Efficient removal of Pb²⁺ by Tb-MOFs: identifying the adsorption mechanism through experimental and theoretical investigations

Abstract

Nanotube-like Tb-based metal–organic frameworks (Tb-MOFs) were synthesized via the assembly of carboxylate-based ligands and Tb³⁺ under solvothermal conditions. The porous Tb-MOFs with high stability and considerable active functional groups make them ideal adsorbents in environment remediation. The factors influencing the adsorption property of Tb-MOFs toward Pb²⁺ ions were studied, comprising pH, ionic strength, adsorbent content, initial Pb²⁺ concentration and contact time. The Tb-MOFs exhibited excellent adsorption property with a maximum removal capacity of 547 mg g⁻¹ and could maintain a high adsorption performance even after five cycles. Results from batch adsorption experiments and X-ray photoelectron spectroscopy (XPS) analysis imply that the formation of the inner-sphere complex (C–/=N⋯Pb) between the nitrogenous groups of Tb-MOFs and Pb²⁺ is the primary adsorption mechanism. Furthermore, density functional theory (DFT) calculations confirm that the most favorable adsorption configuration varies with the reaction conditions and deprotonated functional groups tend to bond with Pb²⁺ at high pH. These results suggest that Tb-MOFs could be used as a promising adsorbent for the removal of Pb²⁺ ions from an aqueous environment.

---

Environmental significance

Lead is one of the world's leading environmental pollutants and has caused severe public health consequences. Reusable Tb-MOFs display highly efficient removal of Pb²⁺ from solutions. The mechanism of Pb²⁺ adsorption by Tb-MOFs was investigated by batch adsorption experiments, XPS analysis and DFT calculations. The inner-sphere complexes (C–/=N⋯Pb) formed with stable adsorption configurations contribute greatly to the adsorption. These findings illustrate that Tb-MOFs can be regarded as promising adsorbents for the removal of Pb²⁺ from aqueous solutions in environmental cleanup.

Introduction

Heavy metal ions are prevalent in contaminated water because of their mass release and their non-biodegradable property, which may lead to severe environmental problems and hazards to human health.^1,2 Lead with unquestionable toxicity has a relatively long biological half-life of 10 to 30 years in the ecosystem.^3–5 Ingestion of lead can result in various diseases, including cancer, anaemia, liver failure and nephritis, even at a very low concentration.^6,7 The maximum permissible levels for Pb²⁺ in potable water are 0.015 and 0.01 mg L⁻¹ according to the Environmental Protection Agency (EPA) and World Health Organization (WHO), respectively.⁸ Therefore, lead treatment employment of efficient techniques including chemical precipitation, electrochemical techniques, flocculation, adsorption, and membrane separation has received considerable attention in various fields.^3,8,9 Among these, adsorption is considered as one of the most effective techniques for the removal of heavy metal ions due to its high efficiency, ease of handling, and applicability even at low ion concentrations.^3,9 Moreover, the regeneration capacity of the adsorbents endows the sustainability of the technique with regards to environmental protection. Conventional adsorbent materials (i.e., natural biosorbents, activated carbons, polymer particles, nanocomposites, mesoporous materials, etc.) are limited in real applications since they are not structurally and functionally tunable and contain limited active sites. These drawbacks of the conventional adsorbents increase the need of designing easily fabricated advanced adsorbents for Pb²⁺ capture.^6,9–11

Metal–organic frameworks (MOFs) as a novel class of superior porous materials, have attracted wide attention due to their facile synthesis, large surface area, chemical stability and potential applications in purification and seperation.^10,12 The outstanding performance of MOFs in these applications depends generally on specific functional groups and controllable pore size/shape in their structures. Recently, tremendous applications have been made for MOFs for their affluent active sites, such as amidogen, alkynyl, or sulfonic acid group, etc.^10,12 However, the main disadvantages of MOF materials, which are their low chemical and mechanical stabilities compared with those of zeolites, may undoubtedly limit their use in large-scale industrial applications.^13–15 Due to its coordination ability to organic functional groups, trivalent Ln³⁺ ions are deemed to be highly promising candidates for designing MOFs.¹⁶ MOFs constructed with Tb as the metal node had higher chemical stability and could be facilely functionalized with abundant functional groups.^13,14,16 Additionally, compared with transition metal-based MOFs, the design and synthesis of innovative porous Ln-MOFs for metal ion adsorption and separation, particularly for Pb²⁺, are less explored because it is difficult to test the accessibility of the porous structures to control and predicate the overall crystal structures of Ln-MOFs.^13,15 However, investigation on Ln-MOFs, particularly Tb-MOFs, would offer an opportunity to efficiently dispose the highly-toxic heavy metal ions present in contaminated water.

Selection of organic linkers for MOFs is essential to establish the skeleton structure and physiochemical characteristics. Hard-acid Ln³⁺ ions would preferably coordinate with hard-base carboxylate donor ligands according to the hard/soft-acids/bases (HSAB) theory.^16,17 Due to the high oxophilicity of lanthanides, carboxylate ligands are believed to be the supreme linkers for building Ln-MOFs.^16,18,19 4,4′,4′′-(1,3,5-triazine-2,4,6-triyltriimino) tris-benzoic acid (H₃TATAB, shown in Fig. S1†) contains abundant N/O-containing groups. Amino, imine and carboxylic acid groups can be introduced into the Ln-MOFs, while the nitrogenous groups do not react with the constitutive metal ions. These unreacted N groups seem to be one of the most prospective ligands for heavy metal ion removal.^14,20 Furthermore, in Ln-MOFs, Ln³⁺ ions can bridge with the deprotonated carboxylic acid groups through oxygen atoms, which can enhance the stability of the MOFs in aqueous solutions and retain the active sites for adsorption.¹⁹ The nitrogenous groups of H₃TATAB have affinity for Pb²⁺ ions through complex formation or electrostatic interactions.²¹ However, due to the sophisticated adsorption mechanism and the lack of understanding of the nature of MOFs, people randomly utilize these materials as adsorbents in practice.

