Synthesis, characterization and catalytic activity of binary metallic titanium and iron containing mesoporous silica

Synthesis, characterization and catalytic activity of binary metallic titanium and iron containing mesoporous silica

Microporous and Mesoporous Materials 162 (2012) 51–59 Contents lists available at SciVerse ScienceDirect Microporous and Mesoporous Materials journa...

1MB Sizes 3 Downloads 57 Views

Microporous and Mesoporous Materials 162 (2012) 51–59

Contents lists available at SciVerse ScienceDirect

Microporous and Mesoporous Materials journal homepage: www.elsevier.com/locate/micromeso

Synthesis, characterization and catalytic activity of binary metallic titanium and iron containing mesoporous silica Yong Wu a,b, Yinjie Zhang a, Jun Cheng a, Zheng Li a, Haiqing Wang a, Qinglin Sun a, Bing Han a,c, Yan Kong a,⇑ a b c

State Key Laboratory of Materials-Oriented Chemical Engineering, College of Chemistry and Chemical Engineering, Nanjing University of Technology, Nanjing 210009, China Jiangsu Key Laboratory of Biofunctional Materials, College of Chemistry and Materials Science, Nanjing Normal University, Nanjing 210046, China School of Material Engineering, Nanjing Institute of Technology, Nanjing 211167, China

a r t i c l e

i n f o

Article history: Received 24 November 2011 Received in revised form 19 April 2012 Accepted 29 April 2012 Available online 15 June 2012 Keywords: Binary metal Mesoporous material Interaction Catalytic activity

a b s t r a c t A series of Fe–Ti incorporated-SBA-15 and MCM-41 has been synthesized using different heteroatom sources and characterized by using XRD, N2-adsorption, HRTEM, ICP, FT-IR, and UV–vis techniques. Catalytic performances of the obtained materials were evaluated in the hydroxylation of styrene with H2O2. Results indicated that (a) all samples exhibited typical hexagonal arrangement of mesoporous structure with high surface areas, (b) the kinds of heteroatoms sources had a significant effect on the heteroatom contents in the samples, i.e. there was a strong interaction between Ti and Fe species in the synthesis procedure, and (c) catalytic results revealed that the selectivity of benzaldehyde and the conversion of styrene were controlled by the contents of Fe and Ti species in the samples, respectively. An appropriate nTi/nFe ratio in catalysts could approach the optimal catalytic performance. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction In 1992, mesoporous materials M41S were successfully synthesized by Mobil researchers [1]. Since then, much attention has been directed toward the synthesis of numerous different mesoporous materials with different pore structures and compositions by changing the surfactants or inorganic precursors such as HMS [2], MSU-n [3], SBA-n [4] and FSM-16 [5]. All of the synthesized mesoporous materials possess a regular array of uniform one- or three-dimensional pores with diameters of 2–10 nm and high surface area (above 1000 m2/g). The most significant disadvantage of these mesoporous materials limiting their application to catalysis, however, is that there are actually few catalytic active sites on their amorphous SiO2 wall. Fortunately, reports have shown that the incorporation of metals including most of the transition metals [6–12] and some maingroup elements such as boron [13], gallium [14], and indium [15] etc. into the walls of mesoporous materials could modify the composition and catalytic activity of the materials. Among them, titanium-containing mesoporous molecular sieves were of great significance in the selective catalytic oxidation processes with bulky molecules, such as the epoxidation of olefins [16], unsaturated alcohols [17], plant oils [18], and the oxidation of organic sulfides [19]. Fe-containing mesoporous silicates also have high ⇑ Corresponding author. Tel./fax: +86 25 83587860. E-mail address: [email protected] (Y. Kong). 1387-1811/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.micromeso.2012.04.046

activities in isomerization of a-pinene oxide [20], the alkylation of aromatic hydrocarbon, selective oxidation of benzene or alkane [21–23], especially in the hydroxylation of phenol [24]. Nevertheless, the incorporation of multi-component of heteroatoms can modify the surface properties of mesoporous silicates more effectively than those of mono-kind of heteroatom and could be widely used in the field of catalysis. Up till now, a few reports have been focused on the incorporation of multi-metal ions containing mesoporous silicates. For example, Calleja et al. [25] reported the synthesis and catalytic activity of Cr/Al-SBA-15, which presented almost four times more activity than a conventional Cr/SiO2 Phillips catalyst in ethylene polymerization. Researches of Balu et al. demonstrated that Fe/Al doped SBA-15 and MCM41 exhibited much higher catalytic activities in the oxidation of benzyl alcohol to benzaldehyde [26]. Selvaraj et al. [27–29] reported the synthesis and catalytic performance of Zn–Al-MCM41. They found that the catalytic acidity of materials increased as zinc was incorporated into Al–MCM41 and exhibited higher activities in the isopropylation of toluene and cyclization of ethanolamine. Telalovic´ et al. also found that Zn–Al–TUD-1 was efficient catalyst in the prins cyclization of citronellal [30]. Li et al. [31] reported the first synthesis of Fe–Al–SBA-15 by introduction of Fe3+ in the synthesis gel of Al–SBA-15. The resulting materials were active for the oxidation of benzene to phenol with nitrous oxide, and the selectivity to phenol was up to 40% at a benzene conversion of 2.3%. In Kalita et al.’s study on the catalytic benzylation of toluene, the conversion over Ce–Al–MCM-41 catalyst was 2–3 times than that

