Cadmium (Cd(II)) is one of the toxic heavy metals from industrial activity such as nickel-cadmium batteries, mining, paint pigmentations, alloy, and electroplating^[1,2]. With the input flux increasing, the Cd pollution is posing a great pressure on environmental safety^[3]. Especially for the human public health, Cd(II) is very nasty due to its long biological half-life, non-biodegradable nature, and carcinogenicity^[4,5]. Then appropriate removal technologies become necessary. Adsorption is considered as one of the promising techniques owing to its low cost, reliability, simplicity and high efficiency^[6,7,8,9]. Since the adsorption reaction, mobility and species of metal ions in the environment determined their fate^{[10,11,12,13,14,15,16]}, it is of great importance to explore the interaction mechanism and microstructure of heavy metals at the solid/water interface in various environmental conditions.

Molybdenum disulfide (MoS₂), a “rising star” material, has attracted tremendous interest in remediation of heavy metals from water over past decades^{[17,18,19,20]}. MoS₂ nanosheets have layered structure with strong in-plane bonding in layers and weak out-of-plane van der Waals interactions between the individual sandwiched S-Mo-S layers^[21,22,23]. Based on the anisotropic structure of MoS₂, two-dimensional (2D) structure with large surface area and permeable channels for ion adsorption and transportation is prone to form^[23]. It is noted that both surfaces of each layer of MoS₂ are fully occupied by S atoms, which may serve as binding sites ^[19,20,21]. In addition, MoS₂ offers some excellent properties like superior mechanical flexibility, excellent chemical and thermal stability. All above properties endow MoS₂ nanosheets with good adsorption performance for the removal of heavy metals. In several cases, the maximum sorption capacities of MoS₂ nanosheets for Hg(II), Ag(I), U(VI), Cd(II), Th(IV) are up to 2506, 1348, 493, 477, 454 mg/g, respectively^[21,24-26].

Many mechanisms have been proposed for the interaction of metals with MoS₂, including ion exchange, inner layer metal-S complexation, outer layer electrostatic attraction, surface precipitation^{[25,26,27,28]}. Extended X-ray adsorption fine structure (EXAFS) spectroscopy provides an effective way to distinguish the interaction mechanism and microstructure^[29]. In the past, the EXAFS technology was widely applied to reveal the interaction mechanisms and microstructures of heavy metals and adsorbents, such as Eu(III) and γ-MnOOH, Eu(III) and δ-MnO₂, Cd(II) and kaolinite, and Cd(II) and HAs^{[29,30,31,32,33]}. However, few studies^[34,35,36] about the interaction mechanism of heavy metals and MoS₂ nanosheets have been conducted at the molecular level.

In present work, both batch experiment and EXAFS technique were employed to analyze the sorption behavior and mechanism of Cd(II) on MoS₂. Firstly, MoS₂ were characterized by SEM, TEM, XRD, FT-IR, etc. Secondly, effects of solution pH, ionic strength, and temperature, contact time on the sorption of Cd(II) and the adsorption kinetics, isotherms, thermodynamics in solutions with various pH were evaluated. Finally, EXAFS technique was used to analyze the sorption mechanisms of Cd(II) on MoS₂ as a function of pH.

Overall chemical reagents including cadmium nitrate (Cd(NO₃)₂·2H₂O), sodium nitrate (NaNO₃), sodium hydroxide (NaOH), nitrite acid (HNO₃), and molybdenum disulfide (MoS₂) were analytical grade purchased from Nanjing XFNANO Materials Tech. Co. Ltd. (China) without further purification. Cd(II) stock solution was obtained by dissolving Cd(NO₃)₂·2H₂O in distilled water.

The intrinsic properties of MoS₂ greatly determinate its sorption capacity, thereby it is helpful to investigate its characterization for revealing the sorption mechanism. Herein, TEM, SEM, EDX analysis and elemental distribution mappings of the MoS₂ sample were carried out by using a transmission electron microscope (JEM-1011, Japan) instrument and a field emission scanning electron microscope (JSM-6360LV, Japan). XRD pattern of the sample was tested by the D8 Discover X-ray diffractometer (Bruker, Germany) with Cu K_α radiation (λ=0.1541 nm) and distinguished according to the JCPD standards. FT-IR spectrometer (NEXUS, America) was employed to evaluate surface functional groups of MoS₂ in the wavelength range of 400-4000 cm^-1. Zeta potential analyzer (Zetasizer Nano ZS, Malvern Co., UK) was used to locate the pH_pzc of the adsorbent, i.e., the suspension of MoS₂ and NaNO₃ adjusted to an appropriate pH was test.