Taking advantage of the excellent properties of Ln-MOFs, H₃TATAB and Tb³⁺ were selected to construct nanotube-like MOFs via solvothermal reactions. The Tb-MOFs were fabricated successfully under mild conditions and examined by scanning electron microscopy (SEM), transmission electron microscopy (TEM), Brunauer–Emmett–Teller (BET), powder X-ray diffraction (XRD), Raman, and Fourier transform infrared spectroscopy (FT-IR). Specific porous structure and the abundant nitrogenous groups of the Tb-based MOFs make it suitable for removal of Pb²⁺ ions from contaminated water. Factors including pH, ionic strength, adsorbent content, initial Pb²⁺ concentration and contact time were investigated to understand the adsorption properties of Tb-MOFs toward Pb²⁺ ions. The reusability of Tb-MOFs was investigated by a simple solvent treatment for five repeated adsorption–desorption cycles. The reaction mechanism between Tb-MOFs and Pb²⁺ was clarified by a combination of batch adsorption experiments, X-ray photoelectron spectroscopy (XPS) analysis and density functional theory (DFT) calculations. The results suggest the potential application of Tb-MOFs for efficient elimination of Pb²⁺ ions from wastewater.

Experimental section

Synthesis of Tb-based MOFs

Tb-Based MOFs were successfully fabricated by using a solvothermal method in accordance with the previous research.^13,14 In brief, Tb(NO₃)₃·6H₂O (0.8 mmol), H₃TATAB (0.4 mmol), Milli-Q water (16 mL) and DMF (24 mL) were added to a Teflon-lined stainless steel autoclave and heated at 100 °C for 72 h, and then cooled to room temperature. The as-prepared Tb-based MOFs were washed with Milli-Q water and ethyl alcohol several times.

Batch adsorption experimental systems

The adsorption of Pb²⁺ ions on Tb-MOFs was conducted in 10 mL polythene centrifugal tubes. Various volumes of Tb-MOFs (1.5 g L⁻¹), Milli-Q water, NaCl and Pb²⁺ (360 mg L⁻¹) were added into the tubes to obtain the required suspensions of 0.001, 0.01 or 0.1 mol L⁻¹ ionic strength. 0.01–0.1 mol L⁻¹ NaOH or HCl aqueous solutions were employed to adjust the pH value of the mixtures. To achieve adsorption equilibrium, the experiments were performed at 298, 308 or 318 K for 24 h and then, the suspensions were separated by centrifugation (8000 rpm, 20 min). Finally, the concentrations of Pb²⁺ were detected by spectrophotometry (λ_max = 616 nm) with chlorophosphonazo III as the chromogenic reagent.

The regeneration of Tb-MOFs was performed by using 0.2 mM EDTA solution as the desorption solvent with solid/liquid 0.15 g L⁻¹. After adsorption, the Tb-MOFs were separated and added in EDTA solution under stirring for 24 h. Then, they were separated by centrifugation and washed thoroughly with Milli-Q water several times. The recycled adsorbents could be extracted and used for Pb²⁺ ion removal again.

The adsorption percentage (%), distribution coefficient (K_d) and adsorption capacity (C_s, mg g⁻¹) were calculated according to the following equations:^22–24

where C₀ and C_e (mg L⁻¹) are the initial and the equilibrium concentrations of Pb²⁺, respectively, V (mL) is the suspension volume, and m (g) represents the weight of Tb-MOFs. Experimental data are presented as the average of triplicate measurements and the relative errors were less than 5%.

Characterization

The morphology of the as-prepared Tb-MOFs was characterized by SEM (S-2500, Japan Hitachi) and TEM (JEM-2010) with EDS mapping. Powder XRD spectra were determined by a diffractometer (Philips X'Pert Pro Super X-ray) with Cu Kα source (λ = 1.54178 Å) and Mercury software (Mercury 3.6) was used to simulate the XRD curve of Tb-MOFs. The as-prepared Tb-MOFs were also investigated by FT-IR spectroscopy (Nicolet 8700, Thermo Scientific) in the range from 400 to 4000 cm⁻¹ and Raman spectroscopy (RAMANLOG 6, SPEX company, USA) at 298 K. The zeta-potential values for Tb-MOFs were recorded over a range of pH via a Nanosizer ZS instrument (Malvern Instrument Co., UK) at 298 K. The specific surface area was measured under N₂ atmosphere using a NOVA 4200e instrument (Quantachrome, FL, USA). Thermogravimetric analysis (TGA) was performed on a thermal gravimetric analyzer (Shimadzu, Kyoto, Japan) under N₂ atmosphere by heating from room temperature to 800 °C at a heating rate of 20 °C per min. The XPS measurement was recorded using ESCALAB 250 (Thermo-VG Scientific, USA).

Computational details

Calculations for all possible adsorption structures were performed in the Gaussian 09 program using the density functional theory (DFT) with WB97XD functional.^4,14 The 6-31G(d, p) basis sets were adopted for C, H, O, and N atoms, while the pseudopotential basis set of Lanl2dz was used for the Pb atom. One or two TATAB ligands were selected from the single crystal structure to investigate the adsorption styles of Pb²⁺ on MOFs. In order to simulate the probable real environment of MOFs, three or six O atoms of the carboxylate terminals (C=O groups) were fixed and the other O atoms of the carboxylate (COO⁻) were saturated by H atoms.¹⁴ Considering the possible deprotonation of N atoms in Tb-MOFs, both deprotonated and undeprotonated TATAB models were used in calculations. All possible initial structures of Pb²⁺ binding on various sites of MOF were considered and optimized to local minima.⁴

Results and discussion

Characterizations of Tb-based MOFs

The morphology and nanostructure of the Tb-MOFs were characterized by SEM and TEM (Fig. 1). The Tb-MOFs display a uniform nanotube morphology with smooth and tidy surface (Fig. 1a–c). Fig. 1d–f clearly show that the Tb-MOFs have nanotube structures with multilayered walls. Closer examination reveals that the average diameter of the nanotubes is 25 nm and the length of nanotubes varies approximately between 200 and 1000 nm. EDS mapping was conducted to study the distribution of C, N, O and Tb elements (Fig. S2†). The EDS maps of Tb and N (Fig. 1h and i) have profiles similar to the TEM image (Fig. 1g), and the spots were homogeneously distributed.

Fig. 1 SEM (a–c) and TEM images (d–g) of Tb-MOFs. EDS mapping (h and i) of Tb-MOFs.

