Selective removal of heavy metal ions in aqueous solutions by sulfide-selector intercalated layered double hydroxide adsorbent

1. Introduction

The pervasive industrialization of human society in past decades has accelerated the contamination of the environment and food chain with heavy metal ions, which has become a serious and public issue. However, the toxicity of some heavy metals, such as Cu, Zn, Ni and Mn, is concentration-dependent; some heavy metals are often essential micronutrients with positive physiological functions in amounts of a few milligrams per day, whereas some metal ions, such as Cd, Pb, As, etc., are toxic if ingested or inhaled [1,2]. In this respect, the adequate intake of some heavy metals has a significant impact on human health, and their intake is important for guaranteeing the growth and development of infants and young children for optimal health [3,4]. On the other hand, the absorption of highly toxic heavy metal ions that are released into the environment can cause a number of health problems, including various diseases and disorders, and should be avoided. Therefore, it is essential and important to remove heavy metal ions from aqueous solutions with the necessary selectivity for human health and food safety.

Adsorption is a method with advantages, such as low cost, simple design and strong operability, and has been considered as one of the most attractive approaches for heavy metal removal. Layered double hydroxides (LDHs), which are known as hydrotalcite-like compounds or anionic clays, have been studied by changing the interlayer anions for the selective adsorption of toxic heavy metal ions [5]. The general formula of this type of lamellar inorganic material is [M²⁺_1-xM³⁺_x(OH)₂]^x+[(A^n-)_x/n]· mH₂O, where M²⁺ and M³⁺ are di- and trivalent metal cations, respectively, x is the molar ratio of M³⁺/(M²⁺ + M³⁺), A^n- is the interlayer anion (or gallery anion), and m is the molar amount of co-intercalated water [6]. LDHs have an abundance of interesting properties, such as a unique layered nanostructure, large surface area and ion exchange ability, and therefore, LDHs have been applied in a wide range of areas, including environmental remediation [7]. Other characteristics, such as the ratio of M²⁺/M³⁺, the type of interlayer anion, the composition of metal elements in the host layers and specific surface area, have important effects on the removal behavior of LDH-based adsorbents [8,9]. However, stone-like and plate-like LDHs are included in the adsorbent composition, which result in low selectivity and a weak affinity for heavy metals [10,11]. Particularly, by intercalating LDH materials with groups of sulfurated species, such as sulfides, thiols and sulfonic groups, substantial improvements have been made, improving the selective sorption properties of LDHs toward heavy metal ions [[12], [13], [14], [15]]. This is a growing class of functional selectors that are soft bases and possess exceptional selectivity and rapid sorption kinetics for soft or relatively soft metal ions [[15], [16], [17], [18], [19]]. For this type of selective heavy metal adsorbent, the multistep procedures are often time consuming and labor intensive. Therefore, the mass production and availability of enormous quantities of these materials at economically viable prices are prerequisites for practical application. Thus, exploring an efficient strategy to overcome the contradiction between the selectivity performance and the complicated fabrication of LDH-based metal ion adsorbents is expected.

In this work, a sulfide group intercalated NiFe-LDH (NFL-S) hierarchical microcomposite has been synthesized by a facile one-pot approach for the selective adsorption of divalent heavy metal ions. Notably, intercalated sulfides and sulfonic groups, which are soft Lewis bases, are capable of the selective adsorption of heavy metal ions. The detailed selective abilities of the resulting NFL-S microcomposites are shown by batch sorption experiments with various divalent heavy metal ions, and the adsorption mechanisms are analyzed by a series of characterizations. Moreover, the as-prepared hierarchical NFL-S microcomposites were also evaluated for the removal of Pb²⁺ and Cu²⁺ in different aqueous samples, and this experiment integrates all the functionalities needed for the capture of efficient heavy metal ions and demonstrates the great potential of NFL-S microcomposites for water safety control.

