The influence of the leaving iodine atom on phyllosilicate syntheses and useful application in toxic metal removal with favorable energetic effects

Abstract

Novel nanostructured 2 : 1 hybrid cobalt and nickel phyllosilicates containing attached diethyl iminodiacetate moieties within the interlamellar spaces were synthesized using a mild non-hydrothermal sol–gel methodology. The organofunctionalization of these hybrids was achieved using silylating agents formed by the reaction of 3-chloropropyltriethoxysilane or 3-iodopropyltrietoxysilane with diethyl iminodiacetate, whereas 3-iodopropyltriethoxysilane was synthesized from the reaction of 3-chloropropyltriethoxysilane with NaI. The degrees of functionalization of the new materials are associated with the nucleophilic displacement of the halide atom by the nitrogen basic center of diethyl iminodiacetate during the silylating agents' syntheses. The incorporation of an iodine atom favored the introduction of bulky basic moieties in the pendant chain. For example, the degree of functionalization for the nickel phyllosilicate functionalized from 3-iodopropyltriethoxysilane, (1.83 ± 0.01) mmol g⁻¹, was higher than that of the nickel phyllosilicate functionalized from 3-chloropropyltriethoxysilane, (0.59 ± 0.02) mmol g⁻¹. Infrared spectroscopy combined with ¹³C and ²⁹Si NMR spectroscopy confirmed the attachment of organic moieties covalently bonded to the silicon sheet network, and the lamellar 2 : 1 trioctahedral phyllosilicate structures were confirmed by XRD. Further characterization was provided by thermogravimetry, SEM and TEM, which exhibited thermally stable hybrids presenting well-formed particles with a homogeneous distribution of cobalt, nickel, and nitrogen. The attached basic centers have the ability to sorb lead and cadmium from aqueous solution. The sorption data fit the Langmuir model and indicated maximum lead sorption values of (2.11 ± 0.16) mmol g⁻¹ for the cobalt phyllosilicate functionalized from 3-iodopropyltriethoxysilane and (1.99 ± 0.11) mmol g⁻¹ for the hybrid prepared from 3-chloropropyltriethoxysilane, for example, reflecting the higher degree of functionalization of the former. All of the other hybrids exhibited the same tendency, even for cadmium sorption. The thermodynamic data obtained from calorimetric titration revealed the spontaneity of these chelating interactive processes, which are enthalpically and entropically favorable for the proposed cation-basic center interactions at the solid–liquid interface.

---

Introduction

The advanced production of functional sorbent materials for toxic metal ion elimination from wastewater is a current research focus for environmental pollution control processes.^1,2 Generally, materials developed for this application should feature specific sites on their surfaces for interaction with metal ions; however, most typical known materials lack such surface properties. Therefore, attaching suitable functional groups containing binding sites to convert classical inorganic structures into powerful sorbents for toxic metal recovery is an appealing approach to this problem.^3,4

Waste streams generated from many industrial activities, such as mining operations, metal plating facilities and tanneries are contaminated by toxic metals. As a consequence, the soils surrounding many of these industries can be contaminated and pose a risk of metal contamination of nearby groundwater and surface water. Some recurring metals associated with these activities are lead, cadmium, chromium and mercury, which are extremely toxic because they are non-biodegradable and tend to accumulate in living organisms, causing many disorders and diseases.^5–8

Chemical precipitation, membrane filtration, ion exchange and carbon sorption are some of the classical treatment processes for the purification of contaminated waste streams. However, alternative low-cost effective technologies and sorbents for the treatment of metal-contaminated industrial waste streams are needed.⁵ Some functional materials that either are naturally available or can be easily synthesized with suitable binding centers may have potential as inexpensive sorbents. Many different organofunctionalized structures have been used as functional sorbents for toxic metallic cation removal from aqueous solutions, such as silica-gel, mesoporous silica, phyllosilicates and biopolymers.^8–12 The functionalization of these structures with organic groups containing basic nitrogen and sulfur has allowed the removal of large amounts of mercury, lead, cadmium and copper ions, for instance.^13–16 Among these structures, synthetic organofunctionalized phyllosilicates are of particular interest because of the high degree of functionalization achieved as a result of the particularities of the synthetic procedures applied in their preparations.^13,17,18 In addition, this lamellar structure also has thermal, chemical and mechanical stability as well as good dispensability and swellability in polar solvents, which make these solids exceptional sorbents. Synthetic phyllosilicates feature high purity, homogeneity and controlled porosity, which are advantages over natural phyllosilicates.^19,20 Because of all of these advantages, the removal of large amounts of toxic metals by organoclays has been reported for many different systems since the pioneer work published almost two decades ago.²¹

Such organoclays have layered structures with an approximate composition of Si₈R₈Mg₆O₁₆(OH)₄, analogous to those found in natural minerals, such as talc, in which the organic functionalities are located in the interlayer space. The stacked lamellas in the organoclay structure are composed by a brucite sheet occupied by a hexacoordinated metal, such as magnesium, nickel or cobalt, in an octahedral environment, which is connected to either one or two single slightly disordered R–SiO₃ tetrahedral sheets to form a 1 : 1 or the well-known talc-like 2 : 1 phyllosilicate structure, respectively.^22–25

Normally, the organofunctionalization of these structures is performed by the chemical modification of a precursor silylating agent with a molecule containing the desired functional group.^19,26 Because the main advantage of organofunctionalized phyllosilicates over other sorbents for toxic metal sorption is the higher degree of functionalization achieved, the development of methods that maximize the amount of pendant organic groups attached to the inorganic structure is extremely desirable. One approach is increasing the yield of the reaction between the precursor silylating agent and the desired organic molecule bearing the functional group.

The precursor 3-chloropropyltriethoxysilane has been widely used in the organofunctionalization of inorganic structures with the aim of a posterior increase of the carbonic chain by introducing different functionalities. The organo-chloro functionalized matrix containing a C–Cl end group can be used to incorporate other organic moieties onto the surface through the displacement of the end chlorine atom by any basic center from another eligible molecule. The capacity of the chlorine atom of this silylating agent to react with basic centers has been widely explored with the goal of obtaining materials with various applications. A wide variety of molecules can be attached to this precursor, and promising features can be attained when incorporated on an inorganic surface.^27–30

To increase the efficiency of this nucleophilic substitution reaction, a leaving atom superior to chlorine must be used. Thus, iodopropyl pendant groups could be a particularly useful candidate for subsequent modification because iodine, located at the extremity of this molecule, is a very good leaving atom in nucleophilic displacement, much better than chlorine.^31,32 Furthermore, 3-iodopropyltriethoxysilane (IPTS) can be produced from the displacement of chlorine by iodine in 3-chloropropyltriethoxysilane (CPTS) by the reaction of the last with sodium iodide. The use of sodium iodide to obtain IPTS from CPTS is also cost-effective, as commercial pure 3-iodopropyltriethoxysilane is much more expensive than 3-chloropropyltriethoxysilane.