Fig. 2a shows the XRD patterns of the as-prepared Tb-MOFs and the simulated Tb-MOFs curve using Mercury software.^13,14 The Tb-MOFs belong to the monoclinic space group (P2₁/c) and exhibited a nanotube-like 3D framework. The patterns of Tb-MOFs agree well with the simulated curve, which indicates that Tb-MOF nanotubes with crystalline structure have been successfully synthesized. The structure of Tb-MOFs can be described as symmetry-related dinuclear Tb carboxylate clusters with Tb as the metal node and deprotonated H₃TATAB as the ligands.^13,14 The Tb-carboxylate chains with dinuclear clusters connected to carboxylate groups are shown in Fig. 2b. The dinuclear Tb cluster consists of two 8-coordinated Tb³⁺ ions bridged by oxygen atoms and are linked with the TATAB ligands (Fig. 2c). The XRD patterns at different pH values were similar to each other, suggesting that the Tb-MOFs are chemically stable over a broad pH range (pH 3.0–8.0).

Fig. 2 XRD patterns of the Tb-MOFs powder (experimental and simulated curves) (a); the coordinated environments of Tb³⁺ (b); and the crystal structure of Tb-MOFs (c).

The FT-IR spectrum of Tb-MOFs is presented in Fig. 3a. A series of characteristic absorption bands in the range of 400–780 cm⁻¹ can be assigned to the Tb–O lattice vibrations.^25,26 The bands at 874, 1243, 1384, 1500 and 3410 cm⁻¹ correspond to N–H, C–NH–C, C–N, C=N and –NH stretching vibrations,^21,27–29 respectively, and indicate the abundance of nitrogenous groups (such as amino and imine) on the surface of Tb-MOFs. The nitrogen-containing functional groups can offer plenty of active sites and thus contribute to the excellent adsorption performance of Tb-MOFs. The other absorption bands at 795 (C–H stretching), 1180 (C–O stretching), 1600 (aromatic ring stretching), 1660 (C=C stretching), and 2930 cm⁻¹ (C–H stretching) suggested that the integrated structure of TATAB still remains unchanged even after the formation of Tb-carboxylate chains.^30–33

Fig. 3 FT-IR spectrum (a); Raman spectrum (b); N₂ adsorption–desorption isotherm (c); pore size distribution curve (d); TAG (e) and the zeta potential (f) of Tb-MOFs.

Raman spectroscopy, a non-destructive technique, was used to characterize the structure of Tb-MOFs. As shown in Fig. 3b, the peak at 1613 cm⁻¹ is ascribed to the C=C stretching mode of the benzene ring in the TATAB linker.^34,35 The peak at 1414 cm⁻¹ is associated with the vibration of triazine and the peak at 977 cm⁻¹ is assigned to the C–N stretching mode.³⁵ The strong broad peak at 665 cm⁻¹ is attributed to the stretching vibration of Tb–O,^36,37 suggesting high affinity between the Tb and O atoms. The peak at 422 cm⁻¹ is attributed to the Tb–O stretching mode.³⁸ The peaks at 579, 736, 1155 and 1306 cm⁻¹ are attributed to the C–H stretching mode of benzene ring and N–H stretching mode of triazine.²⁹ The obtained Raman spectra is further evidence that the functional groups of TATAB were still maintained in the Tb-MOF structures.

The surface area of Tb-MOFs calculated from the N₂ adsorption–desorption isotherm was 56.72 m² g⁻¹ (Fig. 3c). Since the average pore diameter was ∼1.75 nm (Fig. 3d), the Tb-MOFs could be regarded as microporous materials. However, the surface area of Tb-MOFs is not as large as other reported MOFs,^13,14,39 which may due to the lower crystallinity and the relatively limited pore/size distribution of Tb-MOFs. In addition, the adsorption and desorption curves were not completely closed due to the slightly irregular mesoporosity of Tb-MOFs.⁴⁰ TGA measurements were performed to study the thermal stability of Tb-MOFs (Fig. 3e). For a fresh sample of Tb-MOFs, the TGA curve shows the weight loss of moisture and ethanol from 40 to 100 °C. Then, the TGA curve does not show significant changes from 100 to 450 °C, which suggests the relative stability of Tb-MOFs, the compound decomposed quickly upon further heating to 500 °C due to the carbonization of H₃TATAB. These results indicate that the Tb-MOFs have high stability in aqueous systems.

Zeta potential, an important parameter that characterizes the surface charge of Tb-MOFs, is shown in Fig. 3f. The pH_PZC (pH at the point of zero charge) value of Tb-MOFs is ∼4.2. It can be seen that the zeta potential of Tb-MOFs shifts from positive to negative with an increase in pH. This is attributed to the deprotonation of the functional groups of Tb-MOFs at high pH values. The electrostatic attraction between the negatively charged Tb-MOFs and the heavy metal ions is the driving force for the adsorption but the electrostatic repulsion can hinder the metal ions from the liquid to adsorb on the surface of the solid. Hence, according to the zeta potential values, the range of pH for the removal of cations can be selected.

Effect of pH and ion strength

The pH value of a solution plays a vital role in the adsorption process; it not only affects the species of heavy metal ions but also the surface charge and protonation/deprotonation of the adsorbents.³ The effect of pH on the adsorption of Pb²⁺ by Tb-MOFs has been investigated in the range from 3.0 to 7.0 (Fig. 4a). Pb exists in the form of Pb²⁺ at pH < 7.0 (Fig. S3†). It can be observed that the adsorption of Pb²⁺ onto Tb-MOFs was strongly pH-dependent and Pb²⁺ removal increased with the increase in pH values. At pH > pH_PZC, the positively charged cationic Pb²⁺ can be easily adsorbed on the negatively charged Tb-MOFs by electrostatic attraction. As pH increases, the deprotonation of the nitrogenous groups favors the adsorption of Pb²⁺via Pb⋯N complexation, thus considerably increasing Pb²⁺ removal. In contrast, at pH < pH_PZC, the surface of Tb-MOFs is positively charged because the nitrogenous groups are excessively protonated and the abundant H⁺ ions are prone to compete with Pb²⁺ ions, which leads to the decrease in removal percentage at low pH.³

Fig. 4 Effect of pH and ion strength on the adsorption of Pb²⁺ by Tb-MOFs at m/V = 0.15 g L⁻¹ (a). Effect of Tb-MOF content on Pb²⁺adsorption (b). [Pb²⁺] = 90 mg L⁻¹, pH = 5.5.

In order to verify the practical usability, the interaction between Tb-MOFs and Pb²⁺ was investigated in solutions with different ionic strengths. As shown in Fig. 4a, the adsorption of Pb²⁺ on Tb-MOFs is weakly dependent on the ionic strength. It has been reported that the outer-sphere complexation depends on the ionic strength, whereas the inner-sphere complexation exhibits outstanding tolerance towards ionic strength changes because of the strong chemical bonding between the heavy metal ions and the active sites.²¹ The weak interference of ionic strength on Pb²⁺ adsorption suggests that adsorption mainly depends on the inner-sphere complexation (nitrogenous groups with lone pair of electrons binding to Pb²⁺ ions) rather than the outer-sphere surface complexation.

