Hypervalent silicon-based, anionic porous organic polymers with solid microsphere or hollow nanotube morphologies and exceptional capacity for selective adsorption of cationic dyes

Abstract

We report straightforward access to two new anionic porous organic polymers (Si-POPs) based on hexacoordinate [SiO₆]^2– species and polyimine aromatic linkers which exhibit outstanding capabilities for removal of cationic dyes from contaminated industrial water. These multifunctional materials are stable under air and in water; they were synthesized via solvothermal chemistry exploiting Schiff base formation between tri(protocatechuic aldehyde)-silicate and 4,4′-diaminobiphenyl acid (for Si-POP-1) or 4,4′-diamino-2,2′-stilbenedisulfonic acid (for Si-POP-2). Charge neutrality of the Si-POPs was achieved by [Et₃NH]⁺ cations. The structures were determined by spectroscopic techniques (FT-IR and solid ¹³C-NMR spectroscopy), while SEM and TEM revealed a micrometer-scale solid sphere morphology for Si-POP-1 and hollow nanometer-sized tubes for Si-POP-2. BET specific surface area determination gave 376 and 234 m² g⁻¹ for Si-POP-1 and Si-POP-2, respectively. The unprecedented adsorption abilities of these materials for standard cationic dyes such as methylene blue (MB), malachite green (MG), and Basic Blue 7 (BB7) were revealed by studying the effects of system variables such as pH, contact time, temperature and mixed-dye solutions. At 25 °C, the maximum adsorption capacity of Si-POP-1 reached 300.3 mg g⁻¹ for MB and 378.5 mg g⁻¹ for MG, while the performance of Si-POP-2 at the same temperature was substantially higher (3516 mg g⁻¹ for MB and 662 mg g⁻¹ for MG). Furthermore, at 35 °C, the maximum adsorption of MB on Si-POP-2 reached 4098 mg g⁻¹, surpassing the performance of previous adsorbents. It was found that both Si-POP-1 and Si-POP-2 exhibit charge and size-selective adsorption. Thermodynamics studies indicated that the adsorptions were spontaneous and endothermic. A minimal loss of adsorptive capacity of Si-POP-2 (less than 4%) was observed after four cycles, demonstrating an effective reuse of the regenerated POP in discarding MB from effluents. Importantly, this green protocol also allows recovery of the dye retained in the POP. These assets recommend these new hexacoordinate silicate-based adsorbents as viable materials for wastewater remediation by removal of organic dye pollutants affecting nature and public health.

---

Introduction

The unprecedented growth of the global population during the last few decades has triggered tremendous progress in all areas of human activities, including introduction of advanced materials structurally fitted for specific applications. Recently, special purpose networks such as metal–organic frameworks (MOFs),¹ zeolitic imidazolate frameworks (ZIFs),² covalent organic frameworks (COFs)³, porous organic polymers (POPs)^4,5 and porous coordination polymers (PCPs)⁶ endowed with controlled morphologies and tunable porosities, high surface areas, and pre-established chemical functionalities have witnessed spectacular valorization in nanotechnology,⁷ adsorption processes,⁸ energy storage,⁹ catalysis,¹⁰ chemical and biological sensing,¹¹ medical fields,¹² photovoltaic, magnetic or luminescent devices,¹³etc.

In this context, porous organic polymers (POPs), well-established, inexpensive materials that are usually constructed through covalent bonds between C, H, O and N atoms, are equipped with high specific surface areas, low densities and rationally designed, permanent porosities. They have found many attractive applications¹⁴ in gas storage and separation, in catalysis and as catalyst supports, in drug delivery and biomedical practice and also, importantly, for adsorption of dyes, which are noxious contaminants found in effluents flowing from facilities related to textile, leather, paint, paper and other industries. With the growing demand for coloured materials, ever-increasing amounts of harmful dye pollutants are being discharged into oxygen-impoverished wastewater; these pollutants must be removed to clean the water and thus ensure environmental remediation.

Various physical, chemical and biological methods have been employed to remove organic dyes, which are hazardous to health, from industrial wastewater.^15–17 Among these, adsorption on solid networks is the most productive technique due to its low cost, superior efficiency, and simple operation.¹⁸ Because adsorbents are at the core of this strategy, it is imperative to conceptualize new adsorbents with ultrahigh adsorption capacities and outstanding selectivities in order to effectively separate dyes from industrial effluents. Charged adsorbents boast considerably improved adsorption capacities and selectivities due to the concerted effects of host–guest electrostatic (and electronic) interactions and guest–guest exchange interactions.¹⁹ Consequently, charged adsorbents, such as chemically modified activated carbon materials,²⁰ cation/anion metal–organic frameworks (MOFs)^19,21,22 and charged porous organic polymers (COFs, POPs),^23,24 surpass their neutral counterparts. However, the need for this type of adsorbent is currently still largely unmet.

To address the progressive trend of charged adsorbents, which are of utmost economic importance for the capture of noxious substances from the environment, this work aimed to prepare new anionic microporous organic polymers (Si-POP-1 and Si-POP-2) with exceptionally high adsorption strengths and selectivities for cationic dyes. Our protocol started from inexpensive, commercially available materials and employed a facile, easily scalable synthesis that yielded POPs with defined microporosities, morphologies, and specific BET surface areas. The results highlight that a simple ligand modification in the Schiff base formation step of the synthesis has an irrefutable influence on the adsorption capacity of the Si-POP. Four charged organic dyes, malachite green (MG), methylene blue (MB), Basic Blue 7 (BB7) and methyl orange (MO), commonly employed in industry and in other fields, were tested as a proof of concept that our hexacoordinated silicon-based POPs are valuable materials that are unique with regard to their removal ability of cationic dyes from human habitats. The adsorption isotherms, kinetics, thermodynamics, and factors potentially affecting the process, such as pH value, ionic strength of the solution and selective adsorption from mixed dye solutions, are also systematically investigated.

Experimental

Materials and physical measurements

All solvents and reagents for the syntheses were of analytical grade and were used as received from commercial sources without further purification. The Si-precursor (Et₃NH)₂{[(CHO)C₆H₃O₂]₃Si} was prepared by following a slightly modified literature procedure resulting in increased isolated yields, as shown in the ESI.† Elemental analyses (C, H and N) were carried out with a Perkin-Elmer 240C elemental analyzer. The IR spectra were measured in the 4000 to 400 cm⁻¹ region using a Nicolet IR-408 spectrometer and KBr pellets. Thermogravimetric (TG) curves were recorded from room temperature to 800 °C with a heating rate of 10 °C min⁻¹ on a Netzsch TG 209 instrument under nitrogen atmosphere. The N₂ adsorption isotherms were measured by a V-Sorb 2800TP surface area and pore size analyzer at 77 K. The solid ¹³C-NMR spectra were collected on a JNM-ECZ600R instrument. Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) analyses were conducted with JEOL-2100F and JSM-4800F electron microscopes (JEOL, Japan), respectively. UV-vis spectra were measured using a UV-vis 2550 spectrophotometer (Shimadzu, Japan). Powder X-ray diffraction (PXRD) patterns of the samples were performed on a Bruker D8 Advance X-ray diffractometer with Cu Kα (λ = 1.5418 Å) radiation, at room temperature, with a 2θ range from 2 to 40°.