2. Experimental

First, 0.45 mmol of NiAc₂, 0.05 mmol of Fe(NO₃)₃ and 0.1 g of (NH₄)₂MoS₄ were dissolved in 10 mL of distilled water. Then, 0.12 g of urea was added to the solution and stirred for 60 min. The resulting solution was placed in a 20 mL Teflon-lined autoclave and hydrothermally treated at 160 °C for 10 h. The products were washed with deionized water and collected by centrifugation. Finally, the resulting materials were dried by a vacuum oven at 60 °C for 10 h. All of the reagents were purchased from Sigma-Aldrich.

Scanning electron microscopy (SEM) images were recorded on a Hitachi S-4800 (Japan) instrument. The high-resolution transmission electron microscopic (HRTEM) images were performed on JEOL 2010 F operated at 200 kV. A powder X-ray diffraction (XRD) pattern was obtained on a powder diffractometer (Bruker D8 Advanced Diffractometer System) with Cu Kα (λ = 1.5418 Å) radiation. X-ray photoelectron spectroscopy (XPS) data were collected by an Axis Ultra DLD X-ray photoelectron spectrometer equipped with an Al Kα X-ray source (1486.6 eV). Spectra were obtained by Fourier transform infrared spectroscopy (FTIR) and recorded using a Vetex70 (Bruker Corp, Germany) instrument and KBr pellet at room temperature.

The heavy metal adsorption at various concentrations was studied through batch experiments. The five metal ions, which included Cu²⁺, Zn²⁺, Pb²⁺, Cd²⁺ and Mn²⁺, were obtained from their nitrate salts. The concentration of the residual cations in the supernatant was determined by flame atomic absorption spectroscopy (FAAS) (Z-2000, Hitachi) after centrifugation at 8000 rpm for 5 min. The adsorption capacity (q_t) was calculated by Eq. (1):

The distribution coefficient (k_d) is defined by Eq. (2):

where C₀ and C_f are the initial and equilibrium concentrations of M²⁺ (mg/L), respectively, V is the solution volume (mL), and m is the adsorbent amount (g).

To distinguish the selectivity toward Cu²⁺, Pb²⁺, and Zn²⁺, the ions with concentrations of $\widetilde{1}$0 ppm were mixed with different amounts of NFL-S (1 and 2 mg/L). To investigate the removal capacity of each ion, individual solutions containing Cu²⁺, Zn²⁺, Pb²⁺, Cd²⁺ or Mn²⁺ were prepared and batch experiments were performed at room temperature for 10 h to obtain the equilibrium adsorption.

Different isotherm models, including the Langmuir and Freundlich models, are used to analyze the experimental data, and the mathematical expressions are expressed by Eqs. (3) and (4):

where q_e (mg/g) and q_m (mg/g) in Eq. (3) are the equilibrium adsorption and theoretical maximum adsorption capacities, respectively. The concentration of heavy metal ions at equilibrium was C_e (mg/L). The constant of the Langmuir model k_L (L/mg) reflects the energy of the binding sites and the affinity to adsorb. The constants q_m and k_L are deduced from the slope and intercept of the linear plot of C_e/q_e versus C_e, respectively. In the Freundlich isotherm equation, k_F and n are the constants, which correspond to the adsorption capacity and adsorption intensity, respectively. The constants k_F and n are calculated from the slope and intercept of the linear plot of logq_e versus logC_e, respectively.

The adsorption kinetics experiments were performed for various adsorption time (2-420 min). For each experiment, 0.5 mg/L of NFL-S was added to a solution containing the anions with concentrations of $\widetilde{1}$00 ppm. The suspensions were centrifuged at specified time intervals, and the supernatant solutions were taken and analyzed by FAAS to measure the contents of ions.

Pseudo-first-order and pseudo-second-order models are employed to simulate the experimental data in the kinetic study, which can be expressed by Eqs. (5) and (6):

where the contact time is t (min), and the capacity of the adsorbent is q (mg/g). The adsorption capacity at equilibrium is q_e, and the constant of the pseudo-first-order equation is k₁. In Eq. (6), k₂ (mL·mg^-1·min^-1) is the rate constant of the pseudo-second-order equation, and the equilibrium adsorption capacity of q_e can be calculated from t/q_t versus t.