Successful syntheses in homogeneous conditions employing diethyl iminodiacetate can yield a new silylating agent, which can then be reacted via a sol–gel process to form the inorganic layered structures. This established procedure is of extreme importance for the development of this area of chemistry, enabling high-yield syntheses of new silylating agents and thereby providing organoclays with high degrees of functionalization, which can improve the efficiency of these organoclays in several applications.

The objective of this investigation is to explore the influence of the exchange of chlorine by iodine in the precursor silane 3-chloropropyltriethoxysilane in the synthesis of a new silylating agent prepared from the nucleophilic substitution of iodine by diethyl iminodiacetate to yield novel cobalt and nickel phyllosilicates containing pendant nitrogen basic centers. The resultant hybrids thus contain active centers for cation removal from aqueous solution, and this ability was measured via the sorption of toxic divalent lead and cadmium ions. In addition, the energetic features related to metal/basic center interactions at the solid–liquid interface were calorimetrically studied to determine the thermodynamic parameters related to these interactions.

Experimental section

Chemicals

3-Chloropropyltriethoxysilane (CPTS) and diethyl iminodiacetate were obtained as high-grade reagents (Aldrich) and used as supplied. Anhydrous sodium iodide (Synth) was used to convert 3-chloropropyltriethoxysilane into 3-iodopropyltrietoxysilane (IPTS) in dry acetone (Synth) solvent. All other reagents, including nickel nitrate hexahydrate (Vetec), cobalt nitrate hexahydrate (Vetec), ethanol (Synth) and sodium hydroxide (Synth), were also of high grade and were used as received without prior purification. Sodium hydroxide (Synth) solutions were prepared in deionized water (ultra-pure Milli-Q Millipore, 18.2 MΩ cm), as were the lead and cadmium nitrate (Vetec) solutions used in the sorption and calorimetric experiments.

Synthesis of 3-iodopropyltriethoxysilane

The commercial precursor 3-chloropropyltriethoxysilane (15.0 g, 62.35 mmol) was added dropwise to a solution of sodium iodide (14.0 g, 93.5 mmol) dissolved in 200 cm³ of dry acetone. This solution was heated at 343 K and kept under reflux for 6 days under stirring in a dry nitrogen atmosphere. This procedure generated a white sodium chloride precipitate, which was removed by filtration under nitrogen flow to ensure an inert environment to prevent the hydrolysis of the silane alkoxy groups. The solvent was removed under vacuum, and 100 cm³ of diethyl ether was added to the system to precipitate the residual sodium iodide, followed by another filtration under nitrogen atmosphere. This procedure of diethyl ether addition was repeated four times to ensure the complete extraction of the sodium iodide. The product was then distilled under reduced pressure (10 mmHg), and a clear yellowish liquid, 3-iodopropyltriethoxysilane, was recovered at 369–371 K.³³

Synthesis of organofunctionalized phyllosilicates

First, a new silylating agent was synthesized by adding 25.0 cm³ of an ethanolic solution of diethyl iminodiacetate (6.62 g, 35.0 mmol) dropwise to a solution of 3-iodopropyltriethoxysilane (11.63 g, 35.0 mmol) dissolved in 25.0 cm³ of ethanol, followed by the dropwise addition of 1.77 g (17.5 mmol) of triethylamine. The mixture was maintained under reflux in a dry nitrogen atmosphere for 24 h.

For the synthesis of the cobalt phyllosilicate, the new silylating agent was slowly added to a solution containing 7.64 g (26.25 mmol) of Co(NO₃)₂·6H₂O in 100 cm³ of dry ethanol under magnetic stirring at 323 K. The Si/Co ratio was adjusted to 4/3, which is the molar ratio found in a 2 : 1 trioctahedral phyllosilicate. After the addition of the silane, 100 cm³ of an aqueous solution of sodium hydroxide (0.50 mol dm⁻³) was added dropwise under stirring as the temperature was maintained at 323 K. This procedure immediately formed a grey suspension that was left to age for 5 days at the same temperature. The resultant solid was filtered, washed with ethanol in a Soxhlet system and dried for 12 h under vacuum at 323 K to give the new material, named PhCoI.^13,34

The nickel phyllosilicate functionalized with diethyl iminodiacetate groups was obtained through the same procedure using 7.63 g (26.25 mmol) of Ni(NO₃)₂·6H₂O, and the new material was named PhNiI.

Two other cobalt and nickel phyllosilicates were synthesized using the same procedure, substituting the synthesized 3-iodopropyltriethoxysilane by 8.43 g (35.0 mmol) of 3-chloropropyltriethoxisilane, to evaluate the efficiency of the nucleophilic substitution of chlorine by iodine on the degrees of functionalization of the hybrids. The cobalt and nickel phyllosilicates obtained in these processes were named PhCoCl and PhNiCl, respectively.

Characterization

Powder X-ray diffraction patterns were collected on a Shimadzu model XRD-7000 diffractometer using Cu Kα radiation (0.154 nm) at 45 kV and 30 mA at a rate of 1.00 × 10⁻²° s⁻¹ for 2θ measurements in the range of 1.4 to 70°.

The calculations of the amount of organic pendant chains immobilized onto the organofunctionalized phyllosilicates were based on the nitrogen contents determined through elemental analysis on a Perkin Elmer model 2400 elemental analyzer.

Infrared spectra of the samples in potassium bromide pellets were recorded on a Bomem MB-Series FTIR spectrophotometer by accumulating 40 scans in the range of 4000 to 400 cm⁻¹ with a resolution of 4 cm⁻¹.

¹H and ¹³C nuclear magnetic resonance spectra of the liquid silanes 3-chloropropyltriethoxisilane and 3-iodopropyltriethoxysilane were acquired in CD₃COCD₃ solutions on a Bruker Avance III 500 MHz with spectral widths of 20 and 263 ppm for ¹H and ¹³C, respectively. The acquisition and relaxation times were 1.59 and 1 s, respectively, for the ¹H spectra and 0.5 and 2 s, respectively, for the ¹³C spectra. The chemical shifts are reported in ppm and referenced to the residual solvent peak of tetramethylsilane (TMS).

Solid-state ¹³C and ²⁹Si nuclear magnetic resonance spectra were acquired on a Bruker Avance II 400 spectrometer at room temperature by assaying the solid sample compacted into a 7 mm zirconium oxide rotor. The measurements were performed at 75.47 and 59.61 MHz for carbon and silicon, respectively, using the magic angle spinning (MAS) technique at 10 kHz. To increase the signal-to-noise ratio, the cross-polarization (CP) technique was applied. Pulse repetitions of 1 and 3 s and contact times of 1 and 3 ms were set for the acquisition of the ¹³C and ²⁹Si NMR spectra, respectively.