A set of sorption experiments of Cd(II) onto MoS₂ were conducted under N₂ in polyethylene tubes. The stock suspensions of Cd(II), MoS₂, NaNO₃, and distilled water were mixed in polyethylene tubes in order to gain the desired concentrations. The pH was adjusted by adding 0.1 or 0.01 mol/L HNO₃ or NaOH solution with negligible volumes. The tubes containing above mentioned mixtures were shaken for more than 12 h to achieve sorption equilibrium, and the the solid was separated by centrifugation method. Finally, the Cd concentration in the supernatant was tested. The Cd(II) sorption percentage, distribution coefficient (K_d), and sorption amount (q_e) on MoS₂ were calculated based on the following equations:

where C₀ and C_e represent the initial and equilibrium concentrations of Cd(II) (mg/L), respectively. V is the suspension volume (L) and m is the MoS₂ mass (g).

EXAFS data were collected at room temperature on BL14W at Shanghai Synchrotron Radiation Facility (SSRF, China). The obtained EXAFS data was analyzed by using Athnea software. The raw, averaged data were processed to isolate the EXAFS oscillations by removal of the pre-edge background. The k³-weighted EXAFS spectra of Cd(II) were Fourier transformed (FT). The code FEEF7 and the as-known crystal structure of Cd(NO₃)₂, Cd(OH)₂, CdS were used to calculate the theoretical scattering phases and amplitudes. The bond distance (R) and coordination number (CN), and the Debye-Waller factor of sample were optimized for each single peak by performing curve fitting with nonlinear least-squares.

The morphologies and microstructures of MoS₂ were characterized by SEM, TEM, etc., and the results are shown in Fig. S1 in detail. SEM images, EDX spectra, and elemental distribution mapping of MoS₂ are shown in Fig. 1. It is evident that two non-target elements are present in MoS₂ nanosheets, i.e., Cu and Zn. However, the elemental distribution percentages of Mo and S atoms are much higher than those of Cu and Zn, indicating that MoS₂ contains negligible impurities. In other words, the effects of Cu and Zn in MoS₂ materials on sorption can be ignored.

Fig. 2 shows the pH dependence of Cd(II) sorption on MoS₂ in 0.1, 0.01 and 0.001 mol/L NaNO₃ solutions, respectively. It is observed that the sorption is strongly dependent on pH. The percentage of Cd(II) sorption sharply increased at pH 3.3-6.5 and then creeped until a plateau level at pH>6.5. At pH 3.3-6.5, owing to protonation reaction, the number of protonated sites decreased with the increase of pH, leading to stronger affinity between the ions and MoS₂. However, the saturated surface of MoS₂ immobilized more cadmium ions at pH>6.5. It is noted that the species of Cd(II) are highly dependent on the solution pH. As provided by Wang et al.^[25], the main species were Cd²⁺, CdOH⁺ at pH<8.5, whereas the main species were Cd(OH)₂ and Cd(OH)^- at pH>8.5. Therefore, both electrostatic interaction and precipitation might be the main mechanisms of the sorption.Ionic strength is another important factor influencing sorption. Fig. 2 shows that the effect of ionic strength on Cd(II) sorption is not significant, implying inner-sphere complexation dominates the sorption behavior at different pH. The complexation type of Cd(II) with MoS₂ is different from that with other materials, such as montmorillonite^[37], kaolinite^[31], manganese oxide^[38], fibre fruit lufa^[39], magnetic polyvinyl alcohol/ laponite^[40]. The interactions between Cd(II) and these materials were found to contain both outer-sphere complexation at lower pH and inner-sphere complexation at higher pH. The difference may be attributed to various microstructure and intrinsic properties among these adsorbents.

To understand the sorption kinetics and to determine their phenomenological coefficients, the sorption capacity as a function of contact time was shown in Fig. 3(a). The sorption rates at different pH were similar and the contact time needed for complete adsorption of Cd(II) on MoS₂ was 2 h. The final sorption capacities at pH 4.55, 5.34, and 6.12 reached 28.4, 37.2 and 49.3 mg/g, respectively. These results illustrate that pH only enhances the sorption capacity but does not promote the sorption rate. Besides, 3 kinetic models were employed to analyze the kinetic data, including pseudo-first- kinetic model^[41], pseudo-second-kinetic model^[42], and intra-particle diffusion model^[43], which was described by follows:

