN is coordination numbers of neighbors.c σ^2 is the Debye–Waller factor.
UO2^2+	U–Oax	1.74(8)	1.8(7)	0.00466
	U–Oeq	2.25(9)	5.5(5)	0.00573
GO pH 4.0	U–Oax	1.81(1)	2.0(0)	0.00425
	U–Oeq	2.45(8)	4.6(9)	0.00666
	U–C/N	2.93(2)	3.2(5)	0.00482
GO pH 8.0	U–Oax	1.80(1)	2.0(0)	0.00224
	U–Oeq	2.39(1)	4.1(3)	0.00923
	U–C/N	2.91(6)	3.0(7)	0.00226
	U–U	3.95(8)	0.5(7)	0.03274
GO-Ch pH 4.0	U–Oax	1.75(8)	2.0(0)	0.00250
	U–Oeq	2.41(7)	4.5(5)	0.00522
	U–C/N	2.95(8)	1.4(2)	0.01464
GO-Ch pH 8.0	U–Oax	1.79(1)	2.0(0)	0.00081
	U–Oeq	2.35(4)	4.5(2)	0.00987
	U–C/N	2.94(8)	1.3(0)	0.01903
	U–U	3.94(7)	0.5(8)	0.02285

---

4. Conclusions

As prepared GO-Ch was achieved by grafting Ch on the surface of GO via covalent method. The adsorption kinetics showed that the adsorption of U(VI) on GO and GO-Ch can be followed by pseudo-second-order kinetic model very well. The maximum sorption capacity of U(VI) on GO-Ch composites calculated from the Langmuir models was 225.78 mg g⁻¹ at pH 4.0 and T = 303 K, which is slightly higher than that of GO (∼190 mg g⁻¹). The thermodynamic parameters suggested that adsorption of U(VI) on GO-Ch was an endothermic and spontaneous process. The batch desorption revealed the irreversible adsorption of U(VI) on GO-Ch. Based on the analysis of FTIR and EXAFS, the interaction mechanism between U(VI) and GO-Ch was inner-sphere surface complexation. It is determined that GO-Ch composite exhibits higher adsorption capacities than any other natural materials in the preconcentration of U(VI) ions from large volumes of aqueous solutions, which will further enhance the application of GO-based composite in environmental pollution management.

Acknowledgements

Financial support from National Natural Science Foundation of China (21207135, 21207136 and 91126020), National “863” Plan (no. 2011AA10A10401) and Hefei Center for Physical Science and Technology (2012FXZY005) are acknowledged.