Y. Wu et al. / Microporous and Mesoporous Materials 162 (2012) 51–59

Ti–Fe–MCM-41s were synthesized by traditional sol–gel method, employing cethyltrimethylammonium bromide (CTMAB) as structure-directing agent, Na2SiO39H2O, Fe(NO3)39H2O, Ti(SO4)2 and tetra-butyl ortho-titanate (TBOT) as silicon, iron and titanium sources, respectively. A typical synthesis procedure was as follows: 5.7 g sodium silicate and 1.8 g CTMAB were dissolved in 45 ml H2O. Then 5 ml solution with desired amount of Fe(NO3)39H2O was added into the mixture. After stirring for 1 h, certain amount of 1 M sulfuric acid containing desired amount of Ti(SO4)2 was added and the pH value was adjusted to 9–9.5[40]. For TBOT as titanium source, TBOT was first dissolved in 5 ml ethanol and 0.5 ml acetic acid. The solution and 1 M sulfuric acid were added simultaneously into above mixture and the pH value was adjusted to 9–9.5. The resultant was aged at 393 K for 72 h and the resulting solid was recovered by centrifugation, followed by repeated washing with deionized water, EtOH and acetone, and then dried at room temperature. The molar composition of the mixture was 100SiO2: xTiO2: yFe2O3: 25 CTMAB, where x = 0  5 and y = 0  5, respectively. The template was removed at 823 K for 5 h in air stream with a heating rate of 1 K/min to obtain Ti–Fe–MCM-41. The Ti–Fe–MCM-41s using Ti(SO4)2 as Ti source were designated as xTi–yFe–MCM41-S, while the samples using TBOT as Ti source were designated as xTi–yFe–MCM41–T. A synthesis procedure of Ti–Fe–SBA-15 was as follows: 4 g triblock copolymer poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol) (Pluronic P123, molecular weight = 5800, EO20PO70EO20) was dissolved in 90 ml 0.20 M HCl and stirred for 4 h. After the pH value was adjusted to 1.5 with 1:1 ammonia, 9.0 g of tetraethyl orthosilicate (TEOS), desired amount of TBOT (dissolved in 5 ml ethanol and 0.5 ml acetic acid) and the appropriate amount of ferric nitrate were added directly to the homogeneous solution. After stirred at 313 K for 24 h, the resulting gel was put into the teflon autoclave and maintained at 373 K for another 24 h. The precipitant was collected, washed thoroughly with distilled water, absolute ethanol for several times and dried in air. The molar composition of the mixture was 100SiO2:xTiO2:

Powder X-ray diffraction (XRD) patterns were recorded by a Philips PW 170 diffractometer equipped with Cu Ka radiation (k = 0.154168 nm) and Ni filter and operated at 40 kV and 40 mA. The nitrogen adsorption/desorption isotherms of the samples were obtained on a Micromeritics ASAP-2020 analyzer at 77 K. The samples were degassed at 573 K for 300 min before analysis. Specific surface area, pore size distribution and pore volume were determined by conventional BET (Brunauer–Emmett–Teller) and BJH (Barrett–Joyner–Halenda) equations using adsorption data. FT-IR spectra were recorded with Bruker VECTOR 22 FT-IR spectrometer in the range of 4000–400 cm1. Samples were mixed and ground with KBr followed by pressing into pellets. UV–vis diffuse reflectance spectra were measured with a Perkin–Elmer Lambda 35 spectrometer equipped with a Praying-Mantis diffuse reflectance attachment. BaSO4 was used as reference. Chemical compositions were determined using a Jarrell-Ash 1100 inductively coupling plasma (ICP) atomic emission spectrometer. The samples were dissolved in suitable hot acid before analysis. HRTEM images were taken on a JEM-2010 UHRTEM instrument at an acceleration voltage of 200 kV. The samples were crushed in A.R. grade ethanol and supported on copper grids. 2.3. Catalytic tests The catalytic oxidation of styrene with H2O2 was carried out using a three-neck flask equipped with a magnetic stirrer and a reflux condenser. 0.1 g of catalyst was mixed with 2 ml styrene and 20 ml acetonitrile. The reaction temperature was set up to be 343 K and 1.8 ml 30 wt% H2O2 aqueous solution was introduced. After reacted for 12 h, the catalyst was filtrated and all the liquid organic products were identified on an SP-6890 gas chromatograph (Lunan Ruihong Chemical Instrument Co. Ltd., China) with a 0.32 mm  30 m SE-54 capillary column. The analysis conditions were as follows: FID detector, temperature program from 373 to