Apple juice, pear juice, peach juice, mulberry juice and orange juice, were purchased from a local supermarket. All of these drinking samples were treated as follows: 2 mL aliquots of each beverage sample were added into separate beakers and stirred with 3 mL of concentrated HNO₃ (65%, w/w) and 3 mL of HClO₄. Then, the mixture was digested at 90 °C until a clear solution was obtained. The final solution was diluted to 50 mL with deionized water.

3. Results and discussion

3.1. Characterization of the materials

To form the NFL-S microcomposite, Ni²⁺ and Fe³⁺ were added at a molar ratio of 9 during the reaction and hydrolyzed during the gradual thermal decomposition of urea in the presence of the tetrathiomolybdate salt (Scheme 1). The SEM images of NFL-S shown in Fig. 1(a) indicate that the as-synthesized products consist of a high-quality 3D flower-like hierarchical microstructure, which has a relatively uniform size distribution. The detailed features as shown in Fig. 1(b) indicate that the nanosheets have an average thickness of $\widetilde{3}$0 nm, and the nanosheets are interconnected with each other and interlaced by Ostwald ripening. Transmission electron microscopy (TEM) images of NFL-S (Fig. 1(c)) further reveal that the three-dimensional structure of the NFL-S microcomposite was constructed from multiple overlapping LDH nanosheets. A HRTEM image collected from the petal clearly displays well-resolved lattice fringes with an interplanar spacing of approximately 0.25 nm, which could be well-indexed to the (012) plane of NiFe-LDH (Fig. 1(d)). The detailed crystalline structure, as described in Fig. 2(a), was characterized by XRD. From the XRD patterns, the unambiguous characteristic diffraction peaks of (003), (006), (100), (012), (018), and (110) are the same as the standard XRD peaks of the hydrotalcite-like NFL phase (Fig. 2(b)) [20,21]. The as-prepared NFL-S microcomposite can be fit with the majority of unambiguous characteristic diffraction peaks, including the (003), (006), (100), (012), and (110) lattice planes. Other diffraction peaks (at 31.5°, 41.3°,46.2°, 49.9°, and 53.7°) were attributed to the formation of NiMoO₄ (JCPDS No. 16-0291) and FeMoO₄ (JCPDS No. 16-0326) during the hydrothermal process. The weakening of the (006) lattice plane may be attributed to the intercalation of MoS₄^2-, which shifts the position of the diffraction peaks. However, the (110) plane in NFL-S does not change, indicating the stability of the NiFe-LDH layers during the ion-exchange process. Furthermore, the narrow peak and high intensity of (003) suggests that NFL is highly crystalline, and this peak was chosen to calculate the d-spacing [[22], [23], [24], [25], [26], [27]]. The as-prepared NFL-S has an enlarged d_basal of 0.752 nm compared with the 0.711 nm basal spacing (d_basal) of NFL, which could be attributed to the intercalation of sulfide anions into the interlayer space of the NFL nanosheets, thus leading to an expansion of the interlayer spacing. To further verify the composition of the intercalated compound, the valence states of the elements were studied using XPS. The full XPS spectrum shows that Ni, Fe, O, Mo, and S coexist in the NFL-S composite (Fig. 2(c)). The individual elemental spectra shown in Fig. 2(d, e) display that the characteristic peaks of Ni 2p and Fe 2p were located at 879.47/855.42 eV (the spin-energy separation is 24.05 eV) and 725.23/713.42 eV (the spin-energy separation is 11.81 eV), respectively [28,29]. These results correctly prove the intrinsic quality of the NFL structure. Moreover, the exact positions in the range of 160-171 eV correspond to the S 2p of S^-, S^2-and S⁶⁺, which depend on the oxidation states and chemical environments [[16], [17], [18], [19]]. The binding energy located at 162.43 eV contributes to the sulfide state of MoS₄^2- [30]. The S 2p peak at 168.26 eV belongs to the sulfonic group, which might be attributed to the partial oxidation of the tetrathiomolybdate anions (Fig. 2(f)) [31].