Effect of adsorbent content

Adsorbent content is an important factor that affects the adsorption efficiency of a material for a fixed original concentration of the heavy metal ions. Fig. 4b depicts the effect of the Tb-MOFs content on Pb²⁺ removal. A meaningful relationship is that the adsorption increases quickly as the adsorbent content increased. This can be interpreted in this study as the increase in the Tb-MOFs content leads to more active sites for Pb²⁺ removal. Additionally, the distribution coefficient K_d is usually applied to evaluate the interaction ability between the adsorbent and heavy metal ions.^15,21 As shown in Fig. 4b, K_d is hardly dependent on the content of Tb-MOFs, which agrees with its physicochemical properties.^11,15 Furthermore, the values of log K_d for Tb-MOFs are above 4.0 mL g⁻¹, which indicates good affinity of Tb-MOFs for Pb²⁺ ions.^15,21 Therefore, the Tb-MOFs can be considered as an efficient adsorbent for Pb²⁺ removal from large volumes of wastewater.

Adsorption kinetics

The adsorption kinetics of Pb²⁺ ions on Tb-MOFs is shown in Fig. 5a. The removal rate increases quickly within 10 min due to the large number of available sites at the initial stage. Then, it increases slowly as the concentration of Pb²⁺ ions and adsorption sites reduce until reaching the adsorption equilibrium within 50 min. To further study the adsorption mechanism of Tb-MOFs, the pseudo-second-order model and intraparticle diffusion model were used to evaluate the experimental data.^3,23 The equations with fitting parameters are shown in Table S1.† It can be seen that the pseudo-second-order model depicts the whole adsorption process very well with high correlation coefficients (0.999). Adsorption of Pb²⁺ ions on Tb-MOFs is primarily by chemisorption involving the complexation of Pb²⁺ ions with functional groups of Tb-MOFs.¹¹ The intraparticle diffusion model is not merely the limiting step because the straight line does not pass through the origin (Fig. 5b). The intercept (A) of intraparticle diffusion, an important parameter, reflects the thickness of the boundary layer that limits the diffusion rate of ions. The value of A increased with the increase in time (Table S1†), i.e., the larger the A values, the greater would be the boundary layer effect. This corresponds to the fact that the adsorption rate decreases gradually with the increase in reaction time. These results indicate that four steps with different limiting processes are possibly involved: (I) Pb²⁺ ions rapidly reach the surface of Tb-MOFs due to the presence of abundant active sites, resulting in instantaneous or external surface adsorption; (II) adsorption slows down, indicating the rate-limiting stage for Pb²⁺ ions diffusion into the interior of the Tb-MOFs nanotube; (III) adsorption declines quickly due to the relatively low residual Pb²⁺ concentration and the limited amounts of unoccupied active sites; and (IV) Pb²⁺ adsorption reaches a quite stable equilibrium in the final step. It should be noted that intraparticle diffusion is not the sole rate-limiting step and the chemical complex formation might also be involved.⁴¹

Fig. 5 Adsorption kinetics of Pb²⁺ on Tb-MOFs. Pseudo-second-order kinetic model (a) and intraparticle diffusion model (b). [Pb²⁺] = 90 mg L⁻¹, pH = 5.5, m/V = 0.15 g L⁻¹.

Adsorption isotherms and thermodynamics

To further investigate the adsorption performance and the thermodynamic properties of Tb-MOFs, isotherm models, which can demonstrate the interactions between the Tb-MOFs and Pb²⁺ ions, were employed. The adsorption of Pb²⁺ ions increased promptly with an increase in original concentration (Fig. 6a), which is due to abundant active sites available on Tb-MOFs. Then, adsorption gradually increased and reached a plateau as the active sites were occupied completely. Additionally, the adsorption increased monotonically with the increase in temperature in the range from 298 to 313 K, indicating that the increase in the temperature is beneficial for Pb²⁺ removal by Tb-MOFs. The adsorption data could be simulated by the Langmuir and Freundlich models (Fig. 6a), which are shown as eqn (4) and (5), respectively.⁴²

C_s = K_FC_eⁿ (5)

where C_max (mg g⁻¹) is the maximum adsorption capacity, b is the constant related to the heat adsorption, and 1/n and K_F are the Freundlich constants. The fitting parameters obtained by Langmuir and Freundlich models are presented in Table 1. According to the regression coefficients (R²), the Langmuir model is more suitable for simulating the thermodynamic data than the Freundlich model, and suggests that the active sites of Tb-MOFs have equal adsorption performance and are uniformly distributed. Compared with the maximum Pb²⁺ adsorption capacity of other MOFs and adsorbents, Tb-MOFs exhibits higher adsorption capacity in spite of not having sufficiently high surface area (Table 2).

Fig. 6 Isotherms of Pb²⁺ adsorbed onto Tb-MOFs (solid line: Langmuir model; dashed line: Freundlich model) (a). Liner plot of ln K⁰versus 1/T (b).

Table 1 Parameters calculated from Langmuir and Freundlich models for Pb²⁺ adsorption onto Tb-MOFs

T (K)	Langmuir			Freundlich
	C max (mg g^−1)	b (L mg^−1)	R ^2	K F	n	R ^2
298	547	0.025	0.988	71	0.359	0.969
308	576	0.035	0.994	95	0.325	0.956
318	635	0.047	0.983	126	0.299	0.957

---

Table 2 Comparison of the maximum adsorption capacity of Pb²⁺ on Tb-MOFs with those of other adsorbents

Adsorbents	pH	C max (mg g^−1)	Ref.
MOF-derived magnetic inorganic sorbents	4.9	345	3
Magnetic aminofunctionalized aluminium MOF	—	492	4
[Ag12(MA)8(mal)6.18·H2O]n MOF	7.0	120	5
g-C3N4	5.0	65.6	9
MoS2@biochar	5.0	189	11
Zirconium-based porous MOF	7.0	73	22
Amino-functionalization of Cr-based MOFs	6.0	88	42
Bifunctional mesoporous organosilica	6.0	22	43
Tb-MOFs	5.5	547	This study

---

Temperature-dependent adsorption isotherms were plotted to calculate the thermodynamic parameters, namely, entropy change (ΔS⁰), enthalpy change (ΔH⁰) and Gibbs free energy change (ΔG⁰) as follows:¹¹

