Synthesis of C@Ni-Al LDH HSS for efficient U-entrapment from seawater

In this paper, a double hollow spherical shell composite modified with layered double hydroxide (C@Ni-Al LDH HSS) was fabricated for uranium(VI) (U(VI)) adsorption. Various batch experiments were carried out to investigate the influence of pH, concentration, time and coexistence ion on extraction. The results showed that the adsorption processes of U(VI) onto C@Ni-Al LDH HSS were spontaneous and endothermic and closely followed pseudo-second-order and Langmuir isotherm models. The equilibrium time and maximum adsorption capacity of C@Ni-Al LDH HSS was 360 min and 545.9 mg g−1. FT-IR and XPS analyses proved that the adsorption behavior was primarily attributed to the strong interaction between oxygen-containing functional groups and U(VI). Moreover, the extraction of trace U(VI) (μg L−1) in artificial and natural seawater was also studied. The results showed that C@Ni-Al LDH HSS provided a promising application for the efficient extraction of U(VI) from seawater.

The key factors affecting the extraction of U(VI), such as pH, concentration, temperature and contact time, were exploited. Moreover, for economic reasons, the adsorption-desorption of the adsorbent was investigated five times. Finally, the extraction of trace concentration of U(VI) (μg L −1 ) in artificial and natural seawater was also studied.

Results and Discussion
Characterization. To determine the successful synthesis of C@Ni-Al LDH HSS, TEM measurement was used (Figs 1 and S1). For Fig. S1a, the particle size of the SiO 2 was approximately 400 nm, with a smooth surface and good dispersibility. After RF (resorcinol-formaldehyde) in situ polymerization, the precursor SiO 2 @RF was thicker (54 nm) than that of SiO 2 , indicating that the RF successfully coated the surface of SiO 2 (Fig. S1b). After calcination, the thickness of the carbon shell of the SiO 2 @C microspheres was 36 nm (Fig. S1c), suggesting that the RF layer shrunk during calcination. HCS showed a hollow spherical shell structure, indicating that SiO 2 had been successfully removed (Fig. 1a). After AlOOH coating on the HCS surface by a layer by layer deposition process, the thickness of the spherical shell was thickened from 25 to 30 nm (Fig. 1b). Figure 1c shown that the Ni-Al LDH HSS nanosheets uniformly and densely self-assembled on the coordination sites of the HCS shell. After hydrothermal treatment, C@AlOOH HSS was converted into C@Ni-Al LDH HSS double-shell composites. As illustrated in Fig. 1d, the Ni-Al LDH HSS consisted of a large amount of heterogeneous nanoflakes which were superimposed on each other.
The double-hollow spherical structure of C@Ni-Al LDH HSS was further confirmed by the aid of STEM and elemental mapping. It was clear that the hollow architecture of C@Ni-Al LDH HSS had an average diameter of 400 nm (Fig. 2a), consistented with the TEM image (Fig. 1c). Further, the elemental mapping (Fig. 2b) clearly verified that the C, Al, Ni, and O elements were homogeneously distributed in C@Ni-Al LDH HSS.
The XPS spectra of HCS, Ni-Al LDH HSS and C@Ni-Al LDH HSS were shown in Fig. 3b. Compared with HCS, the new peaks of Ni-Al LDH and C@Ni-Al LDH indicated the appearance of Ni and Al elements in the two HSS. For C@Ni-Al LDH HSS (Fig. S2a), the C 1 s spectra could be divided into C-C, C-O and O-C=O, which located at 284.8, 286.1and 289.1 eV, respectively 43,44 . In the O 1 s spectra of C@Ni-Al LDH HSS (Fig. S2b), the peaks with binding energies of 530.8, 531.7 and 533.3 eV correspond to M-O, -OH and H 2 O, respectively 45,46 .
To further study the detailed structural characteristics of the adsorbent, the N 2 adsorption-desorption isotherms and the corresponding pore size distributions of HCS, Ni-Al LDH HSS and C@Ni-Al LDH HSS were shown in Fig. 4. All isotherm curves were type IV that possessed a typical H 3 -type hysteresis loop, which suggested that the existence of mesoporous structures 47 . The BET surface area of three samples decreased in the following order: C@Ni-Al LDH HSS (207.69 m 2 g −1 ) > HCS (166.08 m 2 g −1 ) > Ni-Al LDH HSS (21.70 m 2 g −1 ). www.nature.com/scientificreports www.nature.com/scientificreports/ Concurrently, the respective average pore diameters and pore volumes of each of the above were 17.29 nm and 0.52 cm 3 g −1 , 7.36 nm and 0.18 cm 3 g −1 , and 17.16 nm and 0.12 cm 3 g −1 , respectively. The BJH average pore sizes of the samples were all in the range of 2-50 nm (inset in Fig. 4), further indicating the existence of mesoporous features. Clearly, the C@Ni-Al LDH HSS possessed a larger specific surface area and pore volume than the pure Ni-Al LDH HSS did. In conclusion, the addition of HCS prevented the aggregation of Ni-Al LDH nanosheets, causing more surfaces. Simultaneously, the high surface area of C@Ni-Al LDH HSS furnished more active sites and enhanced the removal of U(VI) ions.
Effect of initial pH. pH played an important role, influencing the distribution of U(VI) species in the solution and the surface properties of the adsorbents. Thus, the impact of the pH upon the U(VI) adsorption behavior of Ni-Al LDH HSS and C@Ni-Al LDH HSS was investigated in the pH range of 2.0-10.0 and shown in Fig. 5. As could be seen, the adsorption capacity of Ni-Al LDH HSS and C@Ni-Al LDH HSS were closely related to pH. For the two materials, the amount of adsorbed U(VI) sharply increased with an increase from pH 2.0-4.0. At a low pH, the poor adsorption was attributed to electrostatic repulsion between U(VI) and the positively charged surface of the protonated adsorbent [48][49][50] . As the pH increased, the adsorption of U(VI) increased due to the deprotonation of hydroxyl functional groups. At pH 6.0-10.0, the adsorption amount sharply decreased with an increase in pH. This was due to [UO 2 (CO 3 ) 2 ] 2− and [UO 2 (CO 3 ) 3 ] 4− etc. predominating in the presence of CO 2 , leading to a reduction in adsorption efficiency 51 .