View original graphic| Download| PPT slide

Scheme 1.   Synthesis schematic of the flower-like hierarchical NFL-S microstructure and SEM images of the NFL-based composites with the addition of varying amounts of (NH₄)₂MoS₄ (0, 0.05 and 0.1 g).

View original graphic| Download| PPT slide

Fig. 1.   (a,b) SEM images of the flower-like NFL-S hierarchical microcomposite at different scales. (c) TEM images of the as-fabricated NFL-S hierarchical microcomposite. (d) HRTEM image of the NiFe-LDH taken from the image location as shown in (c).

View original graphic| Download| PPT slide

Fig. 2.   (a) Schematic crystalline structure of NFL-S displaying the host layered construction and abundant interlayer anions. (b) XRD patterns of NFL (black) and NFL-S (red) and, for comparison, the inset schematics of the crystalline structures representing the enlarged interlayer space after intercalation. XPS spectra of (c) NFL-S, (d) Ni 2p, (e) Fe 2p and (f) S 2p.

For the octahedrally coordinated d° transition metal cations of Mo⁶⁺, the electronic second-order Jahn-Teller effect can occur once the empty d-orbitals of the metal mix with the filled p-orbitals of the ligands [31,32]. This phenomenon lowers the energy of the occupied valence band states and raises the energy of the empty conduction band states, which would be the driving force for the spontaneous distortion of the d° transition metal cations [[32], [33], [34], [35], [36]]. The distortion of an octahedron can occur along one of three directions: toward an edge, face, or corner [37]. In this work, the existence of Mo⁶⁺ exhibits a unique second-order Jahn-Teller effect, thus enabling the positively charged nanosheets thermodynamically or kinetically to further extend along the c-direction and providing the controlled hierarchical self-assembly for the fabrication of microcomposites from 2D to 3D structures. According to the SEM characterization (Scheme 1), a sheet-like NFL was synthesized immediately during the precipitation of the metal ions and alkaline species. By increasing the amount of ammonium tetrathiomolybdate, different secondary structures are constructed by self-assembling the naonosheets to form 3D microstructures. For further validation, with the presence of MoS₄^2-, a series of LDH-based composites, including MgAl-, NiAl-, and CoAl-LDH nanosheets, could also be assembled into hierarchical microstructures (shown in Fig. S1 in Supporting Information). This demonstrates that Mo⁶⁺ could effectively regulate the topotactic conversion of 2D nanosheets to 3D hierarchical microstructures for LDH-based materials via second-order Jahn-Teller effects. Moreover, the hierarchical microstructure has a larger surface area (67.4 m² g^-1) compared to the NFL nanosheets (13.9 m² g^-1), which provides more binding sites to improve the performance of the adsorbent materials (Fig. S2(a) in Supporting Information). The presence of small mesopores is evidenced by the pore size distributions shown in Fig. S2(b) in Supporting Information, which shows that the size distribution of the small mesopores is approximately 19 nm and the pore volume is 0.418 cm³ g^-1. The hierarchical microstructure is favorable for reducing the distance between the adsorbents and adsorbates, which allows the surface area and active sites of the ultrathin 2D nanosheets to be used sufficiently. Moreover, this structure is supposed to overcome the limitation of mass transfer, thus promoting the diffusion of metal ions during the adsorption process.

This work provides a simple and efficient one-step process for the synthesis of sulfur-functionalized LDH-based adsorbents. In addition, the crystal phase can be transformed from β-Ni(OH)₂ to α-Ni(OH)₂-like LDHs of nickel hydroxide (hydrotalcite phase) due to the introduction of iron hydroxide, according to the literature [28]. This change will increase the interlayer distance and provides more space for NFL to intercalate with sulfide anions, which implies a higher selective adsorption performance compared to that of traditional bimetallic hydroxides.