References

1. C. Basset, O. Averseng, P.-J. Ferron, N. Richaud, A. s. Hagège, O. Pible and C. Vidaud, Chem. Res. Toxicol., 2013, 26, 645–653.
2. M. Carrière, C. Thiebault, S. Milgram, L. Avoscan, O. Proux and B. Gouget, Chem. Res. Toxicol., 2006, 19, 1637–1642.
3. L. He, M. Neu and L. Vanderberg, Environ. Sci. Technol., 2000, 34, 1694–1701.
4. Q. Y. Wu, J. H. Lan, C. Z. Wang, C. L. Xiao and Y. L. Zhao, J. Phys. Chem. A, 2014, 118, 2149–2158.
5. Z. Li, F. Chen and L. Yuan, Chem. Eng. J., 2012, 210, 539–546.
6. Y. Sun, J. Li and X. Wang, Geochim. Cosmochim. Acta, 2014, 140, 621–643.
7. G. Zhao, J. Li, X. Ren, C. Chen and X. Wang, Environ. Sci. Technol., 2011, 45, 10454–10462.
8. G. Sheng, Y. Li, X. Yang, X. Ren, S. Yang, J. Hu and X. Wang, RSC Adv., 2012, 2, 12400–12407.
9. X. Yang, C. Chen, J. Li, G. Zhao, X. Ren and X. Wang, RSC Adv., 2012, 2, 8821–8826.
10. A. Y. Romanchuk, A. S. Slesarev, S. N. Kalmykov, D. V. Kosynkin and J. M. Tour, Phys. Chem. Chem. Phys., 2013, 15, 2321–2327.
11. Y. Sun, Q. Wang, C. Chen, X. Tan and X. Wang, Environ. Sci. Technol., 2012, 46, 6020–6027.
12. Y. Sun, D. Shao, C. Chen, S. Yang and X. Wang, Environ. Sci. Technol., 2013, 47, 9904–9910.
13. Q. J. Du, J. K. Sun, Y. H. Li, X. X. Yang, X. H. Wang, Z. H. Wang and L. H. Xia, Chem. Eng. J., 2014, 245, 99–106.
14. T. Yang and H. Chen, RSC Adv., 2014, 4, 29357–29364.
15. C. Yong, H. Yong, L. Yan and G. Jia, RSC Adv., 2013, 3, 7254–7258.
16. C. Ding, W. Cheng, Y. Sun and X. Wang, Dalton Trans., 2014, 43, 3888–3896.
17. A. Green and M. C. Hersam, J. Phys. Chem. Lett., 2009, 1, 544–549.
18. G. L. Li, G. Liu, M. Li, D. Wan, K. Neoh and E. Kang, J. Phys. Chem., 2010, 114, 12742–12748.
19. D. Chen, H. Feng and J. Li, Chem. Rev., 2012, 112, 6027–6053.
20. S. Hasan, A. Krishnaiah, T. K. Ghosh, D. S. Viswanath, V. M. Boddu and E. D. Smith, Ind. Eng. Chem. Res., 2006, 45, 5066–5077.
21. Y. Zhang, C. Gao, W. Zhao, Z. Zhou, W. Yan, X. Li, Y. Liu, Z. Sun, G. Zhao and J. Gao, RSC Adv., 2014, 4, 33840–33847.
22. S. Peng, H. Meng, Y. Ouyang and J. Chang, Ind. Eng. Chem. Res., 2014, 53, 2106–2113.
23. H. Liu and C. Wang, RSC Adv., 2014, 4, 3864–3872.
24. W. Wan Ngah, L. Teong and M. Hanafiah, Carbohydr. Polym., 2011, 83, 1446–1456.
25. G. Crini and P.-M. Badot, Prog. Polym. Sci., 2008, 33, 399–447.
26. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
27. R. Artinger, W. Schuessler, T. Schaefer and J. Kim, Environ. Sci. Technol., 2002, 36(20), 4358–4363.
28. S. Yan, B. Hua, Z. Bao, J. Yang, C. Liu and B. Deng, Environ. Sci. Technol., 2010, 20, 7783–7789.
29. M. Seredych, R. Pietrzak and T. J. Bandosz, Ind. Eng. Chem. Res., 2007, 46, 6925–6935.
30. N. Travlou, G. Kyzas, N. Lazaridis and E. Deliyanni, Langmuir, 2013, 29, 1657–1668.
31. S. Lagergren, K. Sven. Vetenskapsakad. Handl., 1898, 24, 1–39.
32. E. Tertre, S. Castet, G. Berger, M. Loubet and E. Giffaut, Geochim. Cosmochim. Acta, 2006, 70, 4579–4599.
33. T. Rabung, M. Pierret, A. Bauer, H. Geckeis, M. Bradbury and B. Baeyens, Geochim. Cosmochim. Acta, 2005, 69, 5393–5402.
34. V. Sinitsyn, S. Aja, D. Kulik and S. Wood, Geochim. Cosmochim. Acta, 2000, 64, 185–194.
35. M. LeVan and T. Vermeulen, J. Phys. Chem., 1981, 85, 3247–3250.
36. G. Wood, Carbon, 2002, 40, 231–239.
37. Y. Sun, S. Yang, G. Sheng, Z. Guo, X. Tan, J. Xu and X. Wang, Sep. Purif. Technol., 2011, 83, 196–203.
38. D. Shao, Z. Jiang, X. Wang, J. Li and Y. Meng, J. Phys. Chem. B, 2009, 113, 860–864.
39. S. Yusan and S. Akyil, J. Hazard. Mater., 2008, 160, 388–395.
40. R. Crane, M. Dickinson, I. Popescu and T. Scott, Water Res., 2011, 45(9), 2931–2942.
41. T. Scott, I. Popescu, R. Crane and C. Noubactep, J. Hazard. Mater., 2011, 186(1), 280–287.
42. Y. Sun, J. Li and X. Wang, Geochim. Cosmochim. Acta, 2014, 140, 621–643.
43. T. Undabeytia, S. Nir, G. Rytwo, C. Serban, E. Morillo and C. Maqueda, Environ. Sci. Technol., 2002, 36, 2677–2683.
44. C. Gaillard, A. Chaumont, I. Billard, C. Hennig, A. Ouadi and G. Wipff, Inorg. Chem., 2007, 46, 4815–4826.
45. C. Lucks, A. Rossberg, S. Tsushima, H. Foerstendorf, A. C. Scheinost and G. Bernhard, Inorg. Chem., 2012, 51, 12288–12300.
46. G. Vazquez, C. Dodge and A. Francis, Inorg. Chem., 2008, 47, 10739–10743.

---