Mesoporous titanosilicate nanoparticles: facile preparation and application in heterogeneous epoxidation of cyclohexene

Abstract

Mesoporous titanosilicate nanoparticles with a size of 40 to 75 nm (Nano-Ti-MCM-41) were hydrothermally synthesized from a titanosilicate (TiSil) solution with cetyltrimethylammonium bromide (CTAB) as the template and with a cationic polymer as the size-controlling agent. The particle size is proposed to be controlled by adjusting the size of TiSil-CTA⁺ micelles influenced by the charge interaction between negatively charged titanosilicate and polymer cationic ions during the formation of TiSil-CTA⁺ micelles. The presence of n-hexane and hydrogen peroxide in the preparation process proved to favor a well-ordered mesostructure in these nanoparticles: the mesostructure became disordered without using hexane and hydrogen peroxide although the size of mesoporous titanosilicate particles remained in the nanometer scale. The characterization results showed that Nano-Ti-MCM-41 possesses hierarchical porosity (bimodal mesopores) and an ordered mesoporous structure and that the titanium species are predominately located in tetrahedral framework positions. Nano-Ti-MCM-41 is a highly active catalyst for the epoxidation of cyclohexene with hydrogen peroxide, and displayed higher turnover numbers (TONs) based on the cyclohexene conversions and higher selectivity ratio between cyclohexene epoxide and 1,2-cyclohexanediol compared with traditional Ti-MCM-41 prepared without the cationic polymer. The improved catalytic performances are mainly ascribed to the decrease in the particle size of Ti-MCM-41, resulting in the enhanced accessibility of the reactants to the catalytic Ti species via the shorter channels of Nano-Ti-MCM-41 and in the shorter residence time of cyclohexene epoxide in the mesopores of the nanoparticles. Importantly, the mesoporous titanosilicate nanoparticles are stable catalysts immune to titanium leaching, suggesting their good recyclability potential in the epoxidation with aqueous H₂O₂.

---

1 Introduction

Since the discovery of the first mesostructured silicates composed of ordered hexagonally arranged cylindrical mesopores as micrometer-sized and amorphous aggregates by Mobil researchers,¹ the syntheses of ordered mesoporous silicate materials (OMSMs) with various compositions have been sought for application in adsorption and separation, drug storage and release, sensors, nano-scaled electronic devices, and heterogeneous reactions involving bulky molecules.^2–7 It has been revealed that the versatile practical properties of OMSMs are strongly related to their morphologies, as well as extremely high surface areas, high pore volumes, narrow pore size distributions, and ordered pore systems.^8–10 In recent decades, considerable research effort has been devoted globally to the development of novel OMSMs with designed morphology including microspheres, films, fibers, hollow spheres, hollow tubes, gyroids, rods, monoliths, helices, vesicles and nanoparticles,^11–26 in which ordered mesoporous silicate nanoparticles (OMSNs) proved to exhibit more advantages in mass transport and have attracted more and more attentions. Minimization of particle size to the nanometer range for intracellular delivery is considered to be critical in their potential applications as nano-scaled reservoir in drug and biomolecule delivery, as nano-scaled devices in biological sensors and solar cells, and as nano-reactors in heterogeneous catalysis.

The approaches reported so far on preparation of OMSNs are mainly focused on pure silicas. For example, Cai et al. synthesized MCM-41 nanoparticles with an average size of 110 nm in the sodium hydroxide medium;²⁷ Mou and Ostafin et al. individually reported that mesoporous MCM-41 nanoparticles could be prepared using an ammonia base-catalyzed method.^28,29 These results indicate that the dilute and low surfactant conditions are efficient for the preparation of MCM-41 nanoparticles, which is dependent on the length of silicate-CTA micelles. Additionally, the preparation of nanoscaled particles with a diameter of 20–50 nm was successfully achieved by suppression of the grain growth with the presence of a nonionic surfactant.³⁰ Using template-containing silica microspheres as a precursor, ordered mesoporous silica nanospheres could be synthesized by various hydrothermal merging processes. For the reason that these approaches are limited in pure silica mesoporous nanoparticles, most of the studies on OMSNs have been focusing in their biological and electrical applications. Up to now, the research of OMSNs containing active sites in the framework on the application as nano-reactors in heterogeneous catalysis is relatively rare. The limited studies on application of mesoporous nanoparticles in heterogeneous catalysis started from the first example of disordered mesoporous aluminosilicate nanoparticles, which was used in the cracking of high molecular weight hydrocarbons.³¹ The remarkable catalytic reactivity of these nanoparticles was attributable to a unique combination of mesoporosity, high hydrothermal stability and acidity. Other examples rely on the discovery of Ti-MCM-41 and Fe-MCM-41 nanoparticles prepared in the medium of sodium hydroxide via the dilute solution route at room temperature, showing improved catalytic performances in the epoxidation of alkene and in hydroxylation of phenol.^32,33 The accessibility of the catalytic species was proposed to be enhanced in the shorter mesopore channels in these nanoparticles, which is supported by the limited diffusion and transport of reactants on mesoporous Ti-MCM-41 micrometer-sized particles.³⁴ In that work, the epoxidation of PBD-bodipy with tert-butyl hydrogen peroxide (TBHP) on Ti-MCM-41 micrometer-sized particles was in situ monitored by high-resolution single-turnover mapping, indicating that only the titanium sites in the outer 300 nm of the particle are responsible for the observed catalytic activity.³⁴ The study clearly revealed that the intraparticle diffusion must be one of key factors for the relatively lower oxidation ability of mesoporous micrometer-sized particles than that of nanosized titanosilicate zeolites. Although Ti-MCM-41 nanoparticles have been discovered, their broader application in catalytic oxidations is somewhat limited due to the incomplete condensation of the titanosilicate framework due to the room-temperature approach.³² Therefore, it is of great interest to develop improved mesoporous titanosilicate by decreasing the particle size of Ti-MCM-41 into the nanoscale via hydrothermal method.