The thermogravimetric curves were obtained on a TA 2050 instrument using a heating rate of 0.167 K s⁻¹ in an argon atmosphere under a flow of 30 cm³ s⁻¹ at temperatures ranging from room temperature to 1273 K.

Secondary electron images were acquired on a JEOL JSM 6360-LV scanning electron microscope operating at 20 kV. Energy dispersive X-ray spectroscopy (EDS) was used for elemental mapping on a Noran System Six instrument. The samples were fixed onto double-faced carbon tape adhered to an aluminum support and carbon-coated in a Bal-Tec MD20 instrument.

Transmission electron microscopy (TEM) images were obtained on a JEM 2100 FEG microscope operating at 200 kV at the National Laboratory of Nanotechnology (LNNano) located at the National Center for Research in Energy and Materials (CNPEM) in Campinas, SP, Brazil. The samples were ultrasonically suspended in isopropyl alcohol for 30 min, and the suspensions were deposited on carbon-coated copper grids.

The maximum toxic metal sorption capacity was calculated from the difference between the initial metal concentration in aqueous solution and that found in the supernatant at the equilibrium concentration using a Perkin Elmer model 3000 DV inductively coupled plasma optical emission spectrometer (ICP-OES). For each experimental point, the reproducibility was checked by at least one duplicate run.

The thermal effects at the solid–liquid interface were measured using a Thermometrics model LKB 2277 isothermal calorimeter. The calorimetric titrations of lead and cadmium aqueous solutions were performed onto approximately 20 mg of the chemically modified surfaces suspended in 2.0 cm³ of water inside the calorimeter vessel, with mechanical stirring at 100 rpm, at 298.15 ± 0.20 K. A microsyringe containing the corresponding titrand solutions with a known concentration of approximately 0.10 mol dm⁻³ was coupled to the calorimeter vessel. A series of fixed volumes of 20.0 mm³ were incrementally added, and the baseline was reached after 2 h. The thermal effects of cation dilution and the addition of water to the hybrid suspensions were identically verified using the same experimental calorimetric conditions.

Results and discussion

Silylating agent

A new silylating agent was synthesized, through two different routes, from the reactions of 3-chloropropyltriethoxysilane and 3-iodopropyltriethoxysilane with diethyl iminodiacetate individually. The silane IPTS used in this modification was synthesized by the substitution of chlorine in the CPTS agent by iodine. This displacement favors subsequent modifications to attach functional groups to the organic chain through nucleophilic substitution, taking into account that iodine is a better leaving atom than chlorine.³¹ This displacement reaction and the syntheses of the new silylating agent from the reactions of CPTS and IPTS with diethyl iminodiacetate are depicted in Fig. 1.

Fig. 1 Syntheses of the new silylating agent from the reactions of 3-chloropropyltriethoxysilane (a) and 3-iodopropyltriethoxysilane (b) with diethyl iminodiacetate. 3-Iodopropyltriethoxysilane was obtained from the reaction of 3-chloropropyltriethoxysilane with sodium iodide (c).

The precursor CPTS agent and the corresponding derivative IPTS were characterized by ¹H and ¹³C nuclear magnetic resonance spectroscopy. The spectra are presented in Fig. 2 along with insets containing the numbered structures of the silanes, showing the relationship between the generated peaks and the hydrogen and carbon elements in the organic chains.

Fig. 2 ¹H and ¹³C nuclear magnetic resonance spectra of 3-chloropropyltriethoxysilane (a) and (b) and 3-iodopropyltriethoxysilane (c) and (d), respectively.

The peaks in the ¹H spectrum for CPTS, displayed in Fig. 2(a), are in accordance with the respective silane organic structure, as proved by the triplet at 0.75 ppm (3), corresponding to the hydrogen atoms of the methylene group bonded to the silicon atom. Another triplet at 1.30 ppm (1) is attributed to the hydrogen atoms of the methyl groups located at the extremities of the ethoxy groups. A quintet is observed at 1.92 ppm (4) due to the methylene protons located in the middle of the organic chain of the propyl groups, and a triplet at 3.52 ppm (5) is attributed to the protons of the methylene group bonded to a chlorine atom. Finally, a quartet at 3.90 ppm (2) reflects the presence of methylene groups bonded to an oxygen atom in the alkoxy group.⁴

Fig. 2(c) presents the ¹H NMR spectrum of IPTS, which is the product of the nucleophilic substitution of 3-chloropropyltriethoxysilane. This spectrum displays some peaks at similar chemical shifts to 3-chloropropyltriethoxysilane due to the similarities in the structures of both agents. However, an additional peak is observed at 3.22 ppm (5), attributed to the protons of the methylene group bonded to the iodine atom, and the peak at 3.52 ppm previously observed in CPTS spectrum is almost completely suppressed. The displacement of this peak to lower chemical shift values reflects the exchange of chlorine by iodine, which is less electronegative, resulting in a stronger shielding effect for the neighboring protons. Nonetheless, it is clear that the conversion was not complete based on the residual peak at 3.52 ppm, while other peaks presented as multiplets, inferring that the final product contains a small amount of residual 3-chloropropyltriethoxysilane. This residual quantity could be measured through the integration of the remaining signal at 3.52 ppm in the IPTS spectrum, which, when compared to the other peaks, indicates a high nucleophilic substitution degree of 98%.

The ¹³C spectra of 3-chloropropyltriethoxysilane and 3-iodopropyltriethoxysilane, shown in Fig. 2(b) and (d), respectively, also confirm the halogen nucleophilic substitution. Both spectra display well-defined peaks at 12.3, 18.3, 27.6 and 58.4 ppm, whose correspondence to the carbons of the organic chains is represented by the numbered structures in the insets. Again, the difference between these two spectra concern the peaks related to the carbon atoms bonded to the halogens, which are at 47.3 and 10.5 ppm in CPTS and IPTS, respectively, in agreement with the proper substitution of chlorine by iodine.³⁵

The infrared spectrum of the obtained 3-iodopropyltriethoxisilane agent (Fig. S1†) exhibits bands at 2974, 2928 and 2887 cm⁻¹ attributed to the asymmetric stretching of the methyl groups of alkoxy moieties and the asymmetric and symmetric stretching of methylene groups, respectively, pertaining to the organic chain of the synthesized silane. Another characteristic band found at 1110 cm⁻¹ is related to the C–O stretching bonds present in the alkoxy groups. The confirmation of the nucleophilic substitution of chlorine by iodine is given by the presence of the absorption bands at 1167 cm⁻¹, related to the deformation of the H₂C–I bond, and at 687 cm⁻¹, related to the stretching of the C–I bond.³⁵

Organofunctionalized phyllosilicates

The newly synthesized silylating agent was successfully employed in the sol–gel process to obtain the layered inorganic–organic hybrid nanocompounds. The organofunctional trialkoxysilane agent is the silicon source for the formation of the structure with covalent silicon–carbon bonds, while the metal sources were embedded inside the inorganic layer, resulting in nickel or cobalt atoms positioned in an octahedral arrangement. In the 2 : 1 phyllosilicate solids formed, the organic part, containing diethyl iminodiacetate moieties, is distributed inside the interlayer lamellar cavities and on the external surfaces.