Synthetic procedures

Synthesis of Si-POP-1 (Scheme 1). A mixture of Si-precursor (0.10 mmol, 0.0641 g), 4,4′-diaminobiphenyl (0.15 mmol, 0.0276 g) and anhydrous ethanol (10 mL) was stirred for 5 min, then sealed in a 25 mL Teflon-lined autoclave and heated at 135 °C for 4 days. After being cooled to room temperature, Soxhlet extracted in methanol for 72 h and dried, 0.0732 g of a brown powder was obtained. Anal. calcd (%) for (C₁₀₂H₁₁₂N₁₀O₁₂Si₂)_n: C, 70.97; H, 6.55; N, 8.11. Found: C, 70.95; H, 6.55; N, 8.13. Si-POP-1 synthesis was then carried out at 120 °C, 130 °C, 140 °C and 145 °C in order to assess the effects of the reaction temperature on the morphology and adsorption capacity of this material.

Scheme 1 Schematic of the synthetic pathways to obtain the hexacoordinate silicate-based porous organic polymers Si-POP-1 and Si-POP-2.

Synthesis of Si-POP-2 (Scheme 1). A mixture of Si-precursor (0.10 mmol, 0.0641 g), 4,4′-diamino-2,2′-stilbenedisulfonic acid (0.15 mmol, 0.0591 g) and anhydrous ethanol (10 mL) was stirred for 5 min. The mixture was sealed in a 25 mL Teflon-lined autoclave and heated at 135 °C for 4 days. After being cooled to room temperature, Soxhlet extraction in methanol was carried out for 72 h; 0.0935 g of grey powder was obtained. Anal. calcd (%) for (C₁₀₈H₁₁₈N₁₀O₃₀S₆Si₂)_n: C, 56.78; H, 5.20; N, 6.13. Found: C, 56.76; H, 5.21; N, 6.12. Also, synthesis of Si-POP-2 was conducted at 120 °C, 130 °C, 140 °C and 145 °C to check the influence of temperature on the morphologies and adsorption capacities of the obtained materials.

General procedures for dye adsorption experiments

Four model cationic dyes (methylene blue, MB, malachite green, MG, Basic Blue 7, BB7 and Alcian Blue 8GX, AB8GX) and an anionic dye (methyl orange, MO) were used in the study (for their structures, see the ESI, Fig. S1†) to assess the adsorption abilities and selectivities of our materials. MB, at variance with phenothiazine, to which it is structurally connected, is positively charged and adopts a planar configuration, which is important for its interactions with anionic adsorbents. MB has long been considered as an effective medicine and is ubiquitously employed in hospitals, especially its blue water solution; however, it has harmful effects on living organisms, even after short exposure. Methylene blue is used in a variety of other fields, including dyeing. Malachite green, an iminium salt formally related to triphenylmethane, is used in the dyestuff industry (for silk, leather, and cotton) and as an antibacterial agent. Anions (usually chloride or oxalate) associated with the MG cation do not influence the result of the colouring process. Basic Blue 7, also known as Victoria Blue BO, is used in the dye industry and in aquaculture for antifungal treatment. Like MG, BB7 is a triarylmethane dye; however, instead of a phenyl group, it contains a naphtylamine group and hence is substantially bulkier. MG and BB7 can adopt multiple resonance structures, with the aromatic rings flipping in different planes, and have larger volumes than the quasi-planar MB. These factors, in addition to the dissimilar chemisorption and physisorption properties of Si-POP-1 and Si-POP-2 (vide infra), contribute to the distinct observed adsorption behaviours. Alcian Blue 8GX (greenish-black crystals soluble in water) is a cationic dye that has long been applied in biological and medical studies to produce stable and fast light staining.

To investigate the adsorption capacities of Si-POP-1 and Si-POP-2 for MB, MG, BB7, AB8GX and MO, we used aqueous solutions with desired concentrations of the dye, prepared with deionized water. Experiments were performed in a thermostatic shaker bath at 160 rpm for a fixed time (10 min to 13 h). After adsorption for the selected time, the suspension was separated by centrifugation; the residual concentration of dye in the centrifugate was determined by UV-vis spectroscopy using a UV-vis 2550 spectrophotometer (Shimadzu, Japan) at the absorbance wavelengths of 664 nm and 617 nm for MB and MG, respectively. The detailed procedures are described in the ESI.† If necessary, the dye concentrations were diluted before analysis. The dye concentration in the adsorbent phase, at equilibrium or at time t was calculated using the equation q_e = (c₀ − c_e)V/m or q_t = (c₀ − c_t)V/m, where q_e and q_t are the amounts (mg g⁻¹) of dye adsorbed at equilibrium and time t. c₀, c_e, and c_t are the liquid-phase concentrations (mg L⁻¹) of the dye initially, at equilibrium, and at time t, respectively; V and m represent the volume (in L) of the dye solution and the mass (in g) of the adsorbent, respectively.

Results and discussion

The chemical structures of the Si-precursor, Si-POP-1 and Si-POP-2 were inferred from their FT-IR and solid ¹³C NMR spectra. In FT-IR (Fig. 1), a strong absorption peak at 1672 cm⁻¹ can be clearly observed for the Si-precursor, corresponding to the stretching vibration of the C=O bond in protocatechuic aldehyde. This peak disappeared in the FT-IR spectra of Si-POP-1 and Si-POP-2, while new peaks appeared at 1629 cm⁻¹ (for Si-POP-1) and 1645 cm⁻¹ (for Si-POP-2); these are indicative of the formation of –C=N imine linkages.^25a In addition, the wide peaks between 2800 and 3000 cm⁻¹ in the above three FT-IR spectra can be attributed to the asymmetric stretching vibrations of –CH₃ and –CH₂–, indicating the presence of positively charged triethylammonium.

Fig. 1 FT-IR spectrum of the Si-precursor, Si-POP-1 and Si-POP-2.

The solid ¹³C NMR spectra of Si-POP-1 and Si-POP-2 (Fig. 2) confirmed their structures. The chemical shifts at 8.2 ppm for Si-POP-1 and 8.8 ppm for Si-POP-2 were assigned to the methyl groups from trimethylamine, while the chemical shift at 47 ppm, observed for both Si-POP-1 and Si-POP-2, was ascribed to the methylene groups from triethylamine. These signals undoubtedly prove that Si-POP-1 and Si-POP-2 incorporate triethylammonium as the quaternary ammonium cation in order to balance the charge of the system. The chemical shifts at 159 ppm (for Si-POP-1) and 165 ppm (for Si-POP-2), assigned to the –C=N linkage, demonstrate the formation of the Schiff base, further confirming the imine structures proposed for Si-POP-1 and Si-POP-2. The chemical shift at about 190 ppm can be assigned to the aldehyde carbon from unreacted CHO groups at the margins of the Si-POP-1 and Si-POP-2 networks and/or of Si-precursor molecules wrapped in pores. However, the small area of the peak at ca. 190 ppm suggests that very small amounts of unreacted aldehyde groups are present in the Si-POP-1 and Si-POP-2 materials.

Fig. 2 Solid ¹³C NMR spectra of Si-POP-1 and Si-POP-2.