More recently, cationic polymers have been applied to prepare hierarchically porous TS-1 and beta zeolites.^35–37 The successful preparation of such materials strongly relies on controlling the growth of the zeolite TS-1 and beta particles with the cationic polymer, which points to the existence of strong charge interaction between the cationic polymer and silicate species in the alkaline medium. Herein, Ti-MCM-41 nanoparticles were first prepared, to our knowledge, with a cationic polymer as the size-controlled agent via hydrothermal merging process. The obtained nanoparticles were used to catalyze the epoxidation of cyclohexene with H₂O₂, which is frequently used as a test reaction for the catalytic evaluation of titanosilicate catalysts.

2 Experimental

2.1 Synthesis of mesoporous titanosilicate nanoparticles

In a typical synthesis of mesoporous titanosilicate nanoparticles, a mixture was prepared by mixing 0.55 g of cetyltrimethyl ammonium bromide (CTAB), 12.5 ml of H₂O, 6 ml of ammonium hydroxide aqueous solution (20–25%), and polydiallyldimethyl ammonium chloride (PDADMAC, 3 g, 40 wt%, molecular weight ∼ 1.5 × 10⁵). After stirring for 0.25 h at room temperature, 0.1 g of tetrabutyl orthotitanate (TBOT) and 2.5 ml of tetraethyl orthosilicate (TEOS) were added to the mixture under stirring at 800–1000 rpm, yielding a white gel. The white gel was subsequently stirred at room temperature for 4 h and then was transferred into autoclave for hydrothermal treatment at 100 °C for 48 h. The product was collected by filtration, dried in air and calcined at 550 °C for 6 h for removal of the template CTAB and the cationic polymer PDADMAC. The mesoporous titanosilicate nanoparticles are designated as Nano-MesoTS(x), where x is denoted as the mass ratio of PDADMAC to CTAB. Under the given preparation conditions for Nano-MesoTS(1), n-hexane and aqueous H₂O₂ (30%) were added to the preparation process, obtaining a novel sample of mesoporous titanosilicate nanoparticles (Nano-Ti-MCM-41), in which molar ratio between hexane and CTAB is 2 and the value between H₂O₂ and TBOT is 238. For comparison, other titanosilicate samples were prepared as the same procedure used for Nano-Ti-MCM-41, but without using hydrogen peroxide (C-hex-Nano-Ti-MCM-41) or hexane (C-hp-Nano-Ti-MCM-41).

2.2 Materials characterization

X-ray diffraction (XRD) measurements were performed on a Bruker D8 Advance instrument with Cu Kα radiation. UV-Visible spectra were measured with a PE Lambda 750 spectrophotometer in the region of 200–800 nm. The samples were not treated before measurement and the UV-Visible spectra were recorded in air. The isotherms of nitrogen were measured at the temperature of liquid nitrogen using a Micromeritics TriStar II 3020. The pore-size distribution was calculated using the Barrett–Joyner–Halenda (BJH) model. Scanning electron microscopy (SEM) images were taken on a Hitachi SU8010 apparatus and transmission electron microscopy (TEM) images were taken on a Tecnai F30 electron microscope. The particle sizes of Nano-MesoTS and Nano-Ti-MCM-41 were mainly determined by mean of SEM: each sample was measured on more than 30 randomly selected particles and the average of the results was calculated. Si and Ti elemental analysis was performed by inductively coupled plasma-optical emission spectroscopy (ICP-OES) on a PerkinElmer Optima 7300DV.

2.3 Catalytic tests

The epoxidation of cyclohexene was carried out in capped glass vials for 5 h at 60 °C under magnetic stirring (500 rpm). For the epoxidation with aqueous H₂O₂, catalyst (50 mg) were mixed with H₂O₂ (35%, 2.45 mmol), cyclohexene (4.9 mmol) and acetonitrile (4.5 ml). All the tested samples were fleshly calcined. The catalysts were separated by centrifugation and the products were analyzed with a gas chromatograph (GC) SP-6800A equipped with a SE-54 capillary column (30 m, 0.32 mm) and a FID detector. Octane was used as internal standard.