where q_t (mg/g) represents the sorption capacity of Cd(II) at time t (h); k₁ (h^-1), k₂ (g/(mg·h)), k_i (g/(mg·h^1/2)) are the rate constants of 3 models, respectively; C (mg/L) indicates the thickness of boundary layer. The linear plots of 3 models went with Fig. 3(b-d), respectively. The corresponding parameters can be found in Table 1. We can see that the sorption kinetics could be fitted better by the pseudo-second-order kinetic model than pseudo-first-order kinetic model due to the higher correlation coefficients (R²). The result implied the chemical sorption process^[44]. More remarkable, all the sorption processes on MoS₂ at different pH contained 3 stages (Fig. 3(d)). The initial stage might be ascribed to Cd(II) ions diffusion from solution to the surface of MoS₂. Due to enough binding sites, high concentration Cd(II) ions and strong affinity between metal ions and S atoms^[24,25], great numbers of Cd(II) were rapidly adsorbed onto the MoS₂ surface. The second stage is the intra-particle diffusion, which donated as Cd(II) intraparticle diffusion through 2D structure of MoS₂ with permeable channels^[23]. The third stage on MoS₂ was the final equilibrium stage, which might be attributed to lack of binding sites or low concentration Cd(II) ions. Interestingly, the turning points in time at three stages were the same at solution pH 4.55, 5.34, 6.12. The intra-particle diffusion process mainly occurred after 1 h and the sorption reached equilibrium after 2 h. The sorption isotherms studies were carried out at different solution pH (i.e., 4.55, 5.34, 6.12) and 3 temperatures (i.e., 293, 313, 333 K), and the detailed results are shown in Fig. S2, Fig. S3 and Table S1, Table 2. The sorption isotherms were consistent with previous studies^{[45,46,47,48]}. The thermodynamics implied that the sorption was a spontaneous, endothermic process^[49,50,51].

To reveal the interaction mechanisms between Cd(II) and MoS₂ at microscopic level, local structures of Cd(II) adsorbed on MoS₂ were investigated at various pH by EXAFS. The k³-weighted, normalized, background-subtracted EXAFS spectra $\sqrt{\chi -function}$and the corresponding Fourier transformed radial structure functions (RSFs) magnitudes and imaginary parts of Cd(II) reference samples at 3 pH are shown in Fig. S4. The EXAFS oscillation of the Cd(NO₃)₂ (aq) was a single sinusoidal waveform arising from the backscattering of oxygen atoms in the first shell. The RSFs present similar characteristic and possess only one peak at ~0.2 nm. The existence of Cd-O bond and absence of direct binding for Cd(II) to MoS₂ surface imply a typical outer-sphere complexation of Cd(II)^[36]. Cd(II) in Cd(NO₃)₂ (aq) was coordinated with 6.2 O atoms at interatom bond distance of 0.233 nm (Table S3), which differs from the study by Gräfe et al.^[52]. Gräfe et al. found Cd was coordinated by 5.3 O atoms at the distance of 0.227 nm. The difference may attribute to the error of the EXAFS and XAFS methods. Unlike Cd(NO₃)₂ solution, the spectra and RSFs of Cd(OH)₂ precipitation were more complicated due to the presence of higher-shell atoms in the coordination environment of Cd(II). The first Cd-O distance was 0.238 nm coordinated with 6.1 O atoms and the second shell was simulated only with Cd-Cd contribution (R=0.359 nm, CN=5.9), which was close to the results from Gräfe et al. (Cd-O: R=0.229 nm, CN=6.7; Cd-Cd: R=0.353 nm, CN=5.6)^[52]. The EXAFS spectra of the CdS compound show that Cd(II) can be directly bonded to MoS₂ surface by S atoms, hinting that an inner-sphere complexation occurs between Cd and S^[53]. The single peak (0.24 nm) of RSFs shows the single shell of Cd-S (R=0.259 nm, CN=4.1). Fig. 4 shows that the EXAFS spectra and relevant RSFs of the sorption samples at different pH. Similarity of the spectra of the sorption samples at pH 3.56 and 6.48 to that of the CdS compound suggests that the microstructure of Cd adsorbed onto MoS₂ at low pH is similar to that of CdS complex and an inner-sphere complex forms at low pH. This inner-sphere complex is mainly present in 2D structure with large surface area and permeable channels of MoS₂ nanosheets. In addition, the spectrum at pH 9.57 is identical to that of Cd(OH)₂ precipitation. The peak of RSFs in second shell goes with Cd-Cd bonding, indicating the formation of Cd(OH)₂ precipitation at high pH. Totally, we found two sorption types: the inner-sphere complexation and the surface precipitation. The inner- sphere complexation mainly forms at lower pH which is similar to the microstructure of CdS compound. Thus Cd(II) could be directly adsorbed onto the surface of MoS₂ and then bind with S atoms(Fig. S5). The affinity of Cd-S bond is considered to be stronger than Cd-O bond, which determines the stability and excellent sorption capacity of MoS₂ as an adsorbent ^[21]. At higher pH, the Cd(OH)₂ precipitation is the main species, leading to immobility of Cd(II) onto MoS₂ surface. These interaction mechanisms and microstructure of Cd(II) and MoS₂ could explain macroscopic performance of the sorption at various pH and ionic strength.