The detailed elemental analysis results for the synthesized nickel and cobalt phyllosilicates are listed in Table S1.† These results show the quantities of carbon, hydrogen and nitrogen atoms present in the nanocompounds. In addition, based on the nitrogen percentages obtained, the amount of pendant organic groups, represented as the degree of functionalization of all hybrids, was evaluated.

These results indicated that higher degrees of functionalization were achieved for the phyllosilicates functionalized from IPTS, PhCoI ((1.62 ± 0.01) mmol g⁻¹) and PhNiI ((1.83 ± 0.01) mmol g⁻¹), compared to those prepared from CPTS, PhCoCl ((1.29 ± 0.01) mmol g⁻¹) and PhNiCl ((0.59 ± 0.02) mmol g⁻¹). This result corroborates the proposition that the exchange of the chlorine atom by an iodine atom increases the degree of displacement of the halide atoms in nucleophilic substitution to form the new silylating agent because iodine is a much better leaving group than chlorine.³¹

The layered phyllosilicate structures were investigated by X-ray powder diffraction (XRD), and their diffraction patterns are shown in Fig. 3. Four main peaks are observed in all cases, labeled as (001), (020, 110), (130, 200) and (060, 330) reflections according to the general indexation of organic–inorganic talc-like hybrid diffraction patterns. Additionally, the spectra of these hybrids do not show any signals related to the presence of either cobalt or nickel hydroxides, which could have been side products in these syntheses.^36–39

Fig. 3 X-ray diffraction patterns of the synthesized PhCoCl (a), PhCoI (b), PhNiI (c) and PhNiCl (d) phyllosilicates.

The broadness of the peaks compared with those of the similar natural talc was interpreted in terms of intralayer and stacking disorder due to the insertion of organic chains in the interlamellar regions and the smaller particle diameters of the synthetic solids as well. In particular, the (020) reflections were more intense and modified in shape in the new organoclay materials. In addition, the intralayer reflection (060), characteristic of 2 : 1 trioctahedral phyllosilicate, is evidenced in the diffraction patterns of all of the phyllosilicates synthesized, confirming the proper formation of this structure.⁴⁰ In fact, this peak is less intense in the pattern for PhNiCl because of its poorer organization associated with the heterogeneity of the pendant chains since less chlorine atoms were substituted by diethyl iminodiacetate than in the case of the silane containing iodo groups. Such lowest organization is reinforced by the absence of the (002) reflection between 10 to 15° seen in all the other solids.

The calculated basal spacing of 1801, 1184, 2493 and 1410 pm for PhNiI, PhNiCl, PhCoI and PhCoCl was based on the position of the (001) reflection, calculated from Bragg's law. The (001) reflection in PhCoI pattern is not well-defined because it is located at lower limit of the instrument, since this material possess the highest basal distance. These results clearly show that hybrids functionalized with iodopropyl groups featured higher values of basal spacing compared with those functionalized with chloropropyl moieties. This tendency is expected because higher amount of iodine atoms have undergone reaction with diethyl iminodiacetate than chlorine atoms; thus, the phyllosilicates obtained from 3-iodopropyltriethoxysilane should present higher interlayer spacing, as they possess higher amount of bulky pendant diethyl iminodiacetate groups. These results strongly suggest that the organic fragments are indeed located in the interlayer space of the inorganic–organic hybrid materials.

Infrared spectroscopy is a useful technique for proving the effectiveness of the functionalization processes because it can detect the presence of the structural bonds that compose the attached organic chains within the phyllosilicate interlayers. The infrared spectra of all hybrid phyllosilicates (Fig. S2†) are quite similar, with typical bands related to the clay structure, including sorbed water molecules, and other bands associated with the functional organic groups attached as all of them possess the same attached functional groups and very similar inorganic structures.

All spectra show bands at 2975 cm⁻¹ attributed to the asymmetric stretching of C–H bonds in CH₃ groups and modes related to the symmetric and asymmetric stretching of C–H bonds in CH₂ groups at 2883 and 2922 cm⁻¹, respectively. Sharp bands at 1200 cm⁻¹ are attributable to the stretching of Si–C bonds. The C–O stretching of the C=O bonds belonging to the ethyl acetate groups is represented by bands located at 1735 cm⁻¹. This set of bands confirms the presence of pendant organic groups and the retention of the structural integrity of the silane organic chains after the sol–gel component of the syntheses.^41,42

The broad bands at 3460 cm⁻¹, assigned to the ν(O–H) vibration, are related to the presence of intercalated water molecules between the phyllosilicate layers and on the external surfaces and to structural silanol groups bonded to the inorganic structure. This statement is reinforced by the bands at 1625 cm⁻¹ assigned to the angular vibration of water molecules bonded to the inorganic structural backbone.⁴³

The stretching of siloxane (Si–O–Si) groups of the tetrahedral phyllosilicate sheets is evidenced by the bands at 1128 and 1033 cm⁻¹, which correspond to symmetric and asymmetric stretching modes, respectively. The deformation of Si–OH bonds is represented by bands at 920 cm⁻¹, while ν(Si–OH) modes are assigned to the band at 690 cm⁻¹. In addition, bands at 607 and 645 cm⁻¹ are attributable to δNi–O and δCo–O vibrations, respectively. Finally, the bands at 447 and 469 cm⁻¹ are assigned to the stretching of Ni–O and Co–O bonds. The detection of all of these bands is in agreement with the effective formation of the clay structure.^13,41

The solid-state ¹³C NMR spectra of PhCoI and PhNiI hybrids (Fig. S3†) also give valuable information about the presence and integrity of the pendant organic chains and the nature of the attachment to the inorganic structure formed. Both spectra show broad signals at 13.3 ppm, which is an overlapping of peaks attributed to the carbon atoms of the pendant chain bonded to carbon and silicon atoms; at 28.5 ppm, related to the carbons bonded to the more electronegative nitrogen and oxygen atoms; and at 129.7 ppm, assigned to the carbon of the carbonyl group. These correspondences are clearly indicated in Fig. S3†, in agreement with the inserted numbered structure. These signals prove the presence of functional organic groups in the structure and their integrity after the functionalization processes.^9,13

Additional information about the phyllosilicate structures was obtained from the solid-state ²⁹Si nuclear magnetic resonance spectra of PhCoI and PhNiI phyllosilicates (Fig. S4†).