ΔG⁰ = −RT ln K⁰ (7)

where K⁰ represents the thermodynamic equilibrium constant, R is the gas constant (8.314 J mol⁻¹ K⁻¹), and T (K) is the temperature in Kelvin. ln K⁰ was be obtained by plotting ln K_dversus C_e and extrapolating the curve to 0. The thermodynamic parameters can reveal the mechanism of Pb²⁺ ion adsorption onto Tb-MOFs (Fig. 6b and Table S2†). The increase in positive ΔH⁰ with increase in temperature suggests an endothermic adsorption process. A likely explanation for the positive value of ΔH⁰ is that the Pb²⁺ cations dissolve well in water, while the hydration sheath of Pb²⁺ cations may hinder its adsorption on the Tb-MOFs surface. However, with the increase in temperature, the enhanced molecular motion may reduce the thickness of the hydration sheath. The potential barrier between Pb²⁺ ions and Tb-MOFs is reduced, thus enhancing the adsorption of Pb²⁺ ions on the Tb-MOFs. ΔG⁰ becomes more negative with the increase in temperature, which implies that the adsorption process is spontaneous and higher temperature is beneficial to enhance the adsorption performance. Furthermore, the positive value for ΔS⁰ reveals the random increase in the amount of Pb²⁺ adsorbed on the Tb-MOFs at the solid/solution interface and the high affinity between the Tb-MOFs and Pb²⁺.²³

Adsorption mechanism

XPS analysis was performed to identify the surface properties of the Tb-MOFs before and after Pb²⁺ ion adsorption. Fig. 7a shows the coexistence of Tb, N, O, and C elements in the Tb-MOF samples. After the adsorption of Pb²⁺ ions by Tb-MOFs (Tb-MOFs–Pb), new strong peaks appeared at 643.8, 437.9, 415.3, 138.6 and 19.5 eV, corresponding to Pb 4p, Pb 4f and Pb 5d. This directly proves that the Pb²⁺ ions are successfully adsorbed on Tb-MOFs.

Fig. 7 XPS survey of Tb-MOFs before and after Pb²⁺ adsorption (a); high-resolution spectra of C 1s (b), O 1s (c), and N 1s (d) present in Tb-MOFs. (e) O 1s and (f) N 1s spectra of Tb-MOFs after Pb²⁺ adsorption i.e. Tb-MOFs–Pb.

The high-resolution spectrum can be applied to differentiate and identify the elemental composition on the surface of the adsorbent, which would help elucidating the adsorption mechanism. In the high-resolution C 1s curve of Tb-MOFs (Fig. 7b), the four different peaks centered at 288.4, 285.9, 285.0, and 284.5 eV can be ascribed to O–C–O, C–N, C=N and benzene ring,^43–46 respectively, which are consistent with the FT-IR and Raman analyses. In the high-resolution O 1s spectrum of Tb-MOFs, the two peaks at 532.1 and 531.4 eV are attributed to Tb–O and C–O,^47,48 respectively (Fig. 7c). Interestingly, after Pb²⁺ ions adsorbed on the Tb-MOFs, the deconvoluted O1s spectrum of Tb-MOF–Pb shows almost unchanged Tb–O and C–O peaks (Fig. 7e), indicating that the contribution of surface oxygenic functional groups in the ion removal process may have been overlooked. This phenomenon can be attributed to the saturated coordination of Tb atoms and the hydrophobic benzene hindering the heavy metal ions in the liquid from the surface oxygenic functional groups of the Tb-MOFs.¹⁴ The N 1s band (Fig. 7d) can be divided into three peaks of N species, namely, 400.4, 399.7, and 398.7 eV, corresponding to C–N, N–H, and C=N.^49–51 It is clearly seen that the binding energies of C=N, C–N, and N–H increase after Pb²⁺ ions adsorbed on Tb-MOFs. This is because the shared bond between the N atom and Pb²⁺ ions can decrease the electron cloud density on the N atom, and hence increase the binding energy of the nitrogenous groups.⁵² Moreover, two new peaks at 399.3 and 398.3 eV can be probably attributed to amine (–N–) and imine (–N=) from the Pb–N– and Pb–N= bonding modes. Thus, the nitrogenous groups could act as active sites for coordination with Pb²⁺ ions. These results are consistent with the fact that the adsorption of Pb²⁺ ions on the Tb-MOFs is mainly ascribed to the complexation with nitrogenous groups via the inner-sphere mechanism (Pb–N– and Pb–N=).

Desorption mechanism and recycle performance

For the Pb²⁺ adsorption to be useful and economically favorable, the Tb-MOFs must have good adsorption capacity and should be regenerated. The regeneration agents were applied to recycle the spent Tb-MOFs. An effective, strong ligand and non-destructive regeneration approach was employed using the chelating agent (EDTA), which demonstrated both good ion desorption efficiency and capacity recovery due to its strong affinity to Pb²⁺ ions. Fig. 8a shows the regeneration ability of Tb-MOFs after treatment with EDTA. The Pb²⁺ adsorption capacity slightly reduced for EDTA since some active sites on Tb-MOFs occupied by Pb²⁺ ions could not be completely recovered. However, a high adsorption capacity could be maintained even after five adsorption–desorption cycles. XRD was employed to investigate the stability of the Tb-MOFs (Fig. 8b), and no significant changes could be observed in the XRD pattern, suggesting that Tb-MOFs maintains high structural stability after every cycle of regeneration. A minor amount of Tb ions was released in the solution and high adsorption rates were observed through the regeneration cycles, suggesting the stability of Tb-MOFs (Table S3†).¹⁴ Furthermore, the nanotube-like morphology of Tb-MOFs was maintained even after the fifth cycle of regeneration (Fig. S4†). Therefore, Tb-MOFs still kept a high structural stability after every cycle of regeneration. Based on practical applications, the Tb-MOFs show satisfactory adsorption performance after regeneration and can be applied potentially as an efficient adsorbent for Pb²⁺ ion removal from wastewater.

Fig. 8 Recycling of Tb-MOFs by Pb²⁺ removal using EDTA (a). XRD patterns of Tb-MOFs after each desorption cycle (b).