In this work, the interaction mechanisms and microscopic structure of Cd(II) and MoS₂ nanosheets were investigated by batch experiments and EXAFS technology. In batch experiments, the pH-dependent and ionic strength- independent sorption of Cd(II) onto MoS₂ implied an inner-sphere complexation in the range of pH 3.3-9.6. The better fitted pseudo-second-order kinetics confirmed the chemical nature of the sorption and the intra-particle diffusion model reflected the sorption process from the surface to intra-particle diffusion to final equilibrium. The isotherms could be simulated better by Freundlich isotherms than by Langmuir isotherms model, indicating the heterogeneity of active sites on MoS₂. The thermodynamics implied the sorption at each pH was a spontaneous, endothermic, and irreversible process. The EXAFS spectra revealed the coexistence of two sorption types. The inner-sphere complexation was formed in the form of Cd-S bond similar to CdS complex at pH 3.56 and 6.48, while the precipitation occurred in the form of Cd-O and Cd-Cd bonds similar to Cd(OH)₂ at pH 9.57. In summary, the interaction mechanisms and local structure of Cd(II) and MoS₂ are strongly affected by the solution pH.

Supporting materials

Supporting materials related to this article can be found at https://doi.org/10.15541/jim20190381.

Microscopic Insights into pH-dependent Adsorption of Cd(II) on Molybdenum Disulfide Nanosheets

DONG Lijia¹, GUO Xiaojie², LI Xue¹, CHEN Chaogui¹, JIN Yang¹, AHMED Alsaedi³, TASAWAR Hayat^3,4, ZHAO Qingzhou⁵, SHENG Guodong⁶

(1. School of Life Science, Shaoxing University, Shaoxing 312000, China; 2. College of Materials & Environmental Engineering, Hangzhou Dianzi University, Hangzhou 310018, China; 3. NAAM Research Group, Department of Mathematics, Faculty of Science, King Abdulaziz University, Jeddah 21589, Saudi Arabia; 4. Department of Mathematics, Quaid-I-Azam University, Islamabad 44000, Pakistan; 5. College of Resources and Environment, University of Chinese Academy of Sciences, Beijing 100049, China; 6. College of Chemistry and Chemical Engineering, Shaoxing University, Shaoxing 312000, China)

Characterization of MoS₂ nanosheets

The morphologies and microstructures of MoS₂ nanosheets were characterized by SEM, TEM, XRD, FT-IR, Zeta potentials, AFM, and EDX. Both SEM and TEM images (Fig. S1(a,b)) show the obvious 2D layered structure of MoS₂ and its typical diameter ranges from hundreds of nanometers to several micrometers, implying ultrathin nature^[1,2,3]. The peak number of XRD (Fig. S1(c)) implies relatively high purity of MoS₂. The visible dominant peak (2θ=~17°) confirms the nature of MoS₂ nanosheets and the well-stacked crystalline structure^[4]. The FT-IR spectrum (Fig. S1(d)) indicates functional groups on MoS₂. The peak at ~3550 cm^-1 arises from the stretching vibration of -OH groups. The peaks at ~1700 and ~1400 cm^-1 match with the C=O stretching vibration and O-H deformation vibration, respectively^[5]. The peak at ~600 cm^-1 is attributed to the bend vibration of MoS₂ water. The Zeta potential (Fig. S1(e)) suggests that the zero charge (pH_pzc) point of MoS₂ samples be ~3.93. AFM image illustrates that most MoS₂ nanosheets are thin. The height (~2.0 nm) of the cross-section implies the presence of a few layers of MoS₂ nanosheets (Fig. S1(f)).