1

2

3

4

200

2.1. Synthesis

2.2. Characterization

110

2. Experimental

yFe2O3:1.6P123, where x = 0  5 and y = 0 – 5, respectively. The template was removed by the same procedure as that of Ti–Fe– MCM-41. The Ti–Fe–SBA-15 using TBOT as Ti source was designated as xTi–yFe–SBA–15–T.

100

of Al–MCM-41 [32]. Since all these bimetallic mesoporous materials exhibit higher catalytic activity than the conventional single species containing catalysts, the synthesis and characterization of multi-component mesoporous silicates should be important for the improvement of their functionality as catalysts. However, the synthesis of multi-heteroatoms modified mesoporous silicates was harder than those of mono-kind of heteroatom, possibly due to the difficulty of control over simultaneous hydrolyzation of the multi-type of metal and silicon species. Meanwhile, the interactions of metal species under different conditions were relatively complicated and might prevent the incorporation of guest species. The synthesis and catalytic properties of MCM-41 modified with titanium and iron have been reported by Popova et al. These materials exhibited high activity in the full oxidation of toluene [33] and the dehydrogenation of cyclohexanol [34,35]. In the present study, Ti- and Fe-containing SBA-15 and MCM-41 materials with various titanium and iron contents were synthesized by direct hydrothermal synthesis method and characterized by ICP, XRD, N2physisorption, FT-IR and UV–vis techniques. The interactions of Ti and Fe in the synthesis procedure on the composition have also been extensively studied. Considering that iron or titanium containing mesoporous silica was active in the oxidation of styrene with H2O2, [36–39] catalytic activities of the synthesized materials in the same reaction were also investigated as comparison.

Intensity/a.u.

52

5

5Ti0Fe 5Ti1Fe 5Ti2Fe 5Ti5Fe 2Ti5Fe 1Ti5Fe 0Ti5Fe 6

2 theta/deg. Fig. 1a. XRD patterns of Ti–Fe–MCM-41–S.

7

8

53

Y. Wu et al. / Microporous and Mesoporous Materials 162 (2012) 51–59

5Ti0Fe

5Ti0Fe 5Ti1Fe 5Ti2Fe 5Ti5Fe 2Ti5Fe 1Ti5Fe 0Ti5Fe 1

2

3

4

5

6

7

5Ti2Fe

Quantity Adsorbed(a.u.)

Intensity/a.u.

5Ti1Fe

8

5Ti5Fe 2Ti5Fe 1Ti5Fe 0Ti5Fe

0.0

2 theta/deg.

0.4

0.6

0.8

1.0

Relative pressure ( p/p0)

Fig. 1b. XRD patterns of Ti–Fe–MCM-41–T.

Fig. 2a. Nitrogen adsorption/desorption isotherms of Ti–Fe–MCM-41–S.

5Ti0Fe 5Ti1Fe 5Ti2Fe 5Ti5Fe 2Ti5Fe

Quantity Adsorbed(a.u.)

5Ti0Fe 5Ti1Fe 5Ti2Fe 5Ti5Fe 2Ti5Fe

Intensity/a.u.

0.2

1Ti5Fe 0Ti5Fe

0.0

1Ti5Fe 0Ti5Fe

0.2

0.4

0.6

0.8

1.0

Relative pressure ( p/p0) 0

1

2

3

4

5

6

2 theta/deg.

Fig. 2b. Nitrogen adsorption/desorption isotherms of Ti–Fe–MCM-41–T.

Fig. 1c. XRD patterns of Ti–Fe–SBA-15–T.

473 K, detection temperature 523 K, and vaporization temperature 523 K. The content of each component was calculated by a revised area normalization method.