Studies by conventional thermogravimetric analysis (TGA) of Si-POP-1 and Si-POP-2 proved their thermal stability up to 270 °C, as anticipated considering their robust Schiff base and silicate structural motifs. A totally different degradation profile, between 270 °C and 800 °C, was observed for the two adsorbents (Fig. S2†). Si-POP-1 practically underwent a smooth weight loss from 270 °C to 800 °C, while three weight-loss steps were observed for Si-POP-2, i.e. a first slight loss between 270 °C and 420 °C followed by a sharp decrease from 420 °C to 450 °C and, finally, a second negligible drop in weight up to 800 °C. The different thermal degradation behaviours of Si-POP-1 and Si-POP-2 stem from the additional presence of SO₃H groups in Si-POP-2. Meanwhile, for both adsorbents, the smooth mass loss between 270 °C and 420 °C can be assigned to continuous degradation of the organic polymer chain and release of triethylammonium; the abrupt decay between 420 °C and 450 °C for Si-POP-2 can be accounted for by the rapid release of SO₃H groups. The distinct thermal degradation profiles of Si-POP-1 and Si-POP-2 are an important aspect in view of the practical applications of these two materials.

The N₂ adsorption isotherms of Si-POP-1 and Si-POP-2 showed type-IV curves^25b (Fig. 3). The Brunauer–Emmett–Teller surface areas (S_BET) of activated Si-POP-1 and Si-POP-2 were calculated to be 376 and 234 m² g⁻¹, respectively (P/P₀ = 0.05 to 0.25; Fig. S3†). The pore size distributions of Si-POP-1 and Si-POP-2 were calculated based on non-local density function theory (NLDFT) (curves presented in Fig. S4†). In Si-POP-1 and Si-POP-2, the micropores are centered at about 1.4 and 1.3 nm, while the mesopores are centered around 3.2 and 4.1 nm, respectively. The pore volumes for Si-POP-1 and Si-POP-2 are 0.36 and 0.33 cm³ g⁻¹ (P/P₀ = 0.9). The triethylammonium cations, located in the pores, determine the relatively low values of the S_BET and pore volumes.

Fig. 3 The N₂ isotherm adsorption curves of Si-POP-1 and Si-POP-2.

SEM images (Fig. 4, S5 and S6†) disclosed that the morphologies of the Si-POP-1 samples obtained at 130 °C, 135 °C and 140 °C are pure, solid microspheres with sizes of 1 to 5 μm; meanwhile, in the Si-POP-1 samples prepared at 120 °C and 145 °C, some irregular blocks formed among the micron-sized solid spheres. The morphologies of Si-POP-2 synthesized at 120 °C, 130 °C, 135 °C and 140 °C were found to be hollow nanotubes, with diameters of about 0.5 μm and wall thicknesses of about 200 nm. However, larger diameters were observed for the hollow tubes in the Si-POP-2 obtained at 145 °C. Furthermore, HRTEM (Fig. 4) reveals Si-POP-1 and Si-POP-2 to be solid microspheres and hollow nanotubes with pore textures, respectively. In the microporous organic polymer series, the hollow tube morphology is rarely reported.

Fig. 4 SEM, TEM and HRTEM images of Si-POP-1 (a–c) and Si-POP-2 (d–f) obtained at 135 °C.

The rationale for the different morphologies of our two materials may be their template-directed synthesis:²⁶ in Si-POP-1, the triethylammonium cation acted as a template, determining the formation of microspheres, while a mixed template (the triethylammonium cation and the sulphonic group) was responsible for creating the hollow nanotube morphology of Si-POP-2. Because the morphology, particle size, surface area, and pore structure jointly determine the properties of materials, the distinct adsorption behaviours of Si-POP-1 and Si-POP-2, discussed later, stem from the abovementioned differences in these characteristics. In addition, the broad bands extending from about 15° to 35° in the PXRD pattern (Fig. S7†) revealed the amorphous natures of Si-POP-1 and Si-POP-2, possibly with some chain crosslinking.

The temperature at which the synthesis of Si-POP-1 and Si-POP-2 was carried out had definite effects on the morphology of the material; we then sought to determine whether this changed morphology would further affect the efficiency of dye adsorption. To test this hypothesis, POP-s obtained at different temperatures were used to adsorb MB as a representative dye (the results are presented in Fig. S8†). Si-POP-1 samples prepared at 130 °C, 135 °C and 140 °C exhibited similar adsorption behaviours of approximately 295 mg g⁻¹, while the adsorptions of the samples produced at 120 °C and 145 °C were much lower. An analogous pattern was observed for Si-POP-2: specimens produced at 130 °C, 135 °C and 140 °C presented comparable adsorption levels of about 3350 mg g⁻¹. However, the adsorption capacities of Si-POP-2 originating from syntheses at 120 °C and 145 °C were about 3074 and 2600 mg g⁻¹, respectively, which are lower than those of the samples synthesized from 130 °C to 140 °C. To conclude, the various morphologies of Si-POPs stemming from different synthesis temperatures led to obvious changes in their dye adsorption performance. Samples prepared at 135 °C, considered representative, were used in the adsorption experiments unless otherwise specified.

The influence of pH on the MB and MG adsorption patterns of Si-POP-1 and Si-POP-2 was studied in solution over the pH range of 3 to 8 at 25 °C. The pH was controlled by addition of 0.1 M HCl and 0.1 M NaOH as needed (Fig. S9 and S10†). In the cases of MB and MG adsorption on Si-POP-1, the maximum adsorptions occurred at pH 6 and pH 5, respectively. When the dyes were adsorbed on Si-POP-2, the saturated adsorption increased with pH up to pH 6 and then remained constant. These results indicated that Si-POP-2 exhibits the best adsorption performance in near-neutral aqueous solutions. The effects of the initial concentration on the equilibrium adsorption capacity of the dyes are shown in Fig. 5 and 6 at 25 °C, 30 °C and 35 °C. It is clear that at relatively low concentrations of MB and MG, the adsorbed amounts on Si-POP-1 and Si-POP-2 at equilibrium increased linearly with the initial concentration. Conversely, at high initial concentrations, the adsorption became constant, indicating that saturation was reached and the maximum adsorption capacity was attained.

Fig. 5 Effects of the initial concentrations of MB and MG on adsorption by Si-POP-1.

Fig. 6 Effects of the initial concentrations of MB and MG on adsorption by Si-POP-2.

The Langmuir and Freundlich isotherm models were applied to analyse the relationships between the adsorption capacity at equilibrium (q_e, mg g⁻¹) of Si-POP-1 and Si-POP-2 and the equilibrium concentrations of the MB and MG dyes (c_e, mg L⁻¹). The equations of the Langmuir (eqn (1)) and Freundlich (eqn (2)) isotherm models^20b,c are presented below:

where k_L (L mg⁻¹) is the Langmuir constant and q_max (mg g⁻¹) is the maximum monolayer adsorption capacity. k_f (mg g⁻¹) and 1/n (n, in L mg⁻¹) represent the Freundlich constants corresponding to the adsorption capacity and adsorption intensity, respectively. The q_evs. c_e curves of MB and MG on Si-POP-1 (Fig. S11†) and Si-POP-2 (Fig. S12†) and their fitting curves by the Langmuir (Fig. S13†) and Freundlich (Fig. S14†) isotherm models were further considered.

The fitting results are given in Tables 1 and S1.† The values of the correlation coefficient R² obtained from fitting with the Freundlich isotherm model (Table S1†) greatly differ from unity, while the R² values obtained from fitting with the Langmuir isotherm model (Table 1) are closer to 1; this indicates that the Langmuir model gives better fits to the experimental data. Examination of the adsorption capacities q_max and the Langmuir adsorption constants (k_L) of MB and MG on Si-POP-1 and Si-POP-2, estimated from the Langmuir model, reveals that at 25 °C, the maximum adsorption capacities of Si-POP-1 for MB and MG are 300.3 and 378.8 mg g⁻¹, respectively. Importantly, at saturation, the adsorption amounts of Si-POP-2 for MB and MG are 3516 and 662.2 mg g⁻¹, respectively. These theoretical values are very close to the experimental values. Upon increasing the temperature, the adsorption capacities of Si-POP-1 and Si-POP-2 for MB and MG substantially increased. On both Si-POP materials, the adsorption of MB was considerably stronger than that of MG due to their different molecular sizes and shapes: MB, which contains three linearly condensed rings, is planar and smaller, whereas MG is bulkier because its three cycles are angularly disposed (Fig. S1†). Therefore, MB fits better in the empty inner spaces of both Si-POP matrices. Significantly, the adsorption capacity of Si-POP-2 for MB or MG is far superior to that of Si-POP-1. The main reason is that Si-POP-2, in addition to its different morphology (hollow nanotubes), possesses negatively charged sites (RSO₃⁻ groups) that are prone to electrostatic interactions.²⁷

Table 1 Langmuir isotherm parameters for adsorption of MB and MG on Si-POP-1 and Si-POP-2 at 25 °C, 30 °C and 35 °C

Adsorbent	Dye	T (K)	q max (mg g^−1)	q exp (mg g^−1)	k L (L mg^−1)	R ^2
Si-POP-1	MB	298.15	300.3	288.9	0.4076	0.9977
		303.15	330.0	318.5	0.5861	0.9988
		308.15	355.9	350.7	1.254	0.9995
	MG	298.15	378.8	369.0	0.6804	0.9993
		303.15	450.4	436.1	0.6727	0.9991
		308.15	500.0	488.2	1.163	0.9988
Si-POP-2	MB	298.15	3516	3415	0.1151	0.9994
		303.15	3806	3680	0.1245	0.9997
		308.15	4098	3980	0.1341	0.9993
	MG	298.15	662.2	637.6	0.3922	0.9966
		303.15	704.2	679.8	0.5441	0.9973
		308.15	757.6	730.9	0.6875	0.9976

---

Next, the propensity of Si-POP-2 for adsorbing MB and MG was evaluated in comparison with that of other reported materials; note that we only considered materials which exhibit adsorption levels above 900 mg g⁻¹ for MB and 600 mg g⁻¹ for MG (Tables 2 and 3). From Table 2, it can be concluded that the adsorption capacity of Si-POP-2 for MB is ultra-high, greatly surpassing those of other adsorbents. The logical basis for this exceptional tendency is a combination of chemisorption within the pores of the adsorbent with physisorption, which is attributable to strong electrostatic interactions between the positively charged sites of the dyes and the negatively charged sites of the adsorbents. Both chemisorption and physisorption were strongly enhanced for Si-POP-2 due to the supplemental presence of the sulphonic groups (two for each silicate unit). The structural dissimilarity of the two adsorbents Si-POP-1 and Si-POP-2 consists mainly of the presence of a large number of sulphonic groups in the Si-POP-2 network. On account of this, we speculate that the great difference in the adsorption capacities of the two adsorbents is mainly due to the following: (1) sulphonic groups provide a large number of strong electrostatic interaction sites; as a result, cationic dye molecules tend to be adsorbed more densely on the inner surface of the adsorbent. In contrast, in Si-POP-1, the smaller number of sites and the weaker electrostatic interactions lead to sparse adsorption of dye molecules. (2) Sulphonic acid groups are hydrophilic, which is conducive to the diffusion and penetration of water molecules. In particular, water can rapidly and efficiently spread to the interior of the adsorbent, driving the small dye molecules to diffuse rapidly everywhere inside the adsorbent; this is propitious for electrostatic interactions between the dye molecules and the ionizable SO₃H groups on the adsorbent framework. The high efficiency of space utilization provides extremely high dye adsorption capacity. However, in Si-POP-1, which is hydrophobic due to its lack of hydrophilic groups, diffusion of water and dye molecules within the adsorbent is very difficult. Consequently, the adsorption of the dye molecules and the ion exchange occur at the entrance of the channels, partially blocking them. Low space utilization leads to lower dye adsorption capacity. Therefore, the sulphonic groups play a crucial role in the adsorption process and essentially enhance it. Additionally, the hollow nanotube morphology and the adequate channel shape of Si-POP-2 provide better access of MB and MG for sustained chemisorption at the sulphonic groups, in contrast with Si-POP-1.

Table 2 Adsorption capacity of Si-POP-2 for MB in comparison with other materials

Adsorbent	Adsorbed amount (mg g^−1)	Ref.
Si-POP-2	3516 (25 °C)	This work
	3806 (30 °C)	This work
	4098 (35 °C)	This work
Calcium alginate membrane	3056.4 (25 °C)	28
GO/MCC aerogels	2630 (25 °C)	29
Sodium alginate crosslinked polyacrylic acid (SA-cl-poly(AA)–TiO2)	2257.4 (45 °C)	30
Magnetic polyacrylamide microspheres	1977 (30 °C)	31
GK-cl-P(AA-cl-AAM)/SiO2 nanocomposite (HNC)	1408.7 (25 °C)	32
MIL-68	1666 (25 °C)	33
Commercial activated carbon	980.3 (30 °C)	34
Chemically activated carbon spheres derived from poly(vinyl alcohol) microspheres	925.9 (45 °C)	35
Teak wood bark	914.6 (20 °C)	36
PVBA (43%)-UiO-66	909 (30 °C)	37

---

Table 3 Adsorption capacity of Si-POP-2 for MG in comparison with other materials

Adsorbent	Adsorbed amount (mg g^−1)	Ref.
ZIF-67	2430 (20 °C)	38
Granular composite hydrogel (AA-IA-APT hydrogel)	2433	39
Magnetic β-cyclodextrin–graphene oxide nanocomposites (Fe3O4/β-CD/GO)	990.10 (45 °C)	40
Si-POP-2	757 (35 °C)	This work
Novel grape waste activated carbon (GWAC)	667 (room temp.)	41

---

The ultra-high adsorption ability of Si-POP-2 for MB and MG is a prerequisite for its large-scale practical applications.

Adsorption thermodynamics provides information on the inherent energy changes in the adsorbent and on the mechanism involved in the adsorption process. The thermodynamics parameters, Gibbs free energy change (ΔG⁰, kJ mol⁻¹), enthalpy change (ΔH⁰, kJ mol⁻¹), and entropy change (ΔS⁰, J mol⁻¹ K⁻¹) can be calculated using the following equations:⁴²

ΔG⁰ = −RT ln K^d; (4)

In these equations, K^d is the distribution coefficient for the adsorption, q_e (mg g⁻¹) is the amount of dye adsorbed per mass unit of the adsorbent at equilibrium and c_e (mg L⁻¹) is the concentration of the dye solution at equilibrium. R is the gas constant, with a value of 8.314 J mol⁻¹ K⁻¹, and T (K) is the thermodynamic temperature of the dye solution. The results evidenced that all the Gibbs free energy values were negative (Table S2†), ranging from −4.271 to −8.038 kJ mol⁻¹; this suggests that adsorption of MB and MG on Si-POP-1 and Si-POP-2 is spontaneous. The increase of the absolute values of ΔG⁰ as a function of temperature convincingly indicated that the adsorption behaviours were spontaneous at higher temperatures. The large positive ΔH⁰ values of 64.55 kJ mol⁻¹ for MG on Si-POP-1, 48.08 kJ mol⁻¹ for MB on Si-POP-1, 38.43 kJ mol⁻¹ for MG on Si-POP-2 and 42.74 kJ mol⁻¹ for MB on Si-POP-2 show that adsorption was endothermic at the solid–solution interface.⁴³ If ΔH⁰ is greater than 40 kJ mol⁻¹, the adsorption process is considered to be chemisorption. In contrast, if the ΔH⁰ value is less than 40 kJ moL⁻¹, the adsorption is a physisorption process. Therefore, the above high ΔH⁰ values indicate a chemisorption nature of the ion exchange, which is effective during adsorption of MB and MG on Si-POP-1 and Si-POP-2 (Fig. S15†).⁴⁴ In fact, the super-high adsorption capacity of these two adsorbents should be attributed to synergism between chemisorption and physisorption, which was the underlying concept when designing our two adsorbents. The positive ΔS⁰ values suggest that during adsorption, the degrees of freedom increased at the solid–liquid interface.

Two classical kinetics models, the pseudo-first-order kinetics model and pseudo-second-order kinetics model, are frequently used to evaluate the uptake rate and adsorption efficiency of an adsorbent as well as to analyse the mechanism of dye adsorption. These models⁴³ can be expressed as ln(q_e − q_t) = ln q_e − k₁t and t/q_t = 1/k₂(q_e)² + (1/q_e)t, respectively, where q_t and q_e are the adsorbed amounts of dye at time t and at equilibrium (mg g⁻¹), respectively, and t is the time (min); k₁ and k₂ are the pseudo-first-order rate constant (min⁻¹) and the pseudo-second-order rate constant (g mg⁻¹ min⁻¹). The kinetic parameters were obtained by fitting the plots of q_tvs. t (Fig. 7 and S16†) and are listed in Table S3.† In comparison with the pseudo-first-order kinetics model (Fig. S17†), the pseudo-second-order kinetics model is more suitable for analysing the adsorption kinetics because the values of the correlation coefficients R² derived from the pseudo second-order model are closer to 1. Thus, this outcome suggests that MG and MB adsorption on Si-POP-1 and Si-POP-2 follows pseudo second-order kinetics. Thermodynamic and kinetic data collectively suggest that the driving force of MB and MG adsorption is chemisorption by ion exchange and electrostatic interactions. In the case of Si-POP-2, the ion exchange involves the cationic dye and the triethyl ammonium cations or hydrogens in the RSO₃H groups. Electrostatic interactions intervene between the positively charged sites of the dye and the negatively charged sites on the adsorbent (i.e. octahedral silicate cores and RSO₃⁻ groups). Fortunately, π–π stacking interactions between the dye and the adsorbent participate in the process. Physisorption contributes to both diffusion and adsorption, particularly for Si-POP-2, which is endowed with a hollow nanotube morphology.

Fig. 7 Effects of the contact time t on the adsorption of MB and MG by Si-POP-1 (a) and Si-POP-2 (b) at 25 °C. Initial volume and concentration for (a): 50 mL, 80 mg L⁻¹ MB or MG; (b): 75 mL, 550 mg L⁻¹ for MB and 75 mL, 120 mg L⁻¹ for MG. The mass of each adsorbent, Si-POP-1 or Si-POP-2, was 10 mg.

As shown by the thermodynamic and kinetic analysis, ion exchange phenomena play a very important role in dye adsorption. Metal cations present in the solution interfere with the adsorption process. Sodium and calcium ions were used to study interference effects on the adsorption amount of the dye at saturation. The concentration of each metal ion was set at 100 ppm, which is the concentration of these metal ions in hard water. Comparative assessment (Fig. S18 and S19†) evidenced that sodium and calcium ions play negative roles in the adsorption of MB and MG on Si-POP-1, provoking a distinct decrease of the adsorbed amount at saturation. In contrast, sodium and calcium ions had almost no effect on the adsorbed amounts of MB and MG on Si-POP-2, supposedly because these cations readily exchange with the acidic hydrogens from the sulfonic groups of the adsorbent; thus, the replacement of triethylammonium cations is relegated to a minor role.

Selectivity of adsorption was demonstrated by experiments where the cationic dyes MB and MG were adsorbed on Si-POP-1 and Si-POP-2 (Tables 1–3 and S4†); a mixture of the anionic dye MO and MG was also studied (Fig. 8). From the UV-vis spectra of an aqueous solution containing MO (50 ppm) and MG (50 ppm) adsorbed by Si-POP-1 (left) and Si-POP-2 (right), it is obvious that the intensity of the maximum absorption peak (617 nm) of MG decreased with time, while the intensity of the maximum absorption peak (463 nm) of MO remained practically unchanged; this indicates that MO was not adsorbed, being negatively charged, despite the fact that MO is less bulky than MG. These results clearly indicate selective adsorption on both adsorbents of cationic dyes with different sizes or of MG vs. the anionic dye MO. This outcome provides support to the fact that the adsorption processes were dominated by ion exchange/physisorption and chemisorption, as also confirmed by the thermodynamic and kinetic data. At the same time, it stands as proof that the designed adsorbents exhibit highly selective adsorption for cationic dyes such as MG, MB and BB7.

Fig. 8 UV-vis spectra of a mixture of MO and MG aqueous solutions adsorbed by Si-POP-1 (a) and Si-POP-2 (b).

Adsorbents Si-POP-1 and Si-POP-2 clearly show different adsorption levels and adsorption rates for our set of cationic dyes. We then sought to determine whether these adsorbents can exhibit adsorbate size and shape selectivity. When Si-POP-1 and Si-POP-2 were used to adsorb a mixture with equal concentrations of MB and MG, differences in the adsorption profiles were evident (Fig. 9). On Si-POP-1, the intensities of both the peak at 617 nm of MG and the peak at 664 nm of MB decreased with time, suggesting that MG and MB were simultaneously adsorbed by Si-POP-1. This was also true for Si-POP-2; however, the peak at 664 nm decreased much more rapidly, indicating the preferential adsorption of MB. In addition, larger dyes, Basic Blue 7 (BB7) and Alcian Blue 8GX (AB8GX) (Fig. S1†), were used to probe the size selectivities of Si-POP-1 and Si-POP-2 (Fig. S20–S22†). The adsorption experiments were carried out at 25 °C. It was found that the adsorption capacities of Si-POP-1 for BB7 and AB8GX were about 44 and 42 mg g⁻¹, while those for Si-POP-2 were 740 and 56 mg g⁻¹, respectively. These very low adsorption levels indicate that adsorption of bulky dyes can occur only on the outside surface of the adsorbents. In summary, Si-POP-1 can easily adsorb relatively small dyes such as MB and MG; however, it adsorbed almost none of the larger BB7 and AB8GX. Contrastingly, Si-POP-2 could adsorb BB7 more readily than AB8GX. The obvious conclusion is that both Si-POP-1 and Si-POP-2 exhibit size selectivity. An important deduction is that surface adsorption is operative, to a greater or lesser extent, within the whole range of dyes employed, whereas for both adsorbents, inner chemisorption prevails for small dyes, especially in the case of Si-POP-2.

Fig. 9 UV-vis spectra of mixed MB and MG aqueous solutions absorbed on (a) Si-POP-1 (160 ppm MB + 160 ppm MG) and (b) Si-POP-2 (400 ppm MB + 400 ppm MG).

Because it is of particular significance for material recycling and dye recovery, the release process of MB fully adsorbed on 20 mg Si-POP-2 was monitored by UV-vis. To this end, the MB-loaded adsorbent was introduced in methanol (150 mL) and UV-vis spectra of the solution were recorded (Fig. 10). It was found that during the first 4 hours, the intensity of the UV-vis absorption band that is characteristic of MB increased slowly and then levelled to a plateau, indicating limited release of the dye. After 240 min, NaCl was added in one portion; this triggered much faster release of MB, as indicated by the abrupt increase in intensity of the UV-vis peaks of the dye. A large amount of MB was therefore liberated from the adsorbent via ion-exchange between the adsorbed dye and NaCl, favoured by the good solubility of the dye in methanol. The release experiments with the MB-loaded Si-POP-2 confirmed that the adsorption process occurs predominantly within the adsorbent and only to a moderate extent on the surface.

Fig. 10 Release of MB adsorbed on Si-POP-2 (30 mg) in methanol (150 mL) before and after NaCl addition.

Currently, in industry, there is an urgent need for systematic water management by minimizing the amount of the water consumed while maximizing water reuse via wastewater treatment. For practical application, if our material with super-high adsorption capacity for cationic dyes could be regenerated and reused, the total operational cost could be significantly decreased and a higher volume of water could be cleaned. In this context, we chose to investigate recycling of Si-POP-2 for MB adsorption. Subsequent to a first adsorption process, Si-POP-2 was regenerated by short immersion in a NaCl-saturated methanolic solution, dried and reused in a subsequent MB adsorption cycle. After four adsorption–desorption cycles, Si-POP-2 retained 96% of its impressive initial adsorption capacity (Fig. 11). This result proves that productive recycling can largely exceed the four cycles depicted in our experiment. The simplicity and sustainability of the regeneration procedure suggests that MB itself can easily be recovered in view of valorisation in the dye industry, thus decreasing the cost of MB fabrication and the related environmental damage. It should be noted that in addition to this respect, our protocol outclasses reported adsorption on inexpensive, naturally existing materials or wastes that permit neither adsorbent recycling or dye recovery.^45–48

Fig. 11 Recyclability of regenerated Si-POP-2 for adsorption of MB (10 mg Si-POP-2; 150 mL of a 550 mg L⁻¹ MB solution; 25 °C, 6 h).

Conclusions

Discovering advanced materials that allow capture and release of wastewater pollutants under mild conditions is currently an active area of research. In this regard, the present study introduces two new anionic, hypervalent silicate-based microporous organic polymers (Si-POP-1 and Si-POP-2) for adsorption of cationic dyes. These materials are readily accessed from the direct, hydrothermal Schiff base-driven polycondensation of a hexacoordinate Si-precursor (containing an octahedral dianionic “SiO₆” structural fragment) with a bifunctional aromatic amine. Amine source variation was used to control the properties of the networks, resulting in distinct morphologies (solid microspheres for Si-POP-1 and hollow nanotubes for Si-POP-2), BET specific surface areas and, most importantly, remarkable yet different adsorption propensities for model cationic dye contaminants such as MB, MG, BB7 and AB8GX. In particular, Si-POP-2 showed a record adsorption capacity for removal of MB (above 3500 mg g⁻¹), largely surpassing reported values for most known adsorbents.

The outstanding adsorption power of the dyes was assigned to chemisorption by ion exchange, as confirmed by the observed pseudo-second-order kinetics and enthalpy increase. Taken together with the negative values of the Gibbs free energy, it was demonstrated that the adsorptions occurred spontaneously.

These novel anionic adsorbents are versatile, robust, selective for dyes and more efficient than other POPs. They are stable under air and in water for a long time. Recycling experiments of Si-POP-2 proved that it retains its adsorption capability (>96%) for more than four runs. Additionally, in accord with green chemistry principles, this adsorbent is cleanly and readily regenerable; therefore, it may be industrially viable, considering its convenient, scalable synthesis and efficient use over many cycles with positive environmental consequences.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (21501122, 21671139 and 21701116), the Doctoral Scientific Research Foundation of Liaoning Province (201601193) and the Distinguished Professor Project of Liaoning province (No. 2013204). I. D. and V. D. acknowledge financial support from the Romanian Academy, Institute of Organic Chemistry C. D. Nenitescu, Bucharest.

Notes and references

1. (a) S. Yuan, L. Feng, K. Wang, J. Pang, M. Bosch, C. T. Lollar, Y. Sun, J. Qin, X. Yang, P. Zhang, Q. Wang, L. Zou, Y. Zhang, L. Zhang, Y. Fang, J. Li and H.-C. Zhou, Adv. Mater., 2018, 1704303; (b) H. Liang, X. Jiao, C. Li and D. Chen, J. Mater. Chem. A, 2018, 6, 334–341; (c) S. Yuan, J.-S. Qin, C. T. Lollar and H.-C. Zhou, ACS Cent. Sci., 2018, 4, 440–450; (d) L.-X. You, L.-X. Cui, B.-B. Zhao, G. Xiong, F. Ding, B. Ren, Z.-L. Shi, I. Dragutan, V. Dragutan and Y. Sun, Dalton Trans., 2018, 47, 8755–8763; (e) M. J. Van Vleet, T. Weng, X. Li and J. R. Schmidt, Chem. Rev., 2018, 118, 3681–3721; (f) H. L. Nguyen, F. Gandara, H. Furukawa, T. L. Doan, K. E. Cordova and O. M. Yaghi, J. Am. Chem. Soc., 2016, 138, 4330–4333.
2. (a) G. Zhong, D. Liu and J. Zhang, J. Mater. Chem. A, 2018, 6, 1887–1899; (b) S. Tanaka, T. Nagaoka, A. Yasuyoshi, Y. Hasegawa and J. F. M. Denayer, Cryst. Growth Des., 2018, 18, 274–279; (c) E. V. Ramos-Fernandez, A. Grau-Atienza, D. Farrusseng and S. Aguado, J. Mater. Chem. A, 2018, 6, 5598–5602; (d) D. Saliba, M. Ammar, M. Rammal, M. Al-Ghoul and M. Hmadeh, J. Am. Chem. Soc., 2018, 140, 1812–1823.
3. (a) X. Guan, Y. Ma, H. Li, Y. Yusran, M. Xue, Q. Fang, Y. Yan, V. Valtchev and S. Qiu, J. Am. Chem. Soc., 2018, 140, 4494–4498; (b) G. Zhang, M. Tsujimoto, D. Packwood, N. Tuan Duong, Y. Nishiyama, K. Kadota, S. Kitagawa and S. Horike, J. Am. Chem. Soc., 2018, 140, 2602–2609; (c) O. Yahiaoui, A. N. Fitch, F. Hoffmann, M. Fröba, A. Thomas and J. Roeser, J. Am. Chem. Soc., 2018, 140, 5330–5333.
4. (a) B. Li, Z. Guan, X. Yang, W. D. Wang, W. Wang, I. Hussain, K. Song, B. Tan and T. Li, J. Mater. Chem. A, 2014, 2, 11930–11939; (b) Z. Xiang and D. Cao, J. Mater. Chem. A, 2013, 1, 2691–2718; (c) J. X. Jiang and A. I. Cooper, Top. Curr. Chem., 2010, 293, 1–33; (d) R. Castaldo, G. Gentile, M. Avella, C. Carfagna and V. Ambrogi, Polymers, 2017, 9, 651.
5. J. Liang, Y.-B. Huang and R. Cao, Coord. Chem. Rev., 2019, 378, 32–65.
6. (a) H. Zhu, D. Liu, D. Zou and J. Zhang, J. Mater. Chem. A, 2018, 6, 6130–6154; (b) L. You, W. Zong, G. Xiong, F. Ding, S. Wang, B. Ren, I. Dragutan, V. Dragutan and Y. Sun, Appl. Catal., A, 2016, 511, 1–10; (c) L. X. You, W. Zhu, S. Wang, G. Xiong, F. Ding, B. Ren, I. Dragutan, V. Dragutan and Y. Sun, Polyhedron, 2016, 115, 47–53; (d) Y.-G. Sun, J. Li, J. You, Y. Guo, G. Xiong, B.-Y. Ren, L.-X. You, F. Ding, S.-J. Wang, F. Verpoort, I. Dragutan and V. Dragutan, J. Coord. Chem., 2015, 68, 1199–1212; (e) M. L. Foo, R. Matsuda and S. Kitagawa, Chem. Mater., 2014, 26, 310–322.
7. L. Chen, R. Luque and Y. Li, Dalton Trans., 2018, 47, 3663–3668.
8. (a) A. J. Rieth and M. Dincă, J. Am. Chem. Soc., 2018, 140, 3461–3466; (b) D. Zhao, X. H. Liu, J. H. Guo, H. J. Xu, Y. Zhao, Y. Lu and W. Y. Sun, Inorg. Chem., 2018, 57, 2695–2704.
9. (a) Y. Jiao, J. Pei, D. Chen, C. Yan, Y. Hu, Q. Zhang and G. Chen, J. Mater. Chem. A, 2017, 5, 1094–1102; (b) C. A. Downes and S. C. Marinescu, ChemSusChem, 2017, 10, 4374–4392; (c) A. Elsayed, E. Elsayed, R. Al-Dadah, S. Mahmoud, A. Elshaer and W. Kaialy, Appl. Energy, 2017, 186(SI3), 509–519.
10. (a) Y.-S. Kang, Y. Lu, K. Chen, Y. Zhao, P. Wang and W.-Y. Sun, Coord. Chem. Rev., 2018 DOI:10.1016/j.ccr.2018.02.009; (b) Metal-Organic Frameworks: Applications in Separation and Catalysis, ed. H. Garcia and S. Navalon, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 1st edn, 2018; (c) G. Xiong, X.-L. Chen, L.-X. You, B.-Y. Ren, F. Ding, I. Dragutan, V. Dragutan and Y.-G. Sun, J. Catal., 2018, 361, 116–125; (d) H. Kaur, M. Venkateswarulu, S. Kumar, V. Krishnan and R. R. Koner, Dalton Trans., 2018, 47, 1488–1497; (e) Y.-B. Huang, J. Liang, X.-S. Wang and R. Cao, Chem. Soc. Rev., 2017, 46, 126; (f) L.-X. You, H.-J. Liu, L.-X. Cui, F. Ding, G. Xiong, S.-J. Wang, Ba.-Y. Ren, I. Dragutan, V. Dragutan and Y.-G. Sun, Dalton Trans., 2016, 45, 18455–18458.
11. (a) T. Gong, P. Li, Q. Sui, J. Chen, J. Xu and E.-Q. Gao, J. Mater. Chem. A, 2018, 6, 9236–9244; (b) P. Samanta, A. V. Desai, S. Sharma, P. Chandra and S. K. Ghosh, Inorg. Chem., 2018, 57, 2360–2364.
12. (a) U.-J. Ryu, J. Yoo, W. Kwon and K. M. Choi, Inorg. Chem., 2017, 56, 12859–12865; (b) L. Wang, W. Wang, X. Zheng, Z. Li and Z. Xie, Chem.–Eur. J., 2017, 23, 1379–1385; (c) W. Chen and C. Wu, Dalton Trans., 2018, 47, 2114–2133.
13. (a) L. L. Luo, X. L. Qu, Z. Li, X. Li and H. L. Sun, Dalton Trans., 2018, 47, 925–934; (b) P. Tao, C.-H. Zeng, K. Zheng, C.-Q. Huang and S. L. Zhong, J. Inorg. Organomet. Polym. Mater., 2016, 26, 1087–1094; (c) Y.-N. Sun, G. Xiong, V. Dragutan, I. Dragutan, F. Ding and Y.-G. Sun, Inorg. Chem. Commun., 2015, 62, 103–106; (d) Z. Xiang and D. Cao, J. Mater. Chem. A, 2013, 1, 2691–2718.
14. (a) L. Tan and B. Tan, Chem. Soc. Rev., 2017, 46, 3322–3356; (b) W. Zhang, B. Aguila and S. Ma, J. Mater. Chem. A, 2017, 5, 8795–8824; (c) R. Castaldo, G. Gentile, M. Avella, C. Carfagna and V. Ambrogi, Polymers, 2017, 9, 651; (d) C. Lu, S. Liu, J. Xu, Y. Ding and G. Ouyang, Anal. Chim. Acta, 2016, 902, 205–211; (e) C. Perego and R. Millini, Chem. Soc. Rev., 2013, 42, 3956–3976.
15. M. T. Yagub, T. K. Sen, S. Afroze and H. M. Ang, Adv. Colloid Interface Sci., 2014, 209, 172–184.
16. M. Rafatullah, O. Sulaiman, R. Hashim and A. Ahmad, J. Hazard. Mater., 2010, 177, 70–80.
17. (a) E. M. Dias and C. Petit, J. Mater. Chem. A, 2015, 3, 22484–22506; (b) A. A. Adeyemo, I. O. Adeoye and O. S. Bello, Toxicol. Environ. Chem., 2012, 94, 1846–1863.