Adsorption kinetics.
To determine the effect of the reaction time on the adsorption kinetics, the change in U(VI) adsorption capacity on materials with contact time was investigated. As illustrated in Fig. 6a, the adsorption capacity of U(VI) on C@Ni-Al LDH HSS rapidly increased with shaking time until the adsorption reached equilibrium at 360 min. Ni-Al LDH HSS reached equilibrium at 480 minutes. The maximum adsorption capacity of the U(VI) of C@Ni-Al LDH HSS (545.9 mg g −1 ) was much higher than that of the U(VI) on Ni-Al LDH HSS (343.2 mg g −1 ), demonstrating that U(VI) in aqueous solution could be effectively removed by C@Ni-Al LDH HSS.  www.nature.com/scientificreports www.nature.com/scientificreports/ Based on the experimental data, the kinetic processes of Ni-Al LDH HSS and C@Ni-Al LDH HSS were analyzed using pseudo-first-order, pseudo-second-order and Weber-Morris kinetic models (formulas in ESI. 2). As observed in Fig. 6b,c and Table S1, the kinetics of the adsorption of U(VI) by two adsorbents showed a better fit to the pseudo-second-order model (R 2 Ni-Al LDH HSS = 0.9987 and R 2 C@Ni-Al LDH HSS = 0.9976) than to the pseudo-first-order model (R 2 Ni-Al LDH HSS = 0.9879 and R 2 C@Ni-Al LDH HSS = 0.9886). The results were mainly attributed to the chemical action of the U(VI) ions and the functional groups on the surface of the adsorbents. Figure 6d showed the kinetic experiment fitting data for the intra-particle diffusion model on C@Ni-Al LDH HSS and Ni-Al LDH HSS. The adsorption process consisted of three stages for all materials. In the first portion, diffusion occurred from the bulk phase into the pores, with adsorption taking place on the outside surface of the adsorbents 52 . The first-stage straight line corresponded to rapid adsorption in that the instantaneous adsorption could be due to external diffusion. In the second step, the gradual straight line corresponded to the intra-particle diffusion model, which might be ascribed to the loose porous construction of the adsorbents allowing metal ions to diffuse rapidly to the inner surface. Additionally, the second portion of the linear line did not pass through the origin, indicating that intra-particle diffusion was not the sole rate determining factor for controlling U(VI) adsorption. The last straight line showed that the adsorption rate was controlled by the chemical interaction between U(VI) and the effective active site, consistent with the results of the kinetics study 53 . The order of the rate constant k p values, calculated according to the Weber-Morris formula (Eq. S(3)), was as follows: k p1 > k p2 > k p3 (Table S2). Therefore, the adsorption mechanism was regarded as predominantly chemisorption and partly dependent on the pore size of the U(VI) diffusion.
Adsorption isotherms and thermodynamics. The relation between the saturated adsorption amount and the initial concentration of U(VI) was crucial for optimizing the adsorption process. As shown in Fig. 7a, the amount of adsorbed U(VI) on both adsorbents significantly increased with an increase in the initial U(VI) concentration until equilibrium was reached. The maximum adsorption capacity of C@Ni-Al LDH HSS was greater than that of Ni-Al LDH HSS at 298, 308 and 318 K, suggesting that the C@Ni-Al LDH HSS composite provided more active sites to involve U(VI) adsorption. Moreover, the uptake in U(VI) was improved prominently with the increased in temperature, which was potentially due to the diffusion of U(VI) on the surface of the adsorbents being promoted, and the activity of the functional groups on the surface of the adsorbents was enhanced 54 . Three isothermic Langmuir, Freundlich and D-R models 55 were described as:  www.nature.com/scientificreports www.nature.com/scientificreports/ e where Q m (mg g −1 ) was the maximum adsorption amount at complete monolayer coverage; C e (mg L −1 ) and Q