3.2. Heavy metal removal using NFL-S

The selective adsorption performances of NFL-S and NFL were evaluated for different concentrations of heavy metal ions at room temperature (V: m = 1000 mL g^-1, contact exposure time of 10 h). The distribution coefficient of k_d is a constant that reflects the affinities of NFL-S and NFL for metal ions. At first, the adsorption experiments were performed with five divalent metal ions. Subsequently, the ability of NFL-S and NFL to remove all five ions was evaluated simultaneously. Two sets of metal ion concentrations with low values of 10 ppm and high values of 100 ppm were selected to assess the selectivity and removal capacity of NFL-S and NFL. Table 1 summarizes the removal efficiency of NFL-S and NFL for individual ions with concentrations of 10 ppm, and both materials show higher adsorptions of Pb²⁺ and Cu²⁺ than the other ions. For NFL-S, the removal efficiency for lower concentrations of metal ions was evaluated from a starting concentration of $\widetilde{1}$0 ppm to a ppb level over a 10 h contact time, resulting in a nearly 100% removal efficiency. Meanwhile, NFL also displays >95% and $\widetilde{8}$5% removal rates for Pb²⁺ and Cu²⁺, respectively, under the same conditions. However, in sharp contrast to NFL-S, the adsorptive capacity of NFL was decreased with rising concentrations of metal ions. The hierarchical microstructure could improve the removal efficiency of NFL-S for Pb²⁺ from 39.98%-77.86% and Cu²⁺ from 42.98%-77.22% (Table 2) in comparison with the lamellar structure of NFL. This could be attributed to the strong binding capacity of the intercalated sulfurated functional groups and structural differences [38,39]. Typically, once the k_d values are on the order of $\widetilde{1}$0⁴-10⁵ mL/g, the material can be considered an exceptional adsorbent. The difference of k_d for NFL-S and NFL further implies the good selectivity of the former for separating these ions.

Batch adsorption experiments were used to examine the effect of various competing ions present along with the target ion, where 2- and 10-times the amount of SO₄^2-, NO₃-, and Cl- anions were added to the 10 ppm Pb²⁺ aqueous solution before adsorption. As depicted in Fig. S3 in Supporting Information, the Pb²⁺ removal efficiency of NFL-S decreased with increasing concentrations of competing ions, which caused a slight drop from 87.4% to 83.2% for SO₄^2- and from 88.3% to 85.6% for Cl-. For the competing ions of NO₃-, the removal efficiency of NFL-S changed from 94.0% to 93.7%, without a noticeable trend even for an increased concentration of competing ions. Overall, the results revealed that the as-prepared NFL-S exhibits a high binding affinity for heavy metal ions, which is attributed to the exceptional selectivity of intercalated sulfide ions that are soft bases for soft or relatively soft metal ions.

The aqueous solution containing the five mixed ions was evaluated to further investigate the selective adsorption of NFL-S. Table 3 summarizes the removal efficiency of these divalent metal ions and displays the selective capture order as Pb²⁺ > Cu²⁺ ≥ Zn²⁺ > Cd²⁺ > Mn²⁺. The sorption of the heavy metal ions by NFL-S behaves in the case of the Pearson acid-base concept: the sulfide selectors are soft bases and preferentially absorb soft Lewis acids, such as Pb²⁺ and Cu²⁺, which endows the preferential adsorption of the material [12]. The k_d^Pb and k_d^Cu are much higher than k_d^Zn, k_d^Cd and k_d^Mn, reflecting the high preference for Pb²⁺/Cu²⁺ over Zn²⁺, Cd²⁺ and Mn²⁺. Compared to the single ion solution, NFL-S still exhibits excellent removal capacities in the mixed ion experiment, with increased k_d values for Pb²⁺ and Cu²⁺ but not for Zn²⁺. Meanwhile, the adsorption of Cd²⁺ and Mn²⁺ by NFL-S decreases dramatically in the single ion experiment. The high k_d values of NFL-S for Pb²⁺ ($\widetilde{1}$0⁶ mL/g) and Cu²⁺ ($\widetilde{1}$0⁵ mL/g) demonstrate its rapid decontamination ability. This indicates the good selectivity of NFL-S for heavy metal ions and implies its great potential as a highly effective heavy metal filter for water decontamination.