The assignment of these products were confirmed by gas chromatography-mass spectrometry analysis (GC-MS) on an Agilent 7890A gas chromatograph (HP-5MS column, 30 m, 0.25 mm) coupled to an Agilent 5975 MSD mass spectrometer.

3 Results and discussion

3.1 Characterization of double-shelled hollow mesoporous silica spheres

The morphology and particle size of the calcined Nano-MesoTS and Ti-MCM-41 are revealed by SEM images. Whereas Ti-MCM-41 clearly contain bulky particle in micrometer scale (Fig. 1A), Nano-MesoTS (Fig. 1B–F) are composed of nanosized agglomerated spherical particles. The utilization of cationic polymer PDADMAC leads to an obvious decrease in the particle size of Ti-MCM-41; the size of Nano-MesoTS slightly decreases with the increase of the amount of PDADMAC used for the preparation, which is in agreement with the observation on mesoporous beta zeolites described elsewhere.³⁶ The average size of the particles is ca. 83 and 79 nm for Nano-MesoTS(1) and Nano-MesoTS(1.8), respectively. Note that the agglomeration in the cases of Nano-MesoTS(0.18) and Nano-MesoTS(4.5) is more pronounced.

Fig. 1 SEM images of Ti-MCM-41 (A), Nano-MesoTS(0.18) (B), Nano-MesoTS(1) (C), Nano-MesoTS(1.8) (D), Nano-MesoTS(2.7) (E), and Nano-MesoTS(4.5) (F).

The XRD patterns of Ti-MCM-41 and Nano-MesoTS(0.18) samples exhibit reflections corresponding to the (100), (110) and (200) planes,^{1,25,26,32,33} showing an ordered 2D-hexagonal (p6mm) arrangement of channels (Fig. 2A and B). When the mass ratio of PDADMAC to CTAB is more than 1, disordered structures are obtained, which indicates that the cationic polymer would have a significant influence on the mesostructural order of Ti-MCM-41. It has also been mentioned in the preparation of mesoporous beta zeolites that large amounts of added polymer can result in the reduction of zeolite crystallinity.³⁶ These findings clearly point to strong charge interaction between the cationic polymer and silicate species in the formation process of porous silicates in the alkaline medium.

Fig. 2 XRD patterns of Ti-MCM-41 (A), Nano-MesoTS(0.18) (B), Nano-MesoTS(1) (C), Nano-MesoTS(1.8) (D), Nano-MesoTS(2.7) (E), and Nano-MesoTS(4.5) (F).

N₂ adsorption/desorption isotherms of calcined Ti-MCM-41 and Nano-MesoTS(0.18) samples both give typical type-IV isotherms with a sharp inflection at p/p₀ > 0.3 (Fig. 3). This is characteristic of capillary condensation, which points to the uniformity of the mesopore size distribution.^25,26 It is worth noting that N₂ adsorption/desorption isotherms of Nano-MesoTS(1), Nano-MesoTS(1.8), Nano-MesoTS(2.7) and Nano-MesoTS(4.5) have an obvious hysteresis loop above p/p₀ = 0.9 in addition to a less pronounced inflection at p/p₀ > 0.3 (Fig. 3). Their pore size distribution curves display the coexistence of uniform mesopores with the size at around 3.0 nm and the larger mesopores with the size at around 9.6–74.4 nm. The former pores stem from the arrangement of micelles of the template CTAB and the larger mesopores from the interparticle spaces among the nanoparticles, as observed on other SBA-15 and MCM-41 nanoparticles.³⁰ The size of larger mesopores decreases with the increase of PDADMAC used, suggesting the cationic polymer PDADMAC have a big influence on the size of these nanoparticles, in agreement with the results from SEM. Table 1 lists the textural properties of Ti-MCM-41 and Nano-MesoTS samples. Not as expected, the BET surface area of Ti-MCM-41 is a little higher than that of Nano-MesoTS, which may be related with the larger pore wall thickness (Table S1†), disordered mesopores and/or the agglomeration of the resulting nanoparticles. By considering the same amount of the template CTAB used for preparation of various nanoparticles, the highest pore volume of Nano-MesoTS(1) might imply the larger pore volume of mesopores stemming from the interparticle spaces among the nanoparticles.

Fig. 3 Nitrogen adsorption (+)/desorption (×) isotherms (a) and pore size distritution curves (b) of Nano-MesoTS(1) (A), Nano-MesoTS(0.18) (B), Ti-MCM-41 (C), Nano-MesoTS(1.8) (D), Nano-MesoTS(2.7) (E), and Nano-MesoTS(4.5) (F).