e (mg g −1 ) were the U(VI) concentration and adsorption capacity at equilibrium, respectively; b was a constant related to the affinity and energy of the adsorbents; k was a Freundlich constant and 1/n was associated with the adsorption intensity; β (mol 2 J −2 ) was the D-R constant related to the adsorption free energy and ε (J mol −1 ) was the Polanyi potential; and T(K) and R (8.314 J mol −1 K −1 ) were absolute temperature and gas constants, respectively. As shown in Fig. 7b,d and Table 1 Moreover, the maximum adsorption capacity (Q exp,max ) of C@Ni-Al LDH HSS for U(VI) attained 545.9 mg g −1 at 298 K, 652.4 mg g −1 at 308 K, and 695.1 mg g −1 at 318 K. The high adsorption capacity was attributed to the active sites provided by HCS and Ni-Al LDH HSS.
The ΔS θ , ΔH θ and ΔG θ thermodynamic parameters were calculated by the following formulas: where K d was the equilibrium constant (mL g −1 ); T (K) was Kelvin temperature; R (8.314 J mol −1 K −1 ) was the gas constant; ΔH θ (kJ mol −1 ) was the standard enthalpy change; ΔS θ (J mol −1 K −1 ) was the standard entropy change; and ΔG θ (kJ mol −1 ) was the standard change in Gibbs free energy. The values of ΔH θ and ΔS θ were evaluated from the slope and intercept of the plot of ln Kd vs. 1/T (Fig. 8), with the value of ΔG θ calculated using Eq. (6). From Table 2, the positive value of ΔS θ meaned that the disorder degree of the system increased during U(VI) adsorption, attributable to the structural changed in U(VI) loaded at the solid/solution interface 47 . The negative ΔG θ value and positive ΔH θ value indicated that the processes of U(VI) removal on both absorbents were spontaneous and endothermic 57 . As the temperature increased, the value of ΔG θ became more negative, demonstrating more effective adsorption at higher temperatures. Furthermore, the value of ΔH θ could be used to infer the type of adsorption mechanism. A value of less than 21 kJ mol −1 conformed with physical adsorption, whereas the range of 21-418 kJ mol −1 conformed with chemical adsorption. That ΔH θ Ni-Al LDH HSS = 122.50 kJ mol −1 and ΔH θ C@Ni-Al LDH HSS = 211.71 kJ mol −1 indicate that the adsorption of U(VI) ions is achieved via a chemical mechanism.
Effect of the co-existing ions. To evaluate the selectivity of Ni-Al LDH HSS and C@Ni-Al LDH HSS for U(VI) capture, a solution containing 11 kinds of competing metal cations was used (Table S3). As depicted in Fig. 9a,b, the C@Ni-Al LDH HSS composite exhibited better selectivity for U(VI) adsorption than Ni-Al LDH HSS, maintaining a removal rate of up to 90% in the existence of competing ions. www.nature.com/scientificreports www.nature.com/scientificreports/ The above-mentioned data indicated that the introduction of HCS improved Ni-Al LDH HSS selectivity and that the C@Ni-Al LDH HSS composite possessed an outstanding affinity towards U(VI) among competing metal ions.
The recyclability of C@Ni-Al LDH HSS composite. Due to its good adsorption properties, subsequent experiments were carried out with C@Ni-Al LDH HSS. To assess its practicability and regeneration, 0.1-1.0 mol L −1 Na 2 CO 3 eluents on the desorption of U(VI) were carefully investigated, as shown in Fig. 10a. The desorption efficiency was above 80% when Na 2 CO 3 concentration was greater than 0.5 mol L −1 , suggesting that     www.nature.com/scientificreports www.nature.com/scientificreports/ Na 2 CO 3 was an effective eluent for the recovery of the adsorbent. To evaluate the repeatability of the C@Ni-Al LDH HSS composite, the adsorption-desorption experiments with 0.5 mol L −1 Na 2 CO 3 solution were repeated for five cycles. Figure 10b showed the removal rate of C@Ni-Al LDH HSS for five cycles of experiments. The results showed that there was a decrease of only 10% (from 90% to 80%), indicating that C@Ni-Al LDH HSS possessed good reusability for the efficient removal of U(VI).