3. Results and discussions 3.1. Mesoporous structure of the samples The small angle XRD patterns of calcined Ti–Fe–MCM-41-S, Ti–Fe–MCM-41-T and Ti–Fe–SBA-15-T samples were duplicated in Figs. 1a–c, respectively. The narrow (1 0 0) peaks at about 2h = 2.5° for MCM-41 and about 2h = 1° for SBA-15 indicated the mesoporous structure and high orderings of the samples. Also, weak (1 1 0) and (2 0 0) peaks suggested the two-dimensional hexagonal structure of all the samples. The (2 1 0) peak was too weak to be recognized possibly due to the incorporation of heteroatoms

into the framework of mesoporous molecular sieves and the concomitant decrease of order in the mesoporous structure [10,41]. N2 physisorption isotherms and pore size distributions of calcined Ti–Fe–MCM-41–S, Ti–Fe–MCM-41–T, and Ti–Fe–SBA-15–T samples were presented in Figs. 2a–f. All the isotherms were of type IV, typical of mesoporous solid [42]. For the MCM-41 silica samples, the sharp capillary condensation steps were at relative pressure of 0.2 6 p/p0 6 0.4 (Fig. 2a and b). For the SBA-15 silica samples, the isotherm profiles exhibited a sharp capillary condensation step at a relative pressure of ca. p/p0 = 0.5–0.7 and H1 type hysteresis typical of mesoporous materials with uniform tubular pores (Fig. 2c) [43,44]. The sharp increase of nitrogen adsorption quantity appeared at different relative pressure, which was due to the incorporation of heteroatom that induced the pore sizes to vary in a certain range. The average pore sizes of Ti–Fe–MCM-41 were 2–3 nm (Figs. 2d and e) and those of Ti–Fe-SBA-15 were in the range of 5–8 nm (Fig. 2f). The heteroatoms incorporation resulted in the shift of

54

Y. Wu et al. / Microporous and Mesoporous Materials 162 (2012) 51–59

5Ti1Fe

5Ti0Fe

5Ti2Fe

5Ti1Fe Pore volume/ a.u.

Quantity Adsorbed(a.u.)

5Ti0Fe

5Ti5Fe 2Ti5Fe 1Ti5Fe

5Ti2Fe 5Ti5Fe 2Ti5Fe 1Ti5Fe

0Ti5Fe

0Ti5Fe 0.0

0.2

0.4

0.6

0.8

1

1.0

2

3

4

5

6

7

8

Pore diameter / nm

Relative pressure(p/p0)

Fig. 2e. Pore size distributions of Ti–Fe–MCM-41–T.

Fig. 2c. Nitrogen adsorption/desorption isotherms of Ti–Fe–SBA-15–T.

5Ti0Fe 5Ti1Fe

5Ti0Fe 5Ti1Fe

Pore volume(a.u.)

Pore volume/ a.u.

5Ti2Fe 5Ti5Fe 2Ti5Fe 1Ti5Fe

5Ti2Fe 5Ti5Fe 2Ti5Fe 1Ti5Fe 0Ti5Fe

0Ti5Fe 1

2

3

4

5

6

7

8

9

10

Pore diameter / nm Fig. 2d. Pore size distributions of Ti–Fe–MCM-41–S.

pore size. Also, the narrow pore size distribution revealed a uniform mesoporosity. HRTEM images of some samples were duplicated in Fig. 3. The ordered hexagonal pore arrangements of 1Ti–5Fe–MCM-41–T (Fig. 3a) and 1Ti–5Fe–SBA-15–T (Fig. 3c) samples were clearly visible, and the pore sizes were in the range of 2.2–2.6 and 6.8– 7.5 nm, respectively, which were in good agreement with the average pore size calculated by BJH model. The pore arrangements became worm-like with the metal contents increase, as the images of 5Ti–5Fe–MCM-41–T and 5Ti–5Fe–SBA-15–T shown in Fig. 3b and d.

3.2. Composition of the materials 3.2.1. The content of titanium and iron The composition of the mesoporous materials definitely plays a key role in its catalytic activity. Determination of the heteroatom content in mesoporous silica is therefore important. As illustrated

0

5

10

15

20

25

30

35

40

Pore diameter /nm Fig. 2f. Pore size distributions of Ti–Fe–SBA-15–T.

in Table 1, the titanium, iron and total metal contents in Ti–Fe– MCM-41–S and Ti–Fe–MCM-41–T analyzed by ICP varied in distinguishing manners. For the Ti–Fe–MCM–41–S samples synthesized using titanium sulfate as titanium source, the introduction of iron into the sample with Ti/Si molar ratio of 0.05 in the gel caused the slight decrease of titanium content in the product while the iron content went through a slight increase simultaneously, and the total content of metals remained almost untouched with the increase of iron content in the gel. On the contrary, as the Fe/Si molar ratio in the gel maintained to be 0.05, the introduction of titanium caused slight decrease of iron content while the titanium content and total content of metal increased linearly. However, as tetra-butyl orthotitanate (TBOT) was used as titanium source, the total metal content always increased with the metal/silica molar ratio in the gel. When the iron or titanium content in the gel is fixed, the addition of another metal has much smaller influence on their content in the products than sulfuric titanium used as titanium source. The variations of metal content in Ti–Fe–SBA-15-T samples synthesized

55

Y. Wu et al. / Microporous and Mesoporous Materials 162 (2012) 51–59

Fig. 3. HRTEM images of (a) 1Ti–5Fe–MCM-41–T, (b) 5Ti–5Fe–MCM-41–T, (c) 1Ti–5Fe–SBA-15–T and (d) 5Ti–5Fe–SBA-15 samples.