3.3. Relative selectivity for Pb²⁺, Cu²⁺, and Zn²⁺

The three metal ions, which include Pb²⁺, Cu²⁺ and Zn²⁺, exhibit effective adsorption capacities according to previous results, and to determine the profile selectivity of NFL-S for these ions, further study was conducted. The solution containing the three ions, Pb²⁺, Cu²⁺ and Zn²⁺, at concentrations of $\widetilde{1}$0 ppm was in contact with different amounts of NFL-S (1.0 and 2.0 mg/mL), and the corresponding results are summarized in Table 4. To explain the results more clearly, the separation factor (SF_A/B), which is defined as k_d^A/k_d^B, was used to determine the ability of the material to separate certain ions from the mixture. Generally, values of SF_A/B >100 are considered good separation factors to separate A from B [40]. When using 1.0 mg/mL NFL-S, the SF_Pb/Zn and SF_Cu/Zn values are approximately 10⁴ and 10², respectively, which demonstrates an obviously higher affinity of NFL-S for Pb²⁺ and Cu²⁺ than for Zn²⁺. Moreover, when using 2.0 mg/mL of the adsorbent, Pb²⁺ was almost completely removed, and the adsorption capacity for Cu²⁺ was improved, which further demonstrates the good capture ability of NFL-S. By decreasing the adsorbent concentration, an increased SF_Pb/Cu value from 33 to 78 suggests a higher selectivity for Pb²⁺ than Cu²⁺, and a small quantity of NFL-S is sufficient to adsorb only one of the two ions.

3.4. Sorption kinetic and isotherm toward Pb²⁺ and Cu²⁺

Highlighting a key role of the intercalated sulfides in the selective adsorption process, the adsorption performance for lead and copper was further studied. The adsorption kinetics was studied to investigate the rate of anion adsorption for the NFL-S hierarchical microstructure. Here, the amount of Pb²⁺ and Cu²⁺ adsorbed by the microcomposite adsorbents was calculated for an initial concentration of $\widetilde{1}$00 ppm at room temperature. As shown in Fig. 3(a-c), the NFL-S adsorbents can remove anions at a concentration less than 50% for Cu²⁺ and less than 10% for Pb²⁺ within 30 min. The adsorption saturations of Cu²⁺ and Pb²⁺ were 65.75 and 112.75 mg/g, respectively. The calculated parameters arising from the different models are summarized in Table S1 of Supporting Information. The pseudo-second-order model (R² = 0.999 for Pb²⁺, R² = 0.999 for Cu²⁺) is preferable to describe the adsorption behavior of the microcomposite adsorbent for anions, according to the comparison of the correlation coefficients (Fig. 3(d)). This model assumes that the rate-limiting step of the adsorption is the chemisorption between metal ions and the binding sites of the adsorbent [41,42]. The effect of pH on heavy metal adsorption (shown in Fig. S4 in Supporting Information, where Pb²⁺ is the model ion) shows that the removal efficiency of Pb²⁺ increases significantly with rising pHs, which can be attributed to the highly protonated level of the oxygen functional groups at low pH values.

View original graphic| Download| PPT slide

Fig. 3.   Adsorption kinetic curves for Cu²⁺ and Pb²⁺: (a) ion concentration change following the contact time, (b) removal efficiency as a function of the contact time, (c) sorption capacity (q_t) with contact time, (d) pseudo-second-order kinetic plots for the ion sorption, (e) Langmuir equilibrium isotherm curves of NFL-S for Pb²⁺ and Cu²⁺, and (f) C_e/q_e plotted against the equilibrium concentration (C_e).