DFT calculation

To investigate the binding sites of adsorbed Pb²⁺ ions on the Tb-MOFs surface, DFT calculation was adopted to understand the adsorption process at the atomic level.⁵³ One advantage of Tb-MOFs is that the structure offers enough space to adsorb Pb²⁺ ions without any orientation limitation. The Pb²⁺ ions can combine with the functional groups to form the inner-sphere complexes (Pb–N– and Pb–N=). As binding energy E_b can indicate the possible interaction mechanism between the heavy metal ions and adsorbents, it was calculated as follows:⁵⁴

E_b = E_Tb–MOFs + E_Pb − E_Tb-MOFs–Pb (8)

The results are shown in the Fig. 9 and Table S4.† The positive values for E_b indicate a favorable and stable reaction. Moreover, the high values of E_b (> 1 eV) imply that the adsorption of Pb²⁺ ions on Tb-MOFs is a chemical adsorption process and is consistent with the experimental analysis (inner-sphere complex).⁵⁵Fig. 9a and b show the single ligand binding with Pb²⁺ ions. We can see that the E_b of [single(-H)-Tb-MOFs⋯Pb]⁺ is higher than that of [single-Tb-MOFs⋯Pb]²⁺, indicating that [single(-H)-Tb-MOFs⋯Pb]⁺ is more favorable. A similar situation also occurs in the double ligand adsorption process: E_b([double-Tb-MOFs⋯Pb]⁺) < E_b([double(-H)-Tb-MOFs⋯Pb]⁺) < E_b([double(-2H)-Tb-MOFs⋯Pb]). These results show that the deprotonated functional groups can further improve the adsorption performance, which is in agreement with the experimental results (the adsorption capacity increases with the increase in pH). It can be concluded that at low pH, [single-Tb-MOFs⋯Pb]²⁺ and [double-Tb-MOFs⋯Pb]⁺ are the possible forms and [double-Tb-MOFs⋯Pb]⁺ is more favorable. With the increase in pH, different forms such as [single(-H)-Tb-MOFs⋯Pb]⁺, [double(-H)-Tb-MOFs⋯Pb]⁺, and [double(-2H)-Tb-MOFs⋯Pb] may involve in the complexation and [double(-2H)-Tb-MOFs⋯Pb] is the most stable structure.

Fig. 9 The optimized geometries for Pb²⁺ ion adsorbed on Tb-MOFs: (a) the single ligand binding with Pb²⁺, (b) the single ligand binding (N1 removed a proton) with Pb²⁺, (c) double ligands binding with Pb²⁺, (d) double ligands binding (N3 removed a proton) with Pb²⁺, (e) double ligands binding (N1 and N3 removed two protons) with Pb²⁺.

Conclusions

Nanotube-like Tb-MOFs with satisfactory adsorption capacity, high structure stability and good recycle performance were successfully fabricated. The results suggest that Langmuir model is suitable for depicting the adsorption process, and Pb²⁺ adsorption on Tb-MOFs is an endothermic and spontaneous process. An adsorption capacity of 547 mg g⁻¹ was achieved by Tb-MOFs for Pb²⁺ within 50 min with structure stability. The rate-limiting step controls the entire Pb²⁺ removal process and the adsorption mechanism could be explained by the formation of an inner-sphere complex (C–/=N⋯Pb) between the nitrogenous groups and Pb²⁺. The adsorption configurations optimized by DFT suggest that the Tb-MOFs structure offers adequate space to adsorb Pb²⁺ ions and the deprotonated functional groups bond easily with Pb²⁺, forming stable adsorption complexes even at high pH. The adsorption mechanism of Tb-MOFs suggests its great potential in the removal of Pb²⁺ from contaminated water.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

Financial support from National Natural Science Foundation of China (U1607102, 21876047), Science Challenge Project (TZ2016004), the Fundamental Research Funds for the Central Universities (2018ZD11), the Foundation of Basic Research for Application of Qinghai Province (2017-ZJ-706) are acknowledged.