Table 1 Textural properties of calcined Ti-MCM-41, Nano-MesoTS and Nano-Ti-MCM-41 materials^a

Sample	Si/Ti^b (molar)	Pore size (nm)	Pore volume (cm^3 g^−1)	Surface area (m^2 g^−1)
a Pore-size distribution and pore volume were determined from N2 adsorption isotherms at 77 K.b Determined by ICP.
Ti-MCM-41	45.4	3.3	0.89	817.1
Nano-MesoTS(0.18)	95.3	2.9; 74.4	0.92	734.5
Nano-MesoTS(1)	84.6	2.9; 33.0	1.22	428.7
Nano-MesoTS(1.8)	76.2	3.0; 15.4	0.71	437.8
Nano-MesoTS(2.7)	72.7	2.7; 11.5	0.86	559.7
Nano-MesoTS(4.5)	94.8	2.7; 9.6	0.44	328.2
Nano-Ti-MCM-41	77.4	3.0; 32.6	1.28	652.7

---

Though the addition of cationic polymer PDADMAC results in the formation of mesoporous nanoparticles, the mesoporous structure becomes disordered with the increase of PDADMAC added. This might be attributed to the destruction of titanosilicate-CTA⁺ micelles originating from strong charge interaction between CTA⁺ and PDADMA⁺ ions and/or from the nonuniform hydrolysis of titanium and silicon alkoxides. In our study, n-hexane and aqueous hydrogen peroxide (30%, denoted as HP) were added to the above preparation systems, aiming to obtain Ti-MCM-41 nanoparticles with ordered mesopores. Nano-Ti-MCM-41 can be described as mesoporous titanosilicate nanoparticles with an ordered 2D-hexagonal (p6mm) arrangement of mesopores, and a hierarchical porosity that includes ordered mesopores from micelles and interparticle space. This description stems from the full characterization of the material. From the XRD patterns (Fig. 4a-A), hexagonal mesopores proved to be shown in Nano-Ti-MCM-41 prepared with n-hexane and aqueous HP. N₂ adsorption/desorption isotherms of Nano-Ti-MCM-41 give a typical type-IV isotherm with a sharp inflection at low pressure and with an additional hysteresis loop at high pressure (Fig. 4b). Its pore size distribution curve shows the coexistence of uniform mesopores with the size at ca. 3.0 nm and the larger mesopores with the size at ca. 32.6 nm.

Fig. 4 XRD patterns (a) of Nano-Ti-MCM-41 (A), C-hp-Nano-Ti-MCM-41 (B), and C-hex-Nano-Ti-MCM-41 (C) and the nitrogen adsorption (+)/desorption (×) isotherms (b) of Nano-Ti-MCM-41 (A) and the relevant pore size distribution curves (inset).

The morphology, particle size and pore architecture of calcined Nano-Ti-MCM-41 are further revealed by SEM and TEM images. The material proved to be composed of nanosized elongated spherical particles and the size of the particles ranges from 40 to 75 nm (Fig. 5). TEM images of Nano-Ti-MCM-41 show the existence of ordered hexagonal arrays and one-dimensional mesoporous parallel channels within these nanoparticles (Fig. 5). The TEM results confirm that the Ti-MCM-41 nanoparticles are pure phases with short and ordered mesoporous channels. It is worth noting that the pore size and particle size of Nano-Ti-MCM-41 are similar with those of Nano-MesoTS(1) and that the BET surface area is however higher, suggesting the ordered mesopores are in favor of higher surface area.

Fig. 5 SEM and TEM images of Nano-Ti-MCM-41. Characteristic interparticle mesopores and ordered mesopores from surfactant micelles are indicated by the arrows and ellipses, respectively.

Based on the XRD results, it is also observed that the mesostructure is less ordered without using hexane and that it becomes completely disordered without using hydrogen peroxide (Fig. 4a-B and C). When only n-hexene was applied for the preparation, disordered mesopores were still achieved as those in Nano-MesoTS(1) although higher amount of n-hexane was used in the preparation (Fig. S1†). When hydrogen peroxide was present in the preparation, the mesopore ordering was highly improved regardless of amount of n-hexane applied (Fig. S1†). Clearly, the order of mesopores in Nano-Ti-MCM-41 could be affected by the combination of hydrogen peroxide and hexane and hydrogen peroxide plays more important role in formation of ordered mesostructure compared with n-hexane. It has been reported that n-hexane, as a typical micelle expander, significantly increase the volume of the micelles, leading to a marked pore volume increase for the resulting templated silicas.³⁸ In our work, however, the pore volume and pore size of Nano-Ti-MCM-41 (see Table 1) are comparable as those of the sample Nano-MesoTS(1) synthesized in the absence of hexane. This finding suggests that hexane solubilizes in the micelles only to a small extent, as already demonstrated by Nagarajan in the work on solubilization of hydrocarbons in micellar solutions.³⁹ Thus, in the case of Nano-Ti-MCM-41 synthesis, hexane appears to act not as a typical micelle swelling agent located in the center of micelles, but as an additive for the preparation solution. Some of n-hexane molecules as well as the additive hydrogen peroxide molecules are proposed to be located between PDADMA⁺ ions and titanosilicate-CTA⁺ micelles interface. Such a less well-defined position of hydrogen peroxide and hexane molecules in the interface is considered to reduce the charge interaction between CTA⁺ and PDADMA⁺ ions, which may account for an ordered mesostructure in the resulting Nano-Ti-MCM-41 during the formation and cooperative assembly of the micelles in the process of hydrothermal treatment. On the other hand, hydrogen peroxide has been proven as a reagent that might control the hydrolysis of titanium alkoxide,⁴⁰ which could greatly favor to get a homogeneous titanosilicate framework through the uniform hydrolysis and condensation of titanium and silicon alkoxides.