Table 1 The structure and constituent properties of calcined Ti–Fe–MCM-41–S materials with different metal contents. Sample

5Ti 1Fe5Ti 2Fe5Ti 5Fe5Ti 5Fe2Ti 5Fe1Ti 5Fe

nFe/nSi (mol%)

nTi/nSi (mol%) a

Total metal content (mol%)

d100 (nm)

a0 (nm)b

ABET (m2/g)

D (nm)

5.01 5.26 5.37 5.42 3.36 3.03 2.83

3.32 3.37 3.37 3.34 3.56 3.59 3.50

3.84 3.89 3.89 3.86 4.11 4.15 4.05

756 778 803 750 897 878 827

2.23 2.38 2.40 2.33 2.57 2.57 2.57

c

V (cm3/g)

d (nm)

0.68 0.71 0.79 0.85 0.99 0.94 1.00

1.61 1.51 1.49 1.53 1.54 1.58 1.48

a

Gel

Calcined

Gel

Calcined

0 1 2 5 5 5 5

0 0.48 0.82 1.62 1.73 2.05 2.83

5 5 5 5 2 1 0

5.01 4.78 4.55 3.80 1.63 0.98 0

Note: a0 = unit cell parameter, ABET = specific surface area, D = pore diameter, d = wall thickness V = pore volume. a ICP results. p b Unit cell parameter value calculated using a0 = 2d100/ 3. c Pore diameter calculated by BJH method.

using TBOT as titanium source were similar with those of Ti– Fe–MCM–41–T, apart from the slightly lower metal/silicon molar ratio. These facts indicated that the interaction of the heteroatoms in the synthesis procedure was different as diverse kinds of compounds were used as metal sources, and there was a notable relationship with the hydrolyzation of Fe and Ti species in solutions. The iron species in the hydrolysis products of Fe(NO3)3 exist mainly in the cationic form of Fe(OH)2+ in the solution. On the other hand, the titanium species in the hydrolysis products of Ti(SO4)2 and TBOT were mainly in the cationic form of Ti(OH)3+ and the anionic form of TiO44, respectively (as shown in Fig. 4). When titanium sulfate was used as titanium source, the hydrolysis of titanium species were prior to those of iron species, and the

resulting Ti(OH)3+ had a strong electrostatic repulsion to Fe(OH)2+ [45,46], thereby reducing the hydrolysis tendency of trivalent iron (Fig. 3, Eq. (1)), which made the iron content in the product lower and decrease with titanium content. On the contrary, when TBOT was used as titanium source, the TiO44 anion from the hydrolysis of titanium species had electrostatic attraction with Fe(OH)2+, thus accelerating the hydrolysis of trivalent iron. As a consequence, the use of TBOT as titanium source led to complete hydrolysis of iron components, and the iron content in the product did not increase with the change of titanium content, which was also much larger than the sample without titanium species. Meanwhile, the metal contents in Ti–Fe–SBA-15–T samples were lower than those in Ti–Fe–MCM-41–T samples, due to the strong acidic condition

56

Y. Wu et al. / Microporous and Mesoporous Materials 162 (2012) 51–59

Table 2 The structure and constituent properties of calcined Ti-Fe-MCM-41-T materials with different metal contents. Sample

5Ti 1Fe5Ti 2Fe5Ti 5Fe5Ti 5Fe2Ti 5Fe1Ti 5Fe

nFe/nSi (mol%) Gel

Calcined

0 1 2 5 5 5 5

0 1.2 2.34 5.34 5.42 5.58 2.83

nTi/nSi (mol%) a

Total metal content (mol%)

d100 (nm)

a0 (nm)b

ABET (m2/g)

D (nm)c

V (cm3/g)

d (nm)

4.65 6.05 7.67 9.99 7.44 6.68 2.83

3.42 3.40 3.40 3.25 3.40 3.37 3.50

3.95 3.92 3.92 3.75 3.92 3.89 4.05

729 715 698 652 763 804 827

2.43 2.36 2.36 2.23 2.57 2.51 2.49

0.68 0.69 0.65 0.67 0.81 0.86 1.00

1.52 1.56 1.56 1.52 1.35 1.38 1.56

a

Gel

Calcined

5 5 5 5 2 1 0

4.65 4.85 5.33 4.65 2.02 1.10 0

Note: a0 = unit cell parameter, ABET = specific surface area, D = pore diameter, d = wall thickness V = pore volume. a ICP results. p b Unit cell parameter value calculated using a0 = 2d100/ 3. c Pore diameter calculated by BJH method.