Notes and references

1. L. Ma, Q. Wang, S. Islam, Y. Liu, S. Ma and M. Kanatzidis, Highly selective and efficient removal of heavy metals by layered double hydroxide intercalated with the MoS₄²⁻ ion, J. Am. Chem. Soc., 2016, 138, 2858–2866.
2. J. Efome, D. Rana, T. Matsuura and C. Lan, Metal–organic frameworks supported on nanofibers to remove heavy metals, J. Mater. Chem. A, 2018, 6, 4550–4555.
3. D. Chen, W. Shen, S. Wu, C. Chen, X. Luo and L. Guo, Ion exchange induced removal of Pb(II) by MOF-derived magnetic inorganic sorbents, Nanoscale, 2016, 8, 7172–7179.
4. R. Ricco, K. Konstas, M. Styles, J. Richardson, R. Babarao, K. Suzuki, P. Scopece and P. Falcaro, Lead(II) uptake by aluminium based magnetic framework composites (MFCs) in water, J. Mater. Chem. A, 2015, 3, 19822–19831.
5. M. Salarian, A. Ghanbarpour, M. Behbahani, S. Bagheri and A. Bagheri, A metal-organic framework sustained by a nanosized Ag12 cuboctahedral node for solid-phase extraction of ultra traces of lead(II) ions, Microchim. Acta, 2014, 181, 999–1007.
6. J. Efome, D. Rana, T. Matsuura and C. Lan, Insight studies on metal-organic framework nanofibrous membrane adsorption and activation for heavy metal ions removal from aqueous solution, ACS Appl. Mater. Interfaces, 2018, 10, 18619–18629.
7. F. Rouhani and A. Morsali, Fast and selective heavy metal removal by a novel metal-organic framework designed with in-situ ligand building block fabrication bearing free nitrogen, Chem. – Eur. J., 2018, 24, 5529–5537.
8. F. Rouhani and A. Morsali, Goal-directed design of metal-organic frameworks for Hg(II) and Pb(II) adsorption from aqueous solutions: A conceptual review, Chem. – Eur. J., 2018, 24, 17170–17179.
9. R. Hu, X. Wang, S. Dai, D. Shao, T. Hayat and A. Alsaedi, Application of graphitic carbon nitride for the removal of Pb(II) and aniline from aqueous solutions, Chem. Eng. J., 2015, 260, 469–477.
10. J. Li, X. Wang, G. Zhao, C. Chen, Z. Chai, A. Alsaed, T. Hayat and X. Wang, Metal-organic framework-based materials: superior adsorbents for the capture of toxic and radioactive metal ions, Chem. Soc. Rev., 2018, 47, 2322–2356.
11. H. Zhu, X. Tan, L. Tan, C. Chen, N. S. Alharbi, T. Hayate, M. Fang and X. Wang, Biochar derived from sawdust embedded with molybdenum disulfide for highly selective removal of Pb²⁺, ACS Appl. Nano Mater., 2018, 1, 2689–2698.
12. L. Zhu, D. Sheng, C. Xu, X. Dai, M. A. Silver, J. Li, P. Li, Y. Wang, Y. Wang, L. Chen, C. Xiao, J. Chen, R. Zhou, C. Zhang, O. K. Farha, Z. Chai, T. E. Albrecht-Schmitt and S. Wang, Identifying the recognition site for selective trapping of ⁹⁹TcO₄⁻ in a hydrolytically stable and radiation resistant cation metal-organic framework, J. Am. Chem. Soc., 2017, 139, 14873–14876.
13. H. Zhang, D. Chen, H. Ma and P. Cheng, Real-time detection of traces of benzaldehyde in benzyl alcohol as a solvent by a flexible lanthanide microporous metal-organic framework, Chem. – Eur. J., 2015, 21, 15854–15859.
14. W. Liu, X. Dai, Z. Bai, Y. Wang, Z. Yang, L. Zhang, L. Xu, L. Chen, Y. Li, D. Gui, J. Diwu, J. Wang, R. Zhou, Z. Chai and S. Wang, Highly sensitive and selective uranium detection in natural water systems using a luminescent mesoporous metal-organic framework equipped with abundant lewis basic sites: A combined batch, X-ray absorption spectroscopy, and first principles simulation investigation, Environ. Sci. Technol., 2017, 51, 3911–3921.
15. C. Xiao, M. A. Silver and S. Wang, Metal-organic frameworks for radionuclide sequestration from aqueous solution: a brief overview and outlook, Dalton Trans., 2017, 46, 16381–16386.
16. J. Heine and K. M_Üller-Buschbaum, Engineering metal-based luminescence in coordination polymers and metal-organic frameworks, Chem. Soc. Rev., 2013, 42, 9232–9242.
17. X. Zhao, D. Tian, Q. Gao, H. Sun, J. Xu and X. Bu, A chiral lanthanide metal-organic framework for selective sensing of Fe(III) ions, Dalton Trans., 2016, 45, 1040–1046.
18. Z. Wang, H. Liu, S. Wang, Z. Rao and Y. Yang, A luminescent terbium-succinate MOF thin film fabricated by electrodeposition for sensing of Cu²⁺ in aqueous environment, Sens. Actuators, B, 2015, 220, 779–787.
19. S. Kaskel, The chemistry of metal-organic frameworks: Synthesis, characterization, and applications, Wiley, 2016, ISBN: 978-3-527-33874-0.
20. X. Tan, J. Hu, G. Montavon and X. Wang, Sorption speciation of nickel(II) onto Ca-Montmorillonite: Batch, EXAFS techniques and modeling, Dalton Trans., 2011, 40, 10953–10960.
21. H. Zhu, J. Wu, M. Fang, L. Tan, C. Chen, N. Alharbi, T. Hayat and X. Tan, Synthesis of a core-shell magnetic Fe₃O₄-NH₂@PmPD nanocomposite for efficient removal of Cr(VI) from aqueous media, RSC Adv., 2017, 7, 36231–36241.
22. Ş. Tokalıoğlu, E. Yavuz, S. Demir and Ş. Patat, Zirconium-based highly porous metal-organic framework (MOF-545) as an efficient adsorbent for vortex assisted-solid phase extraction of lead from cereal, beverage and water samples, Food Chem., 2017, 237, 707–715.
23. X. Tan, M. Fang, L. Tan, H. Liu, X. Ye, T. Hayat and X. Wang, Core-shell hierarchical C@Na₂Ti₃O₇·9H₂O nanostructures for the efficient removal of radionuclides, Environ. Sci.: Nano, 2018, 5, 1140–1149.
24. J. Wu, H. Zhu, G. Liu, L. Tan, X. Hu, C. Chen, N. S. Alharbi, T. Hayat and X. Tan, Fabrication of core-shell CMNP@PmPD nanocomposite for efficient As(V) adsorption and reduction, ACS Sustainable Chem. Eng., 2017, 5, 4399–4407.
25. V. Kumar, O. Ntwaeaborwa, J. Holsa, D. Motaung and H. Swart, The role of oxygen and titanium related defects on the emission of TiO₂:Tb³⁺ nano-phosphor for blue lighting applications, Opt. Mater., 2015, 46, 510–516.
26. D. Chen, E. Jordan and M. Gell, Sol-gel combustion synthesis of nanocrystalline YAG powder from metal-organic precursors, J. Am. Ceram. Soc., 2008, 91, 2759–2762.
27. S. Khan, S. Shahid, S. Kanwal and G. Hussain, Synthesis characterization and antibacterial activity of Cr (III), Co (III), Fe (II), Cu (II), Ni (III) complexes of 4-(2-(((2-hydroxy-5-nitrophenyl) diazenyl) (phenyl) methylene) hydrazinyl) benzene sulfonic acid based formazan dyes and their applications on leather, Dyes Pigm., 2018, 148, 31–43.