UV/Vis spectroscopy is generally recognized as one of the acceptable methods to characterize the coordination of Ti species in titanosilicate zeolites and mesoporous materials.^41–43 The UV/Vis spectra of both Ti-MCM-41 and Nano-MesoTS exhibit a band centered at ca. 225–235 nm with similar intensity (Fig. 6), which can be attributed to distorted tetrahedral Ti species in amorphous titanosilicates.^44,45 Additionally, a small amount of higher-coordinated Ti species coexist with the tetrahedral Ti sites in the materials, as indicated by the shoulder between 270 and 290 nm. This is assigned to reversible penta-coordinated Ti species that form from the interaction of Ti species with moisture and/or to polymerized hexa-coordinated Ti species.⁴⁴ It is worth noting that the shoulder band at 270–290 nm in Nano-MesoTS is more pronounced than in Ti-MCM-41, suggesting that the use of cationic polymer PDADMAC has an influence on the incorporation of Ti species into the silica framework. Moreover, the larger molar ratio of Si to Ti in Nano-MesoTS than in Ti-MCM-41 (Table 1) could also be a sign that the cationic polymer has an influence on the hydrolysis of titanium alkoxide and/or the condensation of titanium and silicon species. Compared with Ti-MCM-41 and Nano-MesoTS samples, Nano-Ti-MCM-41 exhibits an isolated band centered at 214 nm, showing that the Ti species in Nano-Ti-MCM-41 is dominantly tetrahedral and that no penta-coordinated or polymerized hexa-coordinated Ti species is present.^25,26 Note that, when n-hexane is absent for the preparation, the obtained titanosilicate gives a very similar band as Nano-Ti-MCM-41. This suggests the isolated and tetrahedral Ti species in Nano-Ti-MCM-41 is attributed to the utilization of hydrogen peroxide, which has been proven as a reagent that might control the hydrolysis of titanium alkoxide. Also, the presence of isolated and tetrahedral Ti species confirms the uniform hydrolysis and condensation of titanium and silicon alkoxides, as discussed above.

Fig. 6 UV-Vis spectra of Ti-MCM-41 (A), Nano-MesoTS(0.18) (B), Nano-MesoTS(1) (C), Nano-MesoTS(1.8) (D), Nano-MesoTS(2.7) (E), and Nano-MesoTS(4.5) (F), Nano-Ti-MCM-41 (G) and C-hp-Nano-Ti-MCM-41 (H).

On the basis of the characterization results, it is possible to propose a path for the formation of mesoporous titanosilicate nanoparticles by the means of controlling by a cationic polymer. Under basic condition, negatively charged titanosilicate species and the surfactant (CTAB) of opposite charges form the strong interaction (S⁺–I⁻) in favor of highly ordered titanosilicate-CTA⁺ micelles.⁴ In the traditional preparation systems, bulky irregular particles of mesoporous titanosilicates are formed via the cooperative assembly of the micelles after hydrothermal treatment. In the presence of cationic polymer PDADMAC, the size of silicate-CTA⁺ micelles is controlled by charge interaction between negatively charged titanosilicate and PDADMA⁺ ions during the formation of silicate-CTA⁺ micelles (Scheme S1†), as proven by SEM and XRD results. When hexane and hydrogen peroxide are added in the synthesis system, some of the additive molecules may be located between PDADMA⁺ ions and titanosilicate-CTA⁺ micelles interface. The titanosilicate-CTAB composites evolve through condensation of titanosilicate species but controlled by the surrounding cationic polymer, which takes place during the subsequent hydrothermal treatment (Scheme S1†). In the next step, CTAB, PDADMAC and hexane molecules are removed by calcination, generating the resulting Nano-Ti-MCM-41 with ordered mesostructure. Note that, the agglomeration of the nanoparticles somewhat occur in the process of hydrothermal treatment and/or calcination.

3.2 Catalytic performances

The epoxidation of cyclohexene with aqueous H₂O₂ was chosen as a test reaction because this substrate is too large to diffuse into the pores of TS-1,²⁵ but may be converted on the Ti sites in the hierarchically porous structure of Nano-MesoTS and Nano-Ti-MCM-41 samples. Over the catalysts tested, four different products of the oxidation of cyclohexene with aqueous H₂O₂ were detected: cyclohexene epoxide (CHE), 1,2-cyclohexanediol (CHD), 2-cyclohexene-1-ol (CH-OH), and 2-cyclohexene-1-one (CH-ONE). It has been reported that CHE can convert to CHD by the hydrolysis of the epoxide in presence of acid sites and H₂O₂, which is considered to be one of reasons for low yield of the epoxide on most of mesoporous titanosilicates.⁴⁶ Therefore, the average selectivity ratio between CHE and CHD is generally used to evaluate the influence of catalyst structure on the hydrolysis reaction.³²