Possible mechanism of U(VI) adsorption onto C@Ni-Al LDH HSS.
In order to clarify the interaction mechanism between C@Ni-Al LDH HSS and U(VI), FT-IR and XPS spectroscopies of the C@Ni-Al LDH HSS before and after U(VI) adsorption were recorded (Fig. 11). As can be seen in Fig. 11a, after adsorption, similar peaks of the bands below 800 cm −1 slightly red-shifted to 663 cm −1 , suggesting synergistic effects between U(VI) and metal-oxygen functional groups (M-O). Further, the stretching vibration of the -OH groups red-shift to 3447 cm −1 , showing that complex (such as [-O⋅⋅⋅H⋅⋅⋅U] + or [-O⋅⋅⋅U] + ) formations occurred between U(VI) and C@Ni-Al LDH HSS. Importantly, a new peak at 913 cm −1 corresponded to the antisymmetric stretching vibration of the [O=U=O] 2+ group, indicating that U(VI) was successfully immobilized 58,59 .
The XPS spectra of C@Ni-Al LDH HSS-U showed clear new double U 4f peaks, when compared with C@ Ni-Al LDH HSS (Fig. 11b). As could be seen from the XPS high resolution data in Fig. 11c, strong double U 4f peaks characterized with U 4f 5/2 (393.0 eV) and U 4f 7/2 (382.1 eV) in C@Ni-Al LDH HSS-U appear, revealing that U(VI) was captured onto C@Ni-Al LDH HSS 60,61 . As Fig. 11d showed, when compared with the O 1 s spectra of C@Ni-Al LDH HSS before U(VI) sorption, the binding energy of M-O shifted from 530.8 eV to 531.0 eV, and that of -OH shifted from 531.7 and 532.0 eV. The results suggested that the -OH and M-O functional groups interacted with U(VI) ions.
The possible mechanism of U(VI) adsorption was as follows: firstly, due to the 3D double hollow spherical porous shell structure of C@Ni-Al LDH HSS, U(VI) rapidly interacted with the inner and outer surfaces of the adsorbents; then the high surface area and abundant reactive sites contributed to the complexation of functional groups, such as -OH, Al-O and Ni-O, along with the U(VI) and other uranium species, possibly being the reason for the U(VI) adsorption ability of the prepared C@Ni-Al LDH HSS being better than other LDH-based materials ( Table 4).