Table 3 The structure and constituent properties of calcined Ti-Fe-SBA-15-T materials with different metal contents. Sample

5Ti 1Fe5Ti 2Fe5Ti 5Fe5Ti 5Fe2Ti 5Fe1Ti 5Fe

nFe/nSi (mol%) Gel

Calcined

0 1 2 5 5 5 5

0 0.53 1.33 3.6 3.08 2.1 2

nTi/nSi (mol%) a

Total metal content (mol%)

d100 (nm)

a0 (nm)b

ABET (m2/g)

D (nm)c

V (cm3/g)

d (nm)

4.41 4.84 5.39 7.94 4.74 3.01 2.00

9.83 9.82 9.89 9.71 9.64 9.53 9.24

11.35 11.34 11.42 11.21 11.13 11.01 10.66

697 763 765 687 749 742 673

7.28 7.29 7.33 7.08 7.18 7.05 5.41

0.91 0.94 1.00 0.84 1.02 1.40 0.71

4.07 4.05 4.09 4.13 3.95 3.96 5.25

a

Gel

Calcined

5 5 5 5 2 1 0

4.41 4.31 4.06 4.34 1.66 0.91 0

Note: a0 = unit cell parameter, ABET = specific surface area, D = pore diameter, d = wall thickness V = pore volume. a ICP results. p b Unit cell parameter value calculated using a0 = 2d100/ 3. c Pore diameter calculated by BJH method.

SBA-15 synthesized in. According to Fig. 3, the higher H+ concentration will reduce the trend of metal ion hydrolysis. 3.2.2. Status of the metallic elements The status of heteroatoms in mesoporous silica is critical to their catalytic activity, especially when heteroatoms act as active sites. Up till now, there is no systematic approach to distinguish whether the heteroatoms in the framework or on the surface of mesoporous molecule sieves. Usually it can be judged by means of several hybrid methods. For example, the high angle XRD patterns can be used to determine whether there are crystalline heteroatom oxides in the materials. The existence of amorphous or nanoparticle heteroatomic oxides can be determined through various spectroscopic methods. Meanwhile, the strong interaction between silica and heteroatom can also be detected. The regularity variation of structural parameters such as pore size, specific surface area and thickness of pore wall etc. can also be applied to analyze the status of heteroatoms. 3.2.2.1. High angle XRD patterns. For the prepared mesoporous materials, there was only a weak peak centered at ca. 23.3°, corresponding to amorphous silica in the high angle XRD patterns (figures not given) of all the calcined samples. No obvious diffraction peaks corresponding to crystalline iron and/or titanium oxides were observed. These facts suggested that iron and/or titanium species were successfully introduced into the framework of ordered mesoporous silica or the particle size of crystalline oxides on the surface of the samples were too small to be detected by X-ray diffraction. 3.2.2.2. FT-IR spectra. Fig. 5 shows the FT-IR spectra of Ti-Fe-SBA15-T samples. In the FT-IR spectrum of pure SBA-15, a broad band at ca. 1080 cm1 and a band at ca. 800 cm1 were assigned to the

asymmetric and symmetric Si–O stretching vibrations [47]. The bands at ca. 960 and 460 cm1 were due to the stretching and bending vibrations of surface Si–O groups, respectively [48]. As the heteroatoms were introduced to SBA-15, a slight red shift was observed for almost all the bands. FT-IR spectra of series Ti– Fe–MCM-41–T and Ti–Fe–MCM-41–S (Figures not given) exhibited the same changes. It could be concluded that there was a strong interaction between the atoms of iron, titanium and silicon, i.e. some of the heteroatoms were incorporated into the framework of mesoporous silica and the M–O–Si bonds were formed. Since the ion radii of Fe3+ and Ti4+ were all larger than that of Si4+, the length of M–O (M = Fe and Ti) should be greater than that of Si– O, which led to the decrease of the force constant (k) of the bands. The atomic weights of iron and titanium were all bigger than that of silicon, thus the reduced mass (l) would increase. As a result, qffiffiffi the vibration frequencies calculated from the formula m ¼ 21pc lk decreases. 3.2.2.3. UV–vis spectra. UV–vis spectroscopy (Fig. 6a–c) had been extensively used to characterize the coordination circumstance of Fe3+ and Ti4+ ions in mesoporous materials [49,50]. A broad band between 200 and 350 nm centered at 220 and 255 nm, assigned to the low-energy dQ–pQ charge-transfer transitions between tetrahedral oxygen ligands and central Ti4+ and Fe3+ ion, were observed for all samples. These two bands that were generally

Fig. 4. The hydrolysis balance equation of Fe3+, Ti4+ and TBOT.