Although Nano-MesoTS displayed slightly lower conversions of cyclohexene than Ti-MCM-41 due to the absence of ordered mesopores and to the relatively smaller surface area, higher turnover numbers (TON) on MesoTS NP(1), MesoTS NP(1.8) and MesoTS NP(2.7) were obtained (Fig. 7), where TON were calculated based on moles of cyclohexene converted per mole Ti active site of the catalysts. It is also observed that the average selectivity ratio between CHE and CHD on MesoTS NP(1), MesoTS NP(1.8) and MesoTS NP(2.7), is more than double that on Ti-MCM-41 (Fig. 7), suggesting the hydrolysis of CHE with H₂O was less favorable on Nano-MesoTS. As expected, when Nano-Ti-MCM-41 with ordered mesopores was used as the epoxidation catalyst, conversion of cylcohexene on it is comparable with that on Ti-MCM-41; TON value and the average selectivity ratio between CHE and CHD on Nano-Ti-MCM-41 are higher than those on Nano-MesoTS and on Ti-MCM-41 (Fig. 7). These results confirm the advantages of mesoporous titanosilicate nanoparticles in epoxidation of cyclohexene with H₂O₂. Both the higher TON value based on conversion of cyclohexene and higher average selective ratio between CHE and CHD on Nano-MesoTS(1, 1.8 and 2.7) and Nano-Ti-MCM-41 should be attributed to the decrease in the particle size of Ti-MCM-41 to nanometer scale, which results in the enhanced accessibility of the reactants to the catalytic Ti species via the shorter channels ofNano-MesoTS and Nano-Ti-MCM-41 and in the shorter residence time of cyclohexene epoxide in the pores of the nanoparticles. It is worth noting that an opposite trend was observed in the cases of MesoTS NP(0.18) and MesoTS NP(4.5), i.e. the TON values with MesoTS NP(0.18) and MesoTS NP(4.5) are slightly lower than that with Ti-MCM-41 bulky particles. This result is probably ascribed to the presence of more pronounced agglomeration in the cases, which were evidenced by the SEM images (Fig. 1).

Fig. 7 Catalytic conversions, selective ratios between CHE and CHD, and TONs of cyclohexene with aqueous H₂O₂ on mesoporous titanosilicate nanoparticles. Samples 1–7: Ti-MCM-41, Nano-MesoTS(0.18), Nano-MesoTS(1), Nano-MesoTS(1.8), Nano-MesoTS(2.7), Nano-MesoTS(4.5), and Nano-Ti-MCM-41.

Besides, the selectivity ratio between CHE and CHD should be also described by taking account surface acidity of the catalysts. It has been proposed that H₂O₂ causes the formation of acid sites in the residual surface silanols hydrogen bonded to titanium hydroperoxide surface species in titanosilicates. These sites can play an important role in the hydrolysis of cyclohexene epoxide to diol. That is, the silanol concentration in these titanosilicates is highly related with the selectivities. Consequently, the fraction of Si atoms in the form of SiOH groups in Ti-MCM-41 and Nano-Ti-MCM-41 was determined by ²⁹Si MAS NMR spectroscopy (Fig. S2†). Either of the two materials exhibits three peaks at δ = −110, −100, and −90 ppm, which are attributed to Si(OSi)₄ (Q⁴), HOSi(OSi)₃ (Q³), and (HO)₂Si(OSi)₂ (Q²), respectively. Ti-MCM-41 and Nano-Ti-MCM-41 gave the percentage of SiOH groups on all Si atoms (SiOH content) at 51.9% and 48.2%, respectively. Clearly, it follows that the SiOH content was slightly lower in Nano-Ti-MCM-41. The lower number of silanols leads to a decrease in surface acidity, and, therefore, to less hydrolysis of cyclohexene epoxide. As a result, the epoxide selectivity should be enhanced, which is in good agreement with the catalytic results.

Leaching of titanium species from the catalysts into reaction solution in the presence of hydrogen peroxide is a major drawback of some mesoporous titanosilicates.⁴⁷ Since a small amount of leached Ti species can have a significant effect on the catalytic activity, in our work, a test for Ti leaching on Nano-Ti-MCM-41 was carried out according to the description elsewhere.⁴⁸ The catalyst was filtered from the reaction media after 2 h at the reaction temperature, and the filtrate was divided into two parts: one was immediately analyzed by GC and the other was allowed to react for a further 3 h. The conversion of cyclohexene in the two parts of the filtrate was 9.9 and 11.0%, respectively, which reveals that after removal of the catalyst negligible cyclohexene epoxidation took place by taking into account self-oxidation of cyclohexene with H₂O₂. This result shows that Ti species do not leach from Nano-Ti-MCM-41 during the epoxidation with H₂O₂, which is in accordance with the result that no Ti⁴⁺ was found in the filtrate determined using ICP analysis. The high stability of titanium species in the presence of H₂O₂ strongly suggests good recyclability potential of Nano-Ti-MCM-41 in the epoxidation of cyclohexene with H₂O₂. In order to confirm the high stability of the titanium species, the reusability of Nano-Ti-MCM-41 in consecutive runs was evaluated following the recycling approach via calcination at 550 °C for 4 h in air. Because of the fact that the deposition of reaction residues in the mesopores of Nano-Ti-MCM-41 can be removed via calcination approach and that Ti species do not leach from Nano-Ti-MCM-41 during the epoxidation of cyclohexene with H₂O₂, the recycled Nano-Ti-MCM-41 gave comparable conversion (15.2%) as that of the fresh one (18.7%).