Sorption isotherms and thermodynamics

The sorption isotherms studies were carried out at different pH (i.e., 4.55, 5.34, 6.12) and 3 temperatures (i.e., 293, 313, 333 K) (Fig. S2). In Fig. S2(a), the sorption amount of Cd(II) on MoS₂ was promoted by high temperature at each solution pH and also by high pH at each temperature, implying that pH had no impact on temperature effects. All equilibrium data were modeled using the linear forms of Langmuir and Freundlich isotherms by the following equations^[6]:

where q_max (mg/g) and K_L (L/mg) are the maximum sorption capacity of Cd(II) on per weight unit of MoS₂ and the Langmuir affinity parameter, respectively; n and K_F (mg^1-n·Lⁿ/g) represent the exponent and Freundlich affinity-capacity parameter, respectively. Combined the linear plots of fitted isotherms (Fig. S2(b,c)) with the corresponding isotherm parameters (Table S2), it was concluded that the sorption processes at each pH were more appropriately fitted by Freundlich model than Langmuir model due to higher R² values, which was consistent with previous studies^[7,8,9]. This result could be attributed to the heterogeneity distribution of active sites on MoS₂ surface determined by the anisotropic structure of MoS₂. Thereby, the sorption behavior of Cd(II) on MoS₂ was greatly determined by the special structure of interior sulfur atoms. The thermodynamic parameters, the Gibbs free energy (∆G^θ), the enthalpy (∆H^θ), and the entropy (∆S^θ), were calculated as follows:

where K^θ is the distribution coefficient, and the values of lnK^θ at each pH are assessed by plotting lnK_dvs. C_e (Fig. S3(a)). The ∆S^θ at each pH is calculated from the slope of linear plot of ∆G^θ vs. T (Fig. S3(b)). The correlated thermodynamic parameters are recorded in Table S3. It can be inferred that Cd(II) sorption on MoS₂ at each solution pH was a spontaneous, endothermic, and irreversible process due to ∆G^θ<0, ∆H^θ>0, and ∆S^θ>0^[10]. More negative values of ∆G^θ proved the enhancement of the high temperature on the sorption. In addition, the sizes of ∆G^θ and ∆H^θ revealed chemical nature and existence of physical forces^[11,12].

References:

[1]WANG Q, YANG L, JIA F, et al. Removal of Cd(II) from water by using nano-scale molybdenum disulphide sheets as adsorbents. J. Mol. Liq., 2018, 263: 526-533.

[2]COLEMAN J N, LOTYA M, O’NEILL A, et al. Two dimensional nanosheets produced by liquid exfoliation of layered materials. Science, 2011, 331: 568-571.

[3]SPLENDIANI A, SUN L, ZHANG Y, et al. Emerging photoluminescence in monolayer MoS₂. Nano Lett., 2010, 10: 1271-1275.

[4]KUMAR A S K, JIANG S J, WARCHOL J K. Synthesis and characterization of two-dimensional transition metal dichalcogenide magnetic MoS₂@Fe₃O₄ nanoparticles for adsorption of Cr(VI)/Cr(III). ACS Omega, 2017, 2: 6187-6200.

[5]TONG S, DENG H, WANG L, et al. Multi-functional nanohybrid of ultrathin molybdenum disulfide nanosheets decorated with cerium oxide nanoparticles for preferential uptake of lead (II) ions. Chem. Eng. J., 2018, 335: 22-31.

[6]TEMKIN M J, PYZHEV V. Recent modifications to langmuir isotherms. Acta Physchim., 1940, 12: 217-222.

[7]XUE C, QI P S, LIU Y Z. Adsorption of aquatic Cd²⁺ using a combination of bacteria and modified carbon fiber. Adsorpt. Sci. Technol., 2017, 36: 857-871.

[8]GU P, ZHANG S, ZHANG C, et al. Two-dimensional MAX-derived titanate nanostructures for efficient removal of Pb(II). Dalton Trans., 2019, 48(6): 2100-2107.

[9]CHEN W, LU Z, XIAO B, et al. Enhanced removal of lead ions from aqueous solution by iron oxide nanomaterials with cobalt and nickel doping. J. Clean. Prod., 2019, 211: 1250-1258.

[10]ZHANG D, NIU H Y, ZHANG X L, et al. Strong adsorption of chlorotetracycline on magnetite nanoparticles. J. Hazard. Mater., 2011, 192: 1088-1093.

[11]ZHANG H, YU X, CHEN L, et al. Study of ⁶³Ni adsorption on NKF-6 zeolite. J. Environ. Radioact., 2010, 101: 1061-1069.

[12]BEKCI Z, SEKI Y, YURDAKOC M K. A study of equilibrium and FTIR, SEM/EDS analysis of trimethoprim adsorption onto K10. J. Mol. Struct., 2007, 827: 67-74.