The adsorption capacity of NFL-S for Pb²⁺ and Cu²⁺ was evaluated for different initial ion concentrations. The Langmuir and Freundlich isotherm models, used as empirical equations, were used to analyze the experiment data and describe the surface properties and affinity of the adsorbents. As shown in Fig. 3(e), the adsorption capacity increases by increasing the initial concentration of Pb²⁺ and Cu²⁺ at the same temperature. Table S2 of Supporting Information summarizes the fitting results based on the two isotherm models. Because the correlation coefficient (R²) was near 1.0, it can be concluded that the adsorption of Pb²⁺ and Cu²⁺ on the NFL-S adsorbent could be better estimated by the Langmuir model, which implies that the adsorption of ions onto this hierarchical microcomposite occurred via monolayer adsorption [43,44]. The maximum adsorption capacities calculated from the Langmuir isotherm model are 325.73 mg/g for Pb²⁺ and 284.78 mg/g for Cu²⁺, which shows a much higher removal ability for the NFL-S adsorbent than for LDH-based and other reported nanoadsorbents (Table S3 in Supporting Information); this demonstrates the excellent adsorption performance of the hierarchical microcomposite adsorbents for heavy metal ions.

3.5. Structure and morphology characterization after metal ion adsorption

The NFL-S solid samples after adsorption were centrifuged and then dried to conduct XRD, FT-IR and SEM characterizations. The XRD patterns of the spent NFL-S samples (Fig. 4(b)) show no obvious changes, and the characteristic lattice planes of (006), (100), (012), (018), and (110) indicate that the layered crystal structures of the composites were preserved after adsorption [26,27]. However, there is still difference in the adsorption of various ions at concentrations of 10 ppm. The decreased d_basal compared with the original d_basal of 0.75 nm demonstrates the coordination between the adsorbed ions and the sulfide ions, thus leading to the formation of [M(MoS₄)₂]^2- anionic complexes in the interlayer space of NFL-S [38]. This suggests that the metal ions are inserted and bind in the interlayer space. Further explanation was confirmed by the XPS spectrum of the original and used NFL-S. As shown in Fig. 4(c), no observable chemical shift in either the O 1 s and S 2p spectra can be observed before and after adsorption of the metal ions, which implies that chemisorption has occurred and effectively eliminates the possibility of chelation and the formation of a bridging bidentate bond [[45], [46], [47], [48]]. Moreover, the intensity of the O 1 s and S 2p peaks shows a conspicuous decrease. This indicates that the surface chemical process between M²⁺ (M = Pb, Cu) and NFL-S involves OH-M(II), O-M(II), SO₄-M(II) and MoS₄-M(II) surface complexation. Comparatively, the Pb binding energies of purified Pb(NO₃)₂ are located at 144.5 eV for Pb 4f_5/2 and 139.6 eV for Pb 4f_7/2 [38]. After the Pb²⁺ adsorption, a remarkable shift to a higher binding energy can be observed in both Pb 4f_5/2 and Pb 4f_7/2 (Fig. S5(a) in Supporting Information). Moreover, similar shifts of 0.37 eV and 7.17 eV for Cu 2p_1/2 and Cu 2p_3/2, respectively, are detected in the NFL-S-Cu sample compared to the Cu²⁺ species (Fig. S5(b) in Supporting Information) [38]. This indicates the formation of strong interactions between these two ions and NFL-S. The FT-IR spectra were also used to analyze the adsorption mechanism and the stability of NFL-S. As shown in Fig. 5, a band at 1397 cm^-1 appeared for all solid samples after the adsorption is associated with the NO₃- anions, which is attributed to the charge balance in the adsorbent interlayer [16]. The total stability of the NFL layer undergoing the adsorption process could also be reflected by the unchanged v (M-O) vibrations at 620 cm^-1 and δ (O-M-O) modes at 490 cm^-1 [8]. Moreover, the intensified bands located at 1242 and 1022 cm^-1 in the spectra of NFL-S-M(II) are observed after the removal of the divalent metal ions and may be attributed to the formation of Sur-OH-M(II) or Sur-O-M(II) complexes during the adsorption process [10]. Elemental distribution mappings (Fig. S6(a) in Supporting Information) of NFL-S-Pb show the presence and uniform distribution of a significant amount of Pb²⁺ in the samples. SEM images in Fig. S6(b) in Supporting Information also indicate that the resulting NFL-S samples maintain 3D microstructures after heavy metal adsorption. Moreover, residual amounts of Ni²⁺ and Mo⁶⁺ at different pH values after adsorption were also measured, which shows that the concentrations of Ni²⁺ and Mo⁶⁺ are lower than the drinking water standard recommended by WHO (Fig. S7 in Supporting Information); this reflects the stability of NFL-S to undergo the adsorption process and further implies its suitability for water purification. In addition, a consecutive adsorption-desorption experiment was conducted to evaluate the practical reusability of the material; in the experiment, 0.2 M thiourea was dissolved in a 0.02 M HNO₃ solution to release the adsorbed metal ions, thus regenerating the material. The results show that NFL-S can be reused 3 times, with a slight fading occurring for concentrations over 95%, as mentioned above (Fig. S8 in Supporting Information).