4 Conclusions

The nanosized particles of mesoporous titanosilicates were successfully prepared when a cationic polymer PDADMAC was used as a size-controlled agent. Full characterization of these nanoparticles indicates their high surface area, pore volume, hierarchical porosity and the incorporation of titanium species in the framework. In the preparation process, PDADMAC was considered to have a significant influence on the mesostructural order of the nanopartciles by the charge interaction between CTA⁺ and PDDAMA⁺ ions during the formation of titanosilicate-CTA⁺ micelles under such preparation conditions. The addition of hexane and hydrogen peroxide favors the formation of 2D hexagonal arrays and one-dimensional mesoporous parallel channels and isolated and tetrahedral titanium species in the framework. It is proposed that some of the additive n-hexane molecules may be located in the interface between polymer cations and titanosilicate-CTA⁺ micelles, which would lead to weaker charge interaction between CTA⁺ and PDADMA⁺ ions. This may account for an ordered mesostructure in Nano-Ti-MCM-41 combined with cooperative assembly influenced by the controlled hydrolysis and condensation of titanium and silicon alkoxides.

The mesoporous titanosilicate nanoparticles were tested in the epoxidation of cyclohexene with aqueous H₂O₂ to evaluate their advantages in heterogeneous oxidations. Compared with the traditional Ti-MCM-41 bulky particles, both higher TONs based on conversion of cyclohexene and average selective ratio between CHE and CHD were observed on Nano-MesoTS and Nano-Ti-MCM-41. This result should be mainly attributed to the enhanced accessibility of the reactants to the catalytic Ti species via the shorter channels of Nano-Ti-MCM-41 and to the shorter residence time of cyclohexene epoxide in the pores of the nanoparticles. Another advantage of these nanoparticles in catalytic application lies on its high stability of titanium species in the heterogeneous oxidation with aqueous H₂O₂, which is very attractive for possible industrial applications.

Acknowledgements

This work was supported by National Natural Science Foundation of China (51472062, 21303031), Fundamental Research Funds for the Central Universities (HIT. IBRSEM. 201326) and Program for Science and Technology Innovation Talent in Harbin (2013RFQXJ004).