28. D. Kong, N. Qiao, H. Liu, J. Du, N. Wang, Z. Zhou and Z. Ren, Fast and efficient removal of copper using sandwich-like graphene oxide composite imprinted materials, Chem. Eng. J., 2017, 326, 141–150.
29. C. Müller, L. David, V. Chiş and S. Pînzaru, Detection of thiabendazole applied on citrus fruits and bananas using surface enhanced Raman scattering, Food Chem., 2014, 145, 814–820.
30. Y. Zhao, H. Chen, J. Li and C. Chen, Hierarchical MWCNTs/Fe₃O₄/PANI magnetic composite as adsorbent for methyl orange removal, J. Colloid Interface Sci., 2015, 450, 189–195.
31. Z. Jin, X. Wang, Y. Sun, Y. Ai and X. Wang, Adsorption of 4-n-nonylphenol and bisphenol-a on magnetic reduced graphene oxides: A combined experimental and theoretical studies, Environ. Sci. Technol., 2015, 49, 9168–9175.
32. J. Wang, J. Hao, D. Liu, S. Qin, C. Chen, C. Yang, Y. Liu, T. Yang, Y. Fan, Y. Chen and W. Lei, Flower stamen-like porous boron carbon nitride nanoscrolls for water cleaning, Nanoscale, 2017, 9, 9787–9791.
33. Y. Sun, X. Wang, C. Ding, W. Cheng, C. Chen, T. Hayat, A. Alsaedi, J. Hu and X. Wang, Direct synthesis of bacteria-derived carbonaceous nanofibers as a highly efficient material for radionuclides elimination, ACS Sustainable Chem. Eng., 2016, 4, 4608–4616.
34. F. Fernández-Trillo, J. Hest, J. Thies, T. Michon, R. Weberskirch and N. Cameron, Reversible immobilization onto PEG-based emulsion-templated porous polymers by co-assembly of stimuli responsive polymers, Adv. Mater., 2009, 21, 55–59.
35. Y. Chen, S. Han, X. Li, Z. Zhang and S. Ma, Why does enzyme not leach from metal-organic frameworks (MOFs)? Unveiling the interactions between an enzyme molecule and a MOF, Inorg. Chem., 2014, 53, 10006–10008.
36. C. Williams, A. Tolia, H. Chan, C. Takoudis and M. Weaver, Surface-enhanced raman spectroscopy as an in situ real-time probe of catalytic mechanisms at high gas pressures: The CO-NO reaction on platinum and palladium, J. Catal., 1996, 163, 63–76.
37. S. Kumar, A. Ojha and R. Singh, Synthesis and raman signature for the formation of CdO/MnO₂ (core/shell) nanostructures, J. Raman Spectrosc., 2014, 45, 717–722.
38. S. Liu, Y. Liu, Q. Mu, F. Zhang, H. Li and Y. Wang, Synthesis, characterization and photoluminescent properties of rare-earth hydroxides and oxides nanorods by hydrothermal route, Appl. Phys. A: Mater. Sci. Process., 2013, 111, 1229–1240.
39. N. Bobbitt, M. Mendonca, A. Howarth, T. Islamoglu, J. Hupp, O. Farha and R. Snurr, Metal-organic frameworks for the removal of toxic industrial chemicals and chemical warfare agents, Chem. Soc. Rev., 2017, 46, 3357.
40. B. Ouay, S. Kitagawa and T. Uemura, Opening of an accessible microporosity in an otherwise nonporous metal-organic framework by polymeric guests, J. Am. Chem. Soc., 2017, 139, 7886–7892.
41. T. Wang, L. Zhang, C. Li, W. Yang, T. Song, C. Tang, Y. Meng, S. Dai, H. Wang, L. Chai and J. Luo, Synthesis of core-shell magnetic Fe₃O₄@poly(m-Phenylenediamine) particles for chromium reduction and adsorption, Environ. Sci. Technol., 2015, 49, 5654–5662.
42. X. Luo, L. Ding and J. Luo, Adsorptive removal of Pb(II) ions from aqueous samples with amino-functionalization of metal-organic frameworks MIL-101(Cr), J. Chem. Eng. Data, 2015, 60, 1732–1743.
43. M. Dinker, T. Ajithkumar and P. Kulkarni, _L-Proline functionalized dicationic framework of bifunctional mesoporous organosilica for the simultaneous removal of lead and nitrate ions, ACS Sustainable Chem. Eng., 2017, 5, 4188–4196.
44. C. Ding, W. Cheng, Y. Sun and X. Wang, Novel fungus-Fe₃O₄ bio-nanocomposites as high performance adsorbents for the removal of radionuclides, J. Hazard. Mater., 2015, 295, 127–137.
45. T. Wen, Q. Fan, X. Tan, Y. Chen, C. Chen, A. Xu and X. Wang, A core-shell structure of polyaniline coated protonic titanate nanobelt composites for both Cr(VI) and humic acid removal, Polym. Chem., 2016, 7, 785–794.
46. C. Zhu, A. Soldatov and A. Mathew, Advanced microscopy and spectroscopy reveal the adsorption and clustering of Cu(II) onto TEMPO-oxidized cellulose nanofibers, Nanoscale, 2017, 9, 7419–7428.
47. J. Liang, X. Li, Z. Yu, G. Zeng, Y. Luo, L. Jiang, Z. Yang, Y. Qian and H. Wu, Amorphous MnO₂ modified biochar derived from aerobically composted swine manure for adsorption of Pb(II) and Cd(II), ACS Sustainable Chem. Eng., 2017, 5, 5049–5058.
48. H. Mei, X. Tan, L. Tan, Y. Meng, C. Chen, M. Fang and X. Wang, Retention of U(VI) by the formation of Fe precipitates from oxidation of Fe(II), ACS Earth Space Chem., 2018, 2, 968–976.
49. J. Wu, K. Chen, X. Tan, M. Fang, X. Hu, Z. Tang and X. Wang, Core-shell CMNP@PDAP nanocomposites for simultaneous removal of chromium and arsenic, Chem. Eng. J., 2018, 349, 481–490.
50. A. Khattak, Z. Ghazi, B. Liang, N. Khan, A. Iqbal, L. Li and Z. Tang, A redox-active 2D covalent organic framework with pyridine moieties capable of faradaic energy storage, J. Mater. Chem. A, 2016, 4, 16312–16317.
51. S. Zhang, J. Li, X. Wang, Y. Huang, M. Zeng and J. Xu, Rationally designed 1D Ag@AgVO₃ nanowire/graphene/protonated g-C₃N₄ nanosheet heterojunctions for enhanced photocatalysis via electrostatic self-assembly and photochemical reduction methods, J. Mater. Chem. A, 2015, 3, 10119–10126.
52. R. He, W. Li, D. Deng, W. Chen, H. Li, C. Wei and Y. Tang, Efficient removal of lead from highly acidic wastewater by periodic ion imprinted mesoporous SBA-15 organosilica combining metal coordination and co-condensation, J. Mater. Chem. A, 2015, 3, 9789–9798.
53. H. Yan, Y. Lin, H. Wu, W. Zhang, Z. Sun, H. Cheng, W. Liu, C. Wang, J. Li, X. Huang, T. Yao, J. Yang, S. Wei and J. Lu, Bottom-up precise synthesis of stable platinum dimers on graphene, Nat. Commun., 2017, 8, 1070.
54. Y. Ai, Y. Liu, W. Lan, J. Jin, J. Xing, Y. Zou, C. Zhao and X. Wang, The effect of pH on the U(VI) sorption on graphene oxide (GO): A theoretical study, Chem. Eng. J., 2018, 343, 460–466.
55. M. Batzill and U. Diebold, Surface studies of gas sensing metal oxides, Phys. Chem. Chem. Phys., 2007, 9, 2307–2318.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8en01066h

‡ Hongshan Zhu and Jinyun Yuan contributed equally.

---