View original graphic| Download| PPT slide

Fig. 4.   (a) Reaction scheme of NFL-S and the binding modes of NFL-S with M²⁺ in the LDH gallery. (b) XRD patterns of NFL-S before and after the adsorption of different heavy metal and mixed ions at concentrations of 10 ppm. The characteristic diffraction peaks from left to right represent the lattice planes of (003), (006), (100), (012), and (110), respectively. (c) O 1 s and S 2p of NFL-S before and after the adsorption of heavy metal ions. 1, 2 and 3 correspond to the binding energies of S 2p for S^2-, S^-, and S⁶⁺.

View original graphic| Download| PPT slide

Fig. 5.   FT-IR spectra of the samples obtained after NFL-S adsorbed the 10 ppm Cu²⁺, Cd²⁺, Zn²⁺, Pb²⁺, and Mn²⁺.

3.6. Real food samples application

To assess the NFL-S adsorption effect on removing Pb²⁺ and Cu²⁺, the polluted drinking samples were individually treated with hierarchical microcomposites at a dose of 1 g/L. As shown in Table S4 of Supporting Information, the adsorption by NFL-S at 1 g/L could reduce the heavy metal ion concentrations in each of the drinking samples with a high removal efficiency, which suggests that the as-prepared NFL-S adsorbent can possibly be applied for treating these polluted drinks.

4. Conclusion

In this work, a 3D flower-like structure with sulfide functional groups intercalated LDH hierarchical microcomposite sorbent was prepared for selective heavy metal removal, which displays an effective selective order toward various divalent metal ions (Pb²⁺ > Cu²⁺ ≥ Zn²⁺ > Cd²⁺> Mn²⁺) and has high k_d values for Pb²⁺ ($\widetilde{1}$0⁶ mL/g) and Cu²⁺ ($\widetilde{1}$0⁵ mL/g). The kinetic study reveals that chemisorption is the rate-limiting step, and the isotherm study implies monolayer adsorption between the adsorbent and adsorbate. The surface chemical process between NFL-S and M²⁺ (M = Pb, Cu) was interpreted through multiple characterization techniques, and the process was found to involve the formation of surface complexes, including OH-M(II), O-M(II), SO₄-M(II) and MoS₄-M(II). The as-prepared hierarchical NFL-S microcomposite was also evaluated for the removal of Pb²⁺ in different aqueous samples, which demonstrates its great potential for selective heavy metal removal and offers intriguing opportunities for water safety control.

Acknowledgements

This research was financially supported by the National Natural Science Foundation of China (No. 21675127) and the Shaanxi Provincial Science Fund for Distinguished Young Scholars (No. 2018JC-011).

The authors have declared that no competing interests exist.