Adsorption experiments in artificial and natural seawater.
Based on the aforementioned experimental results, the U(VI) adsorption capacity of C@Ni-Al LDH HSS under both artificial and natural seawater was evaluated. The simulated seawater contained a trace concentration of U(VI) and was formulated according to previous reports 45 , possessing a concentration range of 2.91-74.01 μg L −1 . Fig. 12 showed a high removal rate (more than 90%), implying that C@Ni-Al LDH HSS had potential to effectively extract U(VI) from seawater.
The C@Ni-Al LDH HSS adsorbent was placed in the sea off the coast of Rongcheng, a city in Shandong Province in eastern China. The combination of floating buoys with an anchor made the adsorbent be immersed approximately 3 meters below the surface of the sea stream for 31 days. After 31 days of soaking, the adsorption capacity of C@Ni-Al LDH HSS was 1.24 μg g −1 . The C@Ni-Al LDH HSS adsorbent in natural seawater before and after the process was shown in Fig. S3.

Conclusions
In summary, the 3D double hollow spherical shell C@Ni-Al LDH HSS was successfully synthesized and characterized by TEM, STEM, XPS, XRD, and BET techniques. The kinetics data suggested that the processes of adsorption of U(VI) onto C@Ni-Al LDH HSS and Ni-Al LDH HSS fitted well with pseudo-second-order kinetic model. The equilibrium adsorption capacities of C@Ni-Al LDH HSS and Ni-Al LDH HSS were measured and extrapolated using Langmuir, Freundlich and D-R models, with the experimental data found to best fit a Langmuir model. The reusability studies demonstrated that C@Ni-Al LDH HSS possessed high reusability for the efficient removal of U(VI). The adsorption mechanism of C@Ni-Al LDH HSS was mainly attributed to the higher specific surface area of the www.nature.com/scientificreports www.nature.com/scientificreports/   www.nature.com/scientificreports www.nature.com/scientificreports/ double hollow spherical shell structure and the strong interaction between abundant oxygen-containing functional groups and U(VI). In addition, the adsorption of U(VI) from artificial seawater and natural seawater was assessed, suggesting that it had potential to efficiently immobilize the radioactive element of U(VI) in seawater.

Experimental Section
Materials. Tetraethyl orthosilicate (TEOS) and aluminum isopropoxide (Al(OPr) 3 ) were purchased from the Aladdin Chemistry Co. All other chemical reagents were analytical grade and directly used without any additional treatment.

Synthesis of Ni-Al LDH hollow spherical shell (Ni-Al LDH HSS). SiO 2 microspheres were prepared
from tetraethoxysilane (TEOS) through an improved Stöber method 62,63 . Ni-Al LDH HSS was obtained as follows 64,65 : (1) 0.2 g as-prepared SiO 2 microspheres were placed in 20 mL AlOOH sol at room temperature for stirring overnight. The products were centrifuged, washed with ethanol and dried. This process of centrifugation and washing was repeated five times to obtain SiO 2 @AlOOH. (2) A 70 mL aqueous solution containing 0.2 g SiO 2 @ AlOOH, 2.9 g Ni(NO 3 ) 2 ·6H 2 O and 0.3 g urea was transferred in an autoclave at 100 °C for 48 h and then cooled down naturally to room temperature. (3) The targeted resultant Ni-Al LDH HSS was separated by centrifugation, washed with deionized (DI) water and then dried under freezing.
Synthesis of C@Ni-Al LDH HSS. The intermediate SiO 2 @RF was obtained by SiO 2 microspheres through a polymer coating process (details in ESI. 5) 66 . The intermediate SiO 2 @C spheres were synthesized by the carbonization of the as-prepared SiO 2 @RF, which were heated at 150 °C for 1 h under N 2 atmosphere and subsequently at 800 °C for 3 h in the same atmosphere. After etching in 1.0 mol·L −1 NaOH for 12 h, the intermediate SiO 2 @C spheres were converted into hollow carbon spheres (HCS). The HCS powder was separated by centrifugation, washed and then dried under freezing.
C@Ni-Al LDH HSS was obtained by in situ growth of Ni-Al LDH nanosheets on the surface of HCS by a urea hydrolysis method 67,68 , similar to the preparation method of Ni-Al LDH HSS (details in ESI. 5).
The synthetic route of Ni-Al LDH HSS and C@Ni-Al LDH HSS was illustrated in Fig. 13.
Batch adsorption experiments. 2.110 g UO 2 (NO 3 ) 2 ·6H 2 O was dissolved in 2% nitric acid diluted by DI water, giving a U(VI) concentration of 1000 mg L −1 . The working U(VI) solutions were prepared by a suitable dilution of the original solution with DI water in the following adsorption experiment. Typically, 10 mg adsorbent was added in the 20 mL UO 2 (NO 3 ) 2 ·6H 2 O solution with a given concentration and pH, whose initial pH of the solution was adjusted with 0.5 mol L −1 Na 2 CO 3 or HNO 3 . The mixture was shaken for 6 h in a thermostatic shaker bath at the desired temperature. After adsorption, the adsorbent was centrifugally separated and the solution was determined by ICP-AES or ICP-MS. The adsorption capacity Q e (mg·g −1 ) and the removal rate (R%) of U(VI) were calculated using Eqs (7) and (8):