References

1. C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli and J. S. Beck, Nature, 1992, 359, 710–712.
2. J. Y. Ying, C. P. Mehnert and M. S. Wong, Angew. Chem., Int. Ed., 1999, 38, 56–77.
3. M. E. Davis, Nature, 2002, 417, 813–821.
4. Q. S. Huo, D. I. Margolese and G. D. Stucky, Chem. Mater., 1996, 8, 1147–1160.
5. P. T. Tanev and T. J. Pinnavaia, Science, 1995, 267, 865–867.
6. Y. Ye, C. Jo, I. Jeong and J. Lee, Nanoscale, 2013, 5, 4584–4605.
7. A. Corma, Chem. Rev., 1997, 97, 2373–2420.
8. J. Fan, J. Lei, L. M. Wang, C. Z. Yu, B. Tu and D. Y. Zhao, Chem. Commun., 2003, 2140–2141.
9. A. Katiyar and N. G. Pinto, Small, 2006, 2, 644–648.
10. J. M. Sun, H. Zhang, R. J. Tian, D. Ma, X. H. Bao, D. S. Su and H. F. Zou, Chem. Commun., 2006, 1322–1324.
11. Q. Sun, P. J. Kooyman, J. G. Grossmann, P. H. H. Bomans, P. M. Frederik, P. C. M. M. Magusin, T. P. M. Beelen, R. A. van Santen and N. A. J. M. Sommerdijk, Adv. Mater., 2003, 15, 1097–1100.
12. J. W. Wang, Y. D. Xia, W. X. Wang, R. Mokaya and M. Poliakoff, Chem. Commun., 2005, 210–212.
13. C. E. Fowler, D. Khushalani and S. Mann, Chem. Commun., 2001, 2028–2029.
14. H. P. Lin, Y. R. Cheng and C. Y. Mou, Chem. Mater., 1998, 10, 3772–3776.
15. Y. S. Li, J. L. Shi, Z. L. Hua, H. R. Chen, M. L. Ruan and D. S. Yan, Nano Lett., 2003, 3, 609–612.
16. H. Djojoputro, X. F. Zhou, S. Z. Qiao, L. Z. Wang, C. Z. Yu and G. Q. Lu, J. Am. Chem. Soc., 2006, 128, 6320–6321.
17. H. P. Lin and C. Y. Mou, Science, 1996, 273, 765–768.
18. H. Yang, N. Coombs and G. A. Ozin, Nature, 1997, 386, 692–695.
19. G. A. Ozin, H. Yang, I. Sokolov and N. Coombs, Adv. Mater., 1997, 9, 662–667.
20. H. Zhang, J. M. Sun, D. Ma, X. H. Bao, A. Klein-Hoffmann, G. Weinberg, D. S. Su and R. Schlogl, J. Am. Chem. Soc., 2004, 126, 7440–7441.
21. J. Zhang, M. Liu, A. Zhang, K. Lin, C. Song and X. Guo, Solid State Sci., 2010, 12, 267–273.
22. H. N. Wang, X. F. Zhou, M. H. Yu, Y. H. Wang, L. Han, J. Zhang, P. Yuan, G. Auchterlonie, J. Zou and C. Z. Yu, J. Am. Chem. Soc., 2006, 128, 15992–15993.
23. J. Shah and T. J. Pinnavaia, Chem. Commun., 2005, 1598–1600.
24. P. T. Tanev, M. Chibwe and T. J. Pinnavaia, Nature, 1994, 368, 321–323.
25. T. Blasco, A. Corma, M. T. Navarro and J. Pérez Pariente, J. Catal., 1995, 156, 65–74.
26. K. A. Koyano and T. Tatsumi, Chem. Commun., 1996, 145–146.
27. Q. Cai, Z. S. Luo, W. Q. Pang, Y. W. Fan, X. H. Chen and F. Z. Cui, Chem. Mater., 2001, 13, 258–263.
28. Y. S. Lin, C. P. Tsai, H. Y. Huang, C. T. Kuo, Y. Hung, D. M. Huang, Y. C. Chen and C. Y. Mou, Chem. Mater., 2005, 17, 4570–4573.
29. R. I. Nooney, D. Thirunavukkarasu, Y. Chen, R. Josephs and A. E. Ostafin, Langmuir, 2003, 19, 7628–7637.
30. K. Suzuki, K. Ikari and H. Imai, J. Am. Chem. Soc., 2004, 126, 462–463.
31. Y. Liu and T. J. Pinnavaia, J. Am. Chem. Soc., 2003, 125, 2376–2377.
32. K. Lin, P. P. Pescarmona, H. Vandepitte, D. Liang, G. Van Tendeloo and P. A. Jacobs, J. Catal., 2008, 254, 64–70.
33. Y. Jiang, K. Lin, Y. Zhang, J. Liu, G. Li, J. Sun and X. Xu, Appl. Catal., A, 2012, 445–446, 172–179.
34. G. De Cremer, M. B. J. Roeffaers, E. Bartholomeeusen, K. Lin, P. Dedecker, P. P. Pescarmona, P. A. Jacobs, D. E. De Vos, J. Hofkens and B. F. Sels, Angew. Chem., Int. Ed., 2010, 49, 908–911.
35. X. Wang, G. Li, L. Liu and H. Liu, Chin. J. Catal., 2011, 32, 1656–1661.
36. K. Moller, B. Yilmaz, U. Muller and T. Bein, Chem. Mater., 2011, 23, 4301–4310.
37. F.-S. Xiao, L. Wang, C. Yin, K. Lin, Y. Di, J. Li, R. Xu, D. Su, R. Schlogl, T. Yokoi and T. Tatsumi, Angew. Chem., Int. Ed., 2006, 45, 3090–3093.
38. M. Kruk and L. Cao, Langmuir, 2007, 23, 7247–7254.
39. R. Nagarajan, Colloids Surf., B, 1999, 16, 55–72.
40. B. E. Yoldas, J. Mater. Sci., 1986, 21, 1087–1092.
41. A. Zecchina, G. Spoto, S. Bordiga, A. Ferrero, G. Petrini, M. Padovan and G. Leofanti, Stud. Surf. Sci. Catal., 1991, 69, 251–258.
42. G. Ricchiardi, A. Damin, S. Bordiga, C. Lamberti, G. Spano, F. Rivetti and A. Zecchina, J. Am. Chem. Soc., 2001, 121, 11409–11419.
43. S. Bordiga, S. Coluccia, C. Lamberti, L. Marchese, A. Zecchina, F. Boscherini, F. Buffa, F. Genoni, G. Leofanti, G. Petrini and G. Vlaic, J. Phys. Chem., 1994, 98, 4125–4132.
44. A. Bhaumik and T. Tatsumi, J. Catal., 2000, 189, 31–39.
45. E. Gianotti, C. Bisio, L. Marchese, M. Guidotti, N. Ravasio, R. Psaro and S. Coluccia, J. Phys. Chem. C, 2007, 111, 5083–5089.
46. P. Ratnasamy, D. Srinivas and H. Knozinger, Adv. Catal., 2004, 48, 1–169.
47. L. Y. Chen, G. K. Chuah and S. Jaenicke, Catal. Lett., 1998, 50, 107–114.
48. I. W. C. E. Arends and R. A. Sheldon, Appl. Catal., A, 2001, 212, 175–187.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra10145c

‡ These authors have contributed equally to this study.

---