57

460

5Ti

960

1080

SBA-15

800

3440

Y. Wu et al. / Microporous and Mesoporous Materials 162 (2012) 51–59

Transmittance/a.u.

1Fe5Ti 2Fe5Ti

5Fe 5Fe1Ti 5Fe2Ti 5Fe5Ti

4000

3500

3000

2500

5Ti0Fe 5Ti1Fe 5Ti2Fe

Absorbance(a.u.)

5Fe5Ti

2000

1500

Wavenumber/cm

1000

5Ti5Fe 2Ti5Fe 1Ti5Fe

500

0Ti5Fe

-1

Fig. 5. The FT-IR spectra of xTi–xFe–SBA-15–T.

200

300

400

500

600

700

800

Wavelength(nm) Fig. 6b. UV–vis spectra of Ti–Fe–MCM-41–T.

Absorbance(a.u.)

5Ti0Fe

200

5Ti5Fe 2Ti5Fe 1Ti5Fe 0Ti5Fe 300

400

500

600

700

5Ti0Fe 5Ti1Fe

Absorbance(a.u.)

5Ti1Fe 5Ti2Fe

5Ti2Fe 5Ti5Fe 2Ti5Fe 1Ti5Fe 0Ti5Fe

800

Wavelength(nm) Fig. 6a. UV–vis spectra of Ti–Fe–MCM-41–S.

200

300

400

500

600

700

800

Wavelength(nm) attributed to tetracoordinated Ti(IV) and Fe(III) ions in the framework of mesoporous silica were intensified with the metal content. Otherwise, the bands 270 and 480 nm corresponded to partially polymerized hexacoordinated Ti and Fe species were not detected for the samples with lower metal content. However, as the metal content increased, the emergence of the two bands demonstrated that some M–O–M clusters were suspected to coexist with the isolated Ti/Fe sites. As in the spectra of 5Ti–5Fe–MCM-41/SBA-15, both the bands were intensified, which were assigned to the presence of titanium and iron nanocrystallites although they escaped the XRD detection [51,52]. Meanwhile, the intensities of the two bands in the spectra of xTi–xFe–SBA-15 samples were all higher than those of xTi–xFe–MCM-41 samples. It can be concluded that there were more abundant metal oxides in xTi–xFe–SBA-15 samples, possibly due to using TEOS as silicon sources which did not hydrolyze simultaneously with titanium and/or iron sources.

3.2.2.4. Variation of structral parameter. The textural parameters of the samples were summarized in Table 1, where a0, D, d denoted the unit cell parameter, average pore diameter and thickness of pore wall, respectively.

Fig. 6c. UV–vis spectra of Ti–Fe–SBA-15–T.

In small angle XRD pattern, the shift of diffraction peaks at small angle with different metal content (Fig. 1) indicated the variety of a0 values due to metal incorporation which could also be found in Table 1. As the XRD patterns of Ti–Fe–MCM-41–S series displayed in Fig. 1a, with the increase of iron content, the diffraction peaks of the sample series shifted to lower angle until 5Ti–2Fe–MCM-41– S. The Ti content and a0 of 5Ti–0Fe–MCM-41–S were 5.01 mol% and 3.84 nm, respectively. Because of the similar ion radius of Ti4+ and Fe3+ (Ti4+:60.5 pm, Fe3+:55 pm), a0 of 5Ti–1Fe–MCM-41–S and 5Ti–2Fe–MCM-41–S were increased to 3.89 nm, due to the increased total metal content (5.26 and 5.37 mol%, respectively). But the diffraction peaks of 5Ti–5Fe–MCM-41–S shifted to higher angle and a0 decreased slightly. These phenomena might be resulted from the presence of titanium and iron nanocrystallites. However, a0 of 0Ti–5Fe–MCM-41–S and 1Ti–5Fe–MCM-41–S increased from 4.05 to 4.15 nm as the total metal content increased from 2.83 to 3.03 mol%, yet a0 of 2Ti–5Fe–MCM-41–S and 5Ti–5Fe– MCM-41–S decreased as the total metal content continued to increase.