The signals at −58.0 and −67.8 ppm and at −58.3 and −66.9 ppm observed in the PhCoI and PhNiI spectra are attributed to the silicon atoms of the T² [RSi*(OSi)(OH)(OM)] and T³ [RSi*(OSi)₂(OM)] species, respectively, where R stands for the attached pendant groups and M represents the metals cobalt and nickel that occupy the octahedral sites of the lamellar structures. These signals confirm that the organic chains are covalently bonded to the silicon atoms of the inorganic structure, proving the effectiveness of the functionalization processes. The higher intensity of the T³ signals relative to the T² signals in both spectra, and the absence of Q species, implies that a high condensation degree between the alkoxy groups was obtained during the sol–gel process, leading to well-formed lamellar structures.⁴⁴

Scanning electron microscopy (SEM) was used to verify the particle morphology of the synthesized phyllosilicates at the microscale and to estimate the particle size of the solids. The respective micrographs are exhibited in Fig. 4.

Fig. 4 Scanning electron microscopy images of PhCoCl (a), PhCoI (b), PhNiCl (c) and PhNiI (d).

The images of all phyllosilicates display well-formed particles with an inconstant format and irregular size distribution. These characteristics are common in organofunctionalized phyllosilicates synthesized through the sol–gel method under mild temperature and pressure conditions. The irregular shape and size are also associated with the structural disorder caused by the insertion of organic groups within the lamellas.^26,37,45

The energy dispersive scanning (EDS) results of cobalt or nickel, depending on the phyllosilicate type, and silicon and nitrogen atoms for PhCoI and PhNiI are presented in Fig. 5.

Fig. 5 EDS images of silicon (a), cobalt (b) and nitrogen (c) mappings on the PhCoI hybrid surface (d) as well as silicon (e), nickel (f) and nitrogen (g) mappings on the PhNiI surface (h).

The images reveal cobalt or nickel atoms homogeneously dispersed all over the surfaces of the materials, as well as silicon atoms, inferring good condensation rates between the silicon precursors from silylating agents and metal salts in the form of nitrates. These synthesized hybrids were acquired from the sol–gel process, as there is no evidence of the formation of agglomerates or domains of the corresponding oxides. The fine dispersion of nitrogen atoms also indicates that the substitution reaction between iodine atoms in the silane for diethyl iminodiacetate groups had a good yield. In this process, the basic centers available on a large portion of the surface of the solids can favor the efficiency of these hybrids in toxic metal sorption or other applications in which this property would be useful.¹³

Transmission electron microscopy (TEM) images were acquired for all hybrid phyllosilicates, as displayed in Fig. 6, to unveil the structural characteristics of the solids. The four hybrids presented quite similar structures, showing stacked phyllosilicate platelets approximately 50–300 nm in width. These materials do not exhibit highly crystalline lamellar structures, as already indicated by XRD analyses, due to the insertion of organic groups between the structural layers. The lamellar structure formation is evidenced by the overlapping of the layers clearly shown in some regions of the images. This morphology is typical of organofunctionalized phyllosilicates acquired through the sol–gel method, as previously reported.^45–47

Fig. 6 TEM images of the hybrids PhCoCl (a), PhCoI (b), PhNiCl (c) and PhNiI (d).

Thermogravimetric curves (Fig. S5†) were acquired to evaluate the thermal stability of the solids and the amount of organic groups attached to the inorganic surface, along with their respective derivative curves.

The hybrids presented total mass losses of 47.9, 58.4, 41.7 and 58.5% for PhCoCl, PhCoI, PhNiCl and PhNiI, respectively. These values are attributed to the combination of the release of water molecules physically sorbed on the surface, the decomposition of the attached organic chains, and the collapse of the inorganic structures, leading to the formation of final oxides as residues, such as CoO, NiO and SiO₂.^20,48 In general, these events were observed in the range of 310 to 440, 500 to 790 and 860 to 1090 K, respectively, according to the derivative thermogravimetric curves. The highest mass losses due to organic moiety decompositions are those observed for hybrids synthesized from IPTS silane, as given by the values of 36.5 and 37.3% for PhCoI and PhNiI, respectively. In contrast, the hybrids synthesized from CPTS, PhCoCl and PhNiCl, gave mass losses of 31.1 and 35.2% due to the decomposition of the pendant organic groups, respectively. These data are in agreement with the degrees of functionalization of the materials functionalized from IPTS being highest, inferring that chlorine substitution by iodine in the precursor agent CPTS was favorable, as already proved by elemental analysis.

All of the hybrids except PhNiI presented two peaks in the derivative curve related to the decomposition of organic matter. This may be a result of the mixture of pendant organic chains bonded to the inorganic structure, as the surfaces still feature some chloropropyl and iodopropyl groups that have not reacted with diethyl iminodiacetate.

Sorption experiments

The organofunctionalization of the inorganic structures of cobalt and nickel phyllosilicates with organic molecules containing basic atoms, such as nitrogen and oxygen, attach Lewis basic centers to the surfaces of these solids. The surface basic sites promote the ability of these hybrids to sorb toxic metallic cations that behave as Lewis acids in this interactive effect when available in aqueous medium. Thus, during the sorption process, the formation of complexes with divalent transition metals occurs through the available electron pairs pertaining to the nitrogen and oxygen atoms in the organic chains on the solid surfaces.⁴³

The interactive effects of the chemically modified matrices applied for the sorption of lead and cadmium ions were studied through batchwise experiments at constant temperature. For this purpose, the hybrids were dispersed in different solutions with different concentrations of these metallic cations. The amount of metallic cations remaining in solution after the sorption process was measured by inductively coupled plasma – optical emission spectroscopy (ICP-OES). From the concentrations of the cation in the supernatant after equilibrium (C_s), it was possible to calculate the number of moles of metallic cations fixed, per gram of sorbent, on the hybrid surfaces (N_f) through eqn (1).^4,13,49

N_f = (n_i − n_s)/m (1)

where n_i and n_s are the amounts of ions in solution before the beginning of the process and after equilibrium, respectively, and m is the mass of the functionalized hybrid.

Lead sorption

The capacity of the new synthesized hybrids to sorb lead ions per unit of mass at constant temperature was evaluated by batchwise experiments, and the obtained results were fit to the Langmuir model through non-linear regression.

Important parameters related to the sorption processes between the surfaces and lead ions, such as the number of ions needed to form a monolayer of sorbate on the surface (N^s) and the equilibrium constant related to the sorption process (b), were obtained through the non-linear regression of the Langmuir equation, presented in eqn (2).^4,9

Lead sorption experiments were carried out for all synthesized hybrids, and the corresponding isotherms are displayed in Fig. 7.

Fig. 7 Experimental data and predicted Langmuir isotherms (solid lines) for the sorption of lead ions onto PhCoCl (a), PhCoI (b), PhNiCl (c) and PhNiI (d) surfaces at 298 ± 1 K.

All isotherms exhibited the behavior expected by Langmuir equation, varying only in the intensity of the amount of metallic cations sorbed. These organofunctionalized phyllosilicates presented higher sorption capacities than many materials previously reported due to the high density of organic pendant groups on the inorganic structures.⁵⁰

Among all of the phyllosilicates, the N^s values obtained for PhCoI and PhNiI, synthesized from IPTS, of (1.84 ± 0.07) and (2.11 ± 0.16) mmol g⁻¹, respectively, were higher than those for PhCoCl and PhNiCl, prepared from CPTS, which were (1.67 ± 0.08) and (1.99 ± 0.11) mmol g⁻¹, respectively. These results clearly indicate the advantage of the substitution of chlorine atoms by iodine atoms in the precursor silylating agent. This substitution leads to higher densities of organic pendant groups; as a result, higher amounts of metallic ions can be sorbed. The results of the analysis of lead sorption on the surfaces are listed in more detail in Table S2,† which also presents the equilibrium constants of the processes.

Cadmium sorption

Cadmium sorption experiments were also performed with the phyllosilicates functionalized with diethyl iminodiacetate, namely, PhCoCl, PhCoI, PhNiCl and PhNiI. The corresponding isotherms are shown in Fig. 8.

Fig. 8 Experimental data and predicted Langmuir isotherms (solid lines) for the sorption of Cd²⁺ onto PhCoCl (a), PhCoI (b), PhNiCl (c) and PhNiI (d) surfaces at 298 ± 1 K.

The maximum cadmium sorption capacities (N^s) were lower than those obtained for lead because cadmium is a softer Lewis acid than lead and interacted less favorably with diethyl iminodiacetate groups, as these pendant organic moieties only contain borderline nitrogen and hard oxygen basic centers. Nonetheless, an increase in the maximum sorption value was observed when the chlorine was substituted by iodine in the silylating agent to synthesize nickel phyllosilicates, where PhNiI presented a N^s value of (1.00 ± 0.05) mmol g⁻¹, which is more than double the obtained value for PhNiCl, (0.49 ± 0.05) mmol g⁻¹. In the case of the cobalt phyllosilicates, the sorption value obtained for PhCoCl of 0.58 mmol g⁻¹ was maintained even after halogen substitution in the precursor silylating agent. These N^s values and the equilibrium constant of the sorption processes of cadmium sorption for the four phyllosilicates are presented in Table S2.†

Calorimetry

The use of calorimetry is crucial for the acquisition of important data related to the energetic effects generated by the interactions between metallic cations in aqueous solutions and the pendant basic centers on the hybrid surfaces. The enthalpy of the sorption of divalent cations on the functionalized surfaces and the values of Gibbs energy and entropy related to the interactive processes were obtained from the measurement of the net thermal effects. Sequential additions of aliquots of aqueous metallic solutions in the calorimetric vessel containing the functionalized solids were performed, aiming the acquisition of numeric values that allow the evaluation of the spontaneity of the interactions associated with the studied systems and the stability of the complexes formed after the metallic cation sorption.^13,51

The acquired energetic parameters reflect the thermal effects related to the interactions of the hybrids with metals in solution (Q_sor), while the values related to the dilution of the metallic solutions (Q_dil) and the solid hydration (Q_h) in the calorimetric vessel (Q_dil) were subtracted from the global values (Q_tit).^4,52

The thermal effects associated with each step of the titration of the solids (S) in suspension (sp) with the divalent metals (M²⁺) can be represented by the following equations:

S·nH₂O_(sp) + M²⁺_(aq.) = S·M²⁺_(sp), (Q_sor)

M²⁺_(aq.) + nH₂O = M²⁺·nH₂O_(aq.), (Q_dil)

S_(sp) + nH₂O = S·nH₂O_(sp), (Q_h)

S·nH₂O_(sp) + M²⁺·nH₂O_(aq.) = S·M²⁺_(sp) + 2nH₂O, (Q_tit)

To evaluate the energy associated with the cation sorption processes, it is necessary to subtract the values obtained from the thermal effect generated by the dilution of the metallic solution (Q_dil) and hydration of the solids (Q_h). These values were obtained in a separate titration in which aliquots of lead or cadmium solutions were incrementally added to the calorimetric vessel containing only deionized water in the absence of any material. The thermal effect was subtracted from the data obtained in the experiment, yielding the values related exclusively to the interactions of the pendant basic centers with the metallic ions (Q_sor). Thus, the resulting net thermal effect (Q_tit) of sorption was obtained by the following expression:

Q_tit = Q_sor − Q_dil − Q_h

Because the experimental thermal effect of hydration for the synthesized hybrids was null (Q_h = 0), this expression can be simplified to

Q_tit = Q_sor − Q_dil

This equation can also be described by the calorimetric titration curves obtained from the experiment of Pb²⁺ sorption on the PhNiI material, as displayed in Fig. 9.

Fig. 9 Calorimetric titration curve of 0.02012 g of PhNiI in 2.0 cm³ of deionized water with incrementally added aliquots (ΣV_ad) of 20.0 μdm³ of 0.10 mol dm⁻³ Pb(NO₃)₂ in the same solvent at 298.15 ± 0.20 K. The experimental points represent the sum of the thermal effects of titration (■), dilution (▲) and the net thermal effect (●).

Using the sum of the enthalpic values obtained through calorimetry, it was also possible to build isotherms that followed the behavior predicted by the Langmuir equation applied to calorimetry, as described by eqn (3). The non-linear regression of the Langmuir isotherm allows the determination of the values of Δ_monoH and Θ.⁴

where X corresponds to the molar fraction of the metallic ion in solution in equilibrium after each addition of the metallic solution, Δ_rH is the integral enthalpy of the reaction, Θ is a constant that includes the equilibrium constant and Δ_monoH represents the specific enthalpy of the interactive process to the formation of the cation monolayer.

The sorption Langmuir isotherm of lead ions by the phyllosilicate PhNiI (Fig. S6) is shown in ESI.† All of the sorption isotherms for the other synthesized solids with lead and cadmium cations presented the same pattern with different intensities of the enthalpic values.

As already mentioned, the Langmuir isotherm is fundamentally important for the acquisition of the thermodynamic results of ΔH, ΔG and ΔS, characteristic of the interactive processes.¹³ The Δ_monoH values obtained are used in the calculation of the molar enthalpy variation related to the sorption process. The relationship between Δ_monoH and ΔH is given by the expression described in eqn (4), in which N^s represents the number of moles necessary for the formation of a monolayer of cations on the surface, obtained by the regression of the sorption isotherms.

The molar enthalpy is defined as the thermal effect released from the system or furnished to the system at constant pressure due to the cation/basic center interactions. This variation is measured calorimetrically by following the variation in temperature that occurs in the calorimetric vessel. The enthalpy variation of the system indicates whether the processes are exothermic or endothermic in nature.⁵⁰

The evaluation of the spontaneity of the interactions that occur on the functionalized surfaces, as represented by ΔG obtained at constant temperature and pressure, is in agreement with the decrease in its values, as calculated using eqn (5), where T = 298.15 ± 0.20 K; R is the ideal gas constant, equal to 8.314 J K⁻¹ mol⁻¹; and b is the equilibrium constant of the sorption.⁵¹

ΔG = RT ln b (5)

The entropy of the system is a measure of the molecular disorder and the disordered dispersion of energy in a system that allows the determination of whether a state is accessible from another one through a spontaneous transformation.⁵⁰ This value can be calculated from ΔH and ΔG, as previously obtained, using eqn (6).⁵²

ΔG = ΔH − TΔS (6)

The thermodynamic data obtained from lead and cadmium sorption experiments are listed in Table 1.

Table 1 Thermodynamic results involving the interactions of organofunctionalized hybrids (Hyb) with lead and cadmium cations, given by enthalpy (ΔH), Gibbs energy (ΔG) and entropy (ΔS) at 298.15 ± 0.20 K, and the equilibrium constants of the sorption processes

Hyb	b		−ΔH/kJ mol^−1		−ΔG/kJ mol^−1		ΔS/J K^−1 mol^−1
	Pb^2+	Cd^2+	Pb^2+	Cd^2+	Pb^2+	Cd^2+	Pb^2+	Cd^2+
PhCoCl	1064 ± 19	1352 ± 32	5.73 ± 0.05	5.04 ± 0.03	17.3 ± 0.1	17.9 ± 0.2	39 ± 1	43 ± 3
PhCoI	1872 ± 29	4722 ± 30	6.96 ± 0.03	6.05 ± 0.04	18.7 ± 0.1	21.0 ± 0.1	39 ± 1	50 ± 2
PhNiCl	1275 ± 24	890 ± 29	5.23 ± 0.01	0.91 ± 0.01	17.7 ± 0.5	16.8 ± 0.5	42 ± 1	53 ± 2
PhNiI	1628 ± 44	484 ± 11	7.41 ± 0.01	1.69 ± 0.09	18.3 ± 0.1	15.3 ± 0.1	37 ± 1	46 ± 1

---

The results listed in Table 1 show that the enthalpic values are higher for cobalt and nickel phyllosilicates functionalized from the precursor silane where the original chlorine was substituted by iodine before the synthesis of the silylating agent. This substitution resulted in higher quantities of pendant basic centers promoted by the higher substitution rate of iodine for diethyl iminodiacetate relative to chlorine during the synthesis of the final silylating agent.

In general, the thermodynamic data for all of the materials studied, as obtained from calorimetry, presented negative values of molar enthalpy. These results suggest that the sorption processes of lead and cadmium cations in the solid–liquid interface on these hybrids' surfaces are exothermic, as expected for this type of interaction. There is a release of energy from the calorimetric vessel to the environment during the sorption processes in all systems.^43,13

The calculated Gibbs energy variations also presented negative values, indicating that the sorption processes on the surfaces of all hybrids are spontaneous, as expected due to the potentiality of the nitrogen basic centers present in the pendant organic chains for sorbing metallic cations from aqueous solutions.^4,9,52

Finally, the entropic results presented positive values, suggesting an increase in the disorder of the final system after the sorption process. These results could be justified by the increase in the release of water molecules to the system due to the dehydration of the metallic cations, which were surrounded by the water molecules that formed their coordination spheres before the formation of the cation/basic center complexes on the surface of the solids. In addition, some water molecules that were sorbed on the hybrids'surfaces by hydrogen bonds before the sorption processes were liberated to the environment.^4,13,51

Conclusions

Inorganic–organic cobalt and nickel phyllosilicate hybrids were successfully synthesized through a well-established synthetic route using new silylating agents formed by the individual reactions of 3-chloropropyltriethoxysilane and 3-iodopropyltriethoxysilane with diethyl iminodiacetate. The nucleophilic substitution of chlorine by iodine in the commercial 3-chloropropyltriethoxysilane yielded 3-iodopropyltriethoxisilane with 98% efficiency, which led to hybrids with a higher degree of functionalization, containing attached basic centers in the pendant chains. These new smectite-like structures containing organic groups with oxygen and nitrogen basic sites covalently attached to the inorganic framework favored toxic cation removal from aqueous solutions. The pendant organic chains presented a similar configuration to their precursor silanes, implying that the functionalization processes did not damage the organic structures. The thermodynamic data obtained from the calorimetric titration of the synthesized hybrids revealed an exothermic enthalpy, negative Gibbs energy and positive entropy variations for the interactions of basic centers present on the inorganic matrices with divalent lead and cadmium ions at the solid–liquid interface. These interactive effects are more pronounced for greater availability of basic centers on the surfaces.

Acknowledgements

The authors are indebted to FAPESP for financial support and the National Laboratory of Nanotechnology (LNNano) at National Center of Research on Energy and Materials (CNPEM) for the TEM images.

Notes and references

1. M. X. Tan, Y. N. Sum, J. Y. Ying and Y. Zhang, Energy Environ. Sci., 2013, 6, 3254–3259.
2. D. H. K. Reddy and S.-M. Lee, Adv. Colloid Interface Sci., 2013, 201–202, 68–93.
3. M. R. Awual, T. Yaita, S. A. El-Safty, H. Shiwaku, S. Suzuki and Y. Okamoto, Chem. Eng. J., 2013, 221, 322–330.
4. F. J. V. E. Oliveira, M. A. Melo Jr and C. Airoldi, Mater. Res. Bull., 2013, 48, 1045–1056.
5. S. Babel and T. A. Kurniawan, J. Hazard. Mater., 2003, 97, 219–243.
6. O. Yavuz, Y. Altunkaynak and F. Güzel, Water Res., 2003, 37, 948–952.
7. E. Erdem, N. Karapinar and R. Donat, J. Colloid Interface Sci., 2004, 280, 309–314.
8. B. Samiey, C.-H. Cheng and J. Wu, Materials, 2014, 7, 673–726.
9. M. A. Melo Jr, F. J. V. E. Oliveira, J. A. A. Sales and C. Airoldi, New J. Chem., 2009, 33, 1038–1046.
10. A. Walcarius and L. Mercier, J. Mater. Chem., 2010, 20, 4478–4511.
11. M. Monier, Int. J. Biol. Macromol., 2012, 50, 773–781.
12. Y.-C. Lee, E. J. Kim, D. A. Ko and J.-W. Yang, J. Hazard. Mater., 2011, 196, 101–108.
13. M. A. Melo Jr and C. Airoldi, Dalton Trans., 2010, 39, 10217–10227.
14. A. Bibby and L. Mercier, Chem. Mater., 2002, 14, 1591–1597.
15. D. A. de Sousa, E. de Oliveira, M. C. Nogueira and B. P. Espósito, Bioresour. Technol., 2010, 101, 138–143.
16. I. L. Lagadic, M. K. Mitchell and B. D. Payne, Environ. Sci. Technol., 2001, 35, 984–990.
17. J. Li, Y. Zhang, W. Cai and X. Shao, Talanta, 2011, 84, 679–683.
18. M. Jaber, J. Miehe-Brendle and R. Le Dred, J. Mater. Sci., 2004, 39, 1489–1490.
19. A. S. O. Moscofian, C. R. Silva and C. Airoldi, Microporous Mesoporous Mater., 2008, 107, 113–120.
20. H.-K. Han, Y.-C. Lee, M.-Y. Lee, A. J. Patil and H.-J. Shin, ACS Appl. Mater. Interfaces, 2011, 3, 2564–2572.
21. Y. Fukushima and M. Tani, J. Chem. Soc., Chem. Commun., 1995, 241–242.
22. K. K. R. Datta, A. Achari and M. Eswaramoorthy, J. Mater. Chem. A, 2013, 1, 6707–6718.
23. K. Chabrol, M. Gressier, N. Pebere, M.-J. Menu, F. Martin, J.-P. Bonino, C. Marichal and J. Brendle, J. Mater. Chem., 2010, 20, 9695–9706.
24. T. J. Pinnavaia, Science, 1983, 220, 365–371.
25. J.-C. Gallégo, M. Jaber, J. Miehé-Brendlé and C. Marichal, New J. Chem., 2008, 32, 407–412.
26. M. A. Melo Jr, F. J. V. E. Oliveira and C. Airoldi, Appl. Clay Sci., 2008, 42, 130–136.
27. C. D. Nunes, A. A. Valente, M. Pillinger, A. C. Fernandes, C. C. Romão, J. Rocha and I. S. Gonçalves, J. Mater. Chem., 2002, 12, 1735–1742.
28. R. J. P. Corriu, A. Mehdi, C. Reyé and C. Thieuleux, Chem. Mater., 2004, 16, 159–166.
29. F. Adam, H. Osman and K. M. Hello, J. Colloid Interface Sci., 2009, 331, 143–147.
30. J. J. Allen, E. Rosenberg, E. Johnston and C. Hart, ACS Appl. Mater. Interfaces, 2012, 4, 1573–1584.
31. J. Alauzun, A. Mehdi, C. Reyé and R. Corriu, New J. Chem., 2007, 31, 911–915.
32. I. M. El-Nahhal and N. M. El-Ashgar, J. Organomet. Chem., 2007, 692, 2861–2886.
33. I. Ahmed and R. V. Parish, J. Organomet. Chem., 1993, 452, 23–28.
34. Y. Fukushima and M. Tani, Bull. Chem. Soc. Jpn., 1996, 69, 3667–3671.
35. R. M. Silverstein, F. X. Webster and D. J. Kiemle, Spectrometric Identification of Organic Compounds, John Wiley & Sons, New York, 7th edn, 2005.
36. N. T. Whilton, S. L. Burkett and S. Mann, J. Mater. Chem., 1998, 8, 1927–1932.
37. A. S. O. Moscofian, C. T. G. V. M. T. Pires, A. P. Vieira and C. Airoldi, RSC Adv., 2012, 2, 3502–3511.
38. S. C. Holmström, A. J. Patil, M. Butler and S. Mann, J. Mater. Chem., 2007, 17, 3894–3900.
39. M. Jaber, J. Miehe-Brendle and R. Le Dred, Chem. Lett., 2002, 9, 954–955.
40. S. L. Burkett, A. Press and S. Mann, Chem. Mater., 1997, 9, 1071–1073.
41. L. Yang, S.-K. Choi, H.-J. Shin and H.-K. Han, Int. J. Nanomed., 2013, 8, 4147–4155.
42. I. M. El-Nahhal, N. M. El-Ashgar, M. M. Chehimi, P. Bargiela, J. Maquet, F. Babonneau and J. Livage, Microporous Mesoporous Mater., 2003, 65, 299–310.
43. F. J. V. E. Oliveira, E. C. da Silva Filho, M. A. Melo Jr and C. Airoldi, Surf. Sci., 2009, 603, 2200–2206.
44. G. Chandrasekaran, H.-K. Han, G.-J. Kim and H.-J. Shin, Appl. Clay Sci., 2011, 53, 729–736.
45. Y.-C. Lee, Y.-S. Choi, M. Choi, H. Yang, K. Liu and H.-J. Shin, Appl. Clay Sci., 2013, 83–84, 474–485.
46. A. Dumas, F. Martin, C. Le Roux, P. Micoud, S. Petit, E. Ferrage, J. Brendlé, O. Grauby and M. Greenhill-Hooper, Phys. Chem. Miner., 2013, 40, 361–373.
47. M. Jaber, J. Miéhé-Brendlé, L. Delmotte and R. Le Dred, Solid State Sci., 2005, 7, 610–615.
48. M. G. da Fonseca, C. R. Silva, J. S. Barone and C. Airoldi, J. Mater. Chem., 2000, 10, 789–795.
49. T. R. Macedo, G. C. Petrucelli, A. A. Pinto and C. Airoldi, Mater. Res. Bull., 2011, 46, 2346–2352.
50. J. A. A. Sales, A. G. S. Prado and C. Airoldi, Surf. Sci., 2005, 590, 51–62.
51. T. R. Macedo and C. Airoldi, Microporous Mesoporous Mater., 2006, 94, 81–88.
52. K. S. Sousa, E. C. Silva Filho and C. Airoldi, Carbohydr. Res., 2009, 344, 1716–1723.

---

Footnote

† Electronic supplementary information (ESI) available: Infrared spectrum of IPTS, IR as well as ¹³C and ²⁹Si solid-state NMR spectra of the hybrids, thermogravimetric curves of these solids and elemental analysis and sorption experimental data. See DOI: 10.1039/c4ra06615d

---