18. (a) M. Ge and H. Liu, J. Mater. Chem. A, 2016, 4, 16714–16722; (b) X. Liu, W. Gong, J. Luo, C. Zou, Y. Yang and S. Yang, Appl. Surf. Sci., 2016, 362, 517–524; (c) L. Liu, X.-N. Zhang, Z.-B. Han, M.-L. Gao, X.-M. Cao and S.-M. Wang, J. Mater. Chem. A, 2015, 3, 14157–14164; (d) P. Y. Du, H. Li, X. Fu, W. Gu and X.-A. Liu, Dalton Trans., 2015, 44, 13752–13759; (e) S.-N. Zhao, C. Krishnaraj, H. Jena, D. Poelman and P. Van Der Voort, Dyes Pigm., 2018, 156, 332–337; (f) J. Liu, P. Chen, G. Shao, B. Fan, H. Wang, D. Chen, H. Lu and R. Zhang, J. Inorg. Organomet. Polym. Mater., 2018, 28, 790–799; (g) S. Kaushal, R. Badru, S. Kumar, H. Kaur and P. Singh, J. Inorg. Organomet. Polym. Mater., 2018, 28, 968–977; (h) C. H. Liu, L. J. Zhang, F. Y. Bai, Y. Wang, Y. Z. Hong, C. R. Li and Y. H. Xing, J. Inorg. Organomet. Polym. Mater., 2018, 28, 1839–1849.
19. (a) S. Zhang, J. Li, D. Y. Du, J. S. Qin, S. L. Li, W. W. He, Z. M. Su and Y. Q. Lan, J. Mater. Chem. A, 2015, 3, 23426–23434; (b) J. J. Li, C. C. Wang, H. F. Fu, J. R. Cui, P. Xu, J. Guo and J. R. Li, Dalton Trans., 2017, 46, 10197–10201; (c) Z. Zhu, Y. L. Bai, L. Zhang, D. Sun, J. Fang and S. Zhu, Chem. Commun., 2014, 50, 14674–14677.
20. (a) K. Gupta and O. P. Khatri, J. Colloid Interface Sci., 2017, 501, 11–21; (b) Z. Jia, Z. Li, S. Li, Y. Li and R. Zhu, J. Mol. Liq., 2016, 220, 56–62; (c) T. Ma, P. R. Chang, P. Zheng, F. Zhao and X. Ma, Chem. Eng. J., 2014, 240, 595–600.
21. (a) X. Zhao, X. Bu, T. Wu, S. T. Zheng, L. Wang and P. Feng, Nat. Commun., 2013, 4, 2344; (b) Y. C. He, J. Yang, W. Q. Kan, H. M. Zhang, Y. Y. Liu and J. F. Ma, J. Mater. Chem. A, 2015, 3, 1675–1681; (c) D. M. Chen, J. Y. Tian, Z. W. Wang, C. S. Liu, M. Chen and M. Du, Chem. Commun., 2017, 53, 10668–10671; (d) H. Hahm, S. Kim, H. Ha, S. Jung, Y. Kim, M. Yoon and M. Kim, CrystEngComm, 2015, 17, 8418–8422.
22. S.-N. Zhao, C. Krishnaraj, H. Sekhar Jena, D. Poelman and P. Van Der Voort, Dyes Pigm., 2018, 156, 332–337.
23. (a) Z. Li, H. Li, X. Guan, J. Tang, Y. Yusran, Z. Li, M. Xue, Q. Fang, Y. Yan, V. Valtchev and S. Qiu, J. Am. Chem. Soc., 2017, 139, 17771–17774; (b) Y. Liu and K. Landskron, J. Mater. Chem. A, 2017, 5, 23523–23529; (c) Z. Xiang and D. Cao, J. Mater. Chem. A, 2013, 1, 2691–2718.
24. (a) J. Roeser, D. Prill, M. J. Bojdys, P. Fayon, A. Trewin, A. N. Fitch, M. U. Schmidt and A. Thomas, Nat. Chem., 2017, 9, 977–982; (b) H. Ma, B. Liu, B. Li, L. Zhang, Y. G. Li, H. Q. Tan, H. Y. Zang and G. Zhu, J. Am. Chem. Soc., 2016, 138, 5897–5903; (c) S. Fischer, J. Schmidt, P. Strauch and A. Thomas, Angew. Chem., Int. Ed., 2013, 52, 12174–12178; (d) H. Zhu, D. Liu, D. Zou and J. Zhang, J. Mater. Chem. A, 2018, 6, 6130–6154.
25. (a) L. Xiao, D. C. Sun, T. Li Niu and Y.-W. Yao, Phosphorus, Sulfur Silicon Relat. Elem., 2014, 189, 1564–1571; (b) M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier, F. Rodriguez-Reinoso, J. Rouquerol and K. S. W. Sing, Pure Appl. Chem., 2015, 87, 1051–1069.
26. (a) Y. Xie, D. Kocaefe, C. Chen and Y. Kocaefe, J. Nanomater., 2016, 2302595,  DOI:10.1155/2016/2302595; (b) S. Aghamohammadi and M. Haghighi, Chem. Eng. J., 2015, 264, 359–375.
27. A. A. Narvekar, J. B. Fernandes and S. G. Tilve, J. Environ. Chem. Eng., 2018, 6, 1714–1725.
28. Q. Li, Y. Li, X. Ma, Q. Du, K. Sui, D. Wang, C. Wang, H. Li and Y. Xia, Chem. Eng. J., 2017, 316, 623–630.
29. X. Wei, T. Huang, J. H. Yang, N. Zhang, Y. Wang and Z. W. Zhou, J. Hazard. Mater., 2017, 335, 28–38.
30. S. Thakura, S. Pandeya and O. Arotibaa, Carbohydr. Polym., 2016, 153, 34–46.
31. T. Yao, S. Guo, C. Zeng, C. Wang and L. Zhang, J. Hazard. Mater., 2015, 292, 90–97.
32. H. Mittal, A. Maity and S. S. Ray, Chem. Eng. J., 2015, 279, 166–179.
33. M. S. Tehrani and Z. D. Rouholah, RSC Adv., 2016, 6, 27416–27425.
34. N. Kannan and M. M. Sundaram, Dyes Pigm., 2001, 51, 25–40.
35. Z. Jia, Z. Li, S. Li, Y. Li and R. Zhu, J. Mol. Liq., 2016, 220, 56–62.
36. G. McKay, J. F. Porter and G. R. Prasad, Water, Air, Soil Pollut., 1999, 114, 423–438.
37. Y. F. Yang, Z. H. Niu, H. Li, Y. H. Ma, Y. Zhang and H. F. Wang, Dalton Trans., 2018, 47, 6538–6548.
38. K. Y. A. Lin and H. A. Chang, Chemosphere, 2015, 139, 624–631.
39. Y. Zheng, Y. Zhu and A. Wang, Chem. Eng. J., 2014, 257, 66–73.
40. D. Wang, L. Liu, X. Jiang, J. Yu and X. Chen, Colloids Surf., A, 2015, 466, 166–173.
41. H. Sayğılı and F. Güzel, Chem. Eng. Res. Des., 2015, 100, 27–38.
42. Y. Liu, J. Chem. Eng. Data, 2009, 54, 1981–1985.
43. (a) K. L. Tan and B. H. Hameed, J. Taiwan Inst. Chem. Eng., 2017, 74, 25–48; (b) A. Asfaram, M. R. Fathi, S. Khodadoust and M. Naraki, Spectrochim. Acta, Part A, 2014, 127, 415–421; (c) O. D. Uluozlu, A. Sari, M. Tuzen and M. Soylak, Bioresour. Technol., 2008, 99, 2972–2980.
44. (a) N. S. A. Mubarak, A. H. Jawad and W. I. Nawawi, Energy Ecol. Environ., 2017, 2, 85–93; (b) M. E. Argun, D. Güclü and M. Karatas, J. Ind. Eng. Chem., 2014, 20, 1079–1084.
45. L. dos Santos Silva, J. de Oliveira Carvalho, R. D. de Sousa Bezerra, M. S. da Silva, F. J. L. Ferreira, J. A. Osajima and E. C. da Silva Filho, Molecules, 2018, 23, 743.
46. M. Mon, R. Bruno, J. Ferrando-Soria, D. Armentano and E. Pardo, J. Mater. Chem. A, 2018, 6, 4912–4947.
47. D. S. Tong, C. W. Wu, M. O. Adebajo, G. C. Jin, W. H. Yu, S. F. Ji and C. H. Zhou, Appl. Clay Sci., 2018, 161, 256–264.
48. L. Mouni, L. Belkhiri, J.-C. Bollinger, A. Bouzaza, A. Assadi, A. Tirri, F. Dahmoune, K. Madani and H. Remin, Appl. Clay Sci., 2018, 153, 38–45.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8ta07109h

---