58

Y. Wu et al. / Microporous and Mesoporous Materials 162 (2012) 51–59

Table 4 The catalytic activity of Ti-Fe-MCM-41-S in the styrene oxidation reaction. Catalysts

Conversion of styrene/%

5Ti 1Fe5Ti 2Fe5Ti 5Fe5Ti 5Fe2Ti 5Fe1Ti 5Fe

55.3 50.0 32.9 29.5 31.1 30.5 30.1

Products distribution/% Benzaldehyde

Hyacinthin

Epoxide

Benzoic acid

91.1 92.0 93.4 91.4 94.0 96.3 97.4

0 0 0 0 0 0 0

1.3 2.6 3.3 6.0 2.0 2.8 1.3

0 0 0 0 0 0 0

Reaction condition: Styrene 2.0 ml, nStyrene/Nh2O2 = 1, temperature 70 °C, catalyst 0.1 g, reaction time 12 h, solvent 20 ml CH3CN.

Table 5 The catalytic activity of Ti-Fe-MCM-41-T in the styrene oxidation reaction. Catalysts

5Ti 1Fe5Ti 2Fe5Ti 5Fe5Ti 5Fe2Ti 5Fe1Ti 5Fe

Conversion of styrene/%

51.0 50.0 42.3 37.5 36.0 33.9 30.1

Products distribution/% Benzaldehyde

Hyacinthin

Epoxide

Benzoic acid

90.2 90.1 90.7 90.2 92.5 93.1 97.4

0 0 0 0 0 0 0

5.9 4.6 3.6 5.0 3.2 3.1 1.3

0 0 0 0 0 0 0

Reaction condition: Styrene 2.0 ml, nStyrene/nH2O2 = 1, temperature 70 °C, catalyst 0.1 g, reaction time 12 h, solvent 20 ml CH3CN.

Table 6 The catalytic activity of Ti–Fe–SBA-15–T in the styrene oxidation reaction. Catalysts

5Ti 1Fe5Ti 2Fe5Ti 5Fe5Ti 5Fe2Ti 5Fe1Ti 5Fe

Conversion of styrene/%

44.0 37.1 32.3 25.7 12.2 9.3 8.1

Products distribution/% Benzaldehyde

Hyacinthin

Epoxide

Benzoic acid

89.6 86.3 86.1 92.0 92.9 92.5 93.5

1.3 0.8 0 0 0 0 0

9.2 12.9 13.9 8.0 7.1 7.5 6.5

0 0 0 0 0 0 0

Reaction condition: Styrene 2.0 ml, nStyrene/nH2O2 = 1, temperature 70 °C, catalyst 0.1 g, reaction time 12 h, solvent 20 ml CH3CN.

Similar considerations can be applied to the analysis of Figs. 1b and c. As the XRD patterns of Ti–Fe–MCM-41–T series displayed in Fig. 1b, the shift of diffraction peaks to higher angle and the decrease of a0 as the total metal content increased might be resulted from the presence of titanium and iron nanocrystallites. However, as the XRD patterns of Ti–Fe–SBA-15–T series displayed in Fig. 1c, the diffraction peaks shifted to lower angle and a0 increased as the total metal content increased up to maximum titanium and iron concentration. 3.3. Catalytic activity for the oxidation of styrene The distributions of products in the oxidation of styrene with H2O2 are closely related to the used solvents. Using alkaline DMF as solvents and Fe–MCM-41 as catalyst, Wang et al. [36] reported 41.8% and 37.3% selectivity of styrene oxide and benzaldehyde, respectively. In contrast, Zhang et al. [37] obtained more than 99% selectivity of benzaldehyde using acetonitrile as solvents and Fe–SBA-15 as catalyst. To facilitate the studying on the catalytic activity with heteratoms content, acetonitrile is the preferred solvent in our research. Tables 4–6 summarize the catalytic activities of the above synthesized catalysts in the oxidation of styrene with H2O2 as an oxidant.

With every catalyst tested, the main product in the oxidation of styrene was benzaldehyde. Compared with mesoporous catalysts containing only iron, the conversion of styrene were higher than with only titanium-doping catalyst, while the benzaldehyde selectivity were lower. This phenomenon was similar with the reports in literatures. In Zhang et al.’s report [37], the conversion of styrene and selectivity of benzaldehyde were 9.52% and 99%, respectively when Fe–SBA-15 was used as catalyst. In another report using Ti–SBA-15 as catalyst [38], the conversion of styrene was 13.9%, whereas only 63.2% selectivity of benzaldehyde was achieved. When Ti–Fe–MCM-41–S was used as catalysts, the conversion of styrene distinctly decreased while the selectivity of benzaldehyde went up as Ti content decreased and Fe content increased (Table 4). However, in the Ti–Fe–MCM-41–T series catalysts (Table 5), when Ti content nearly unchanged and Fe content increased, the conversion of styrene decreased while the selectivity of benzaldehyde was almost invariant. When Ti content decreased and Fe content nearly unchanged, the conversion of styrene still decreased while the selectivity of benzaldehyde was observed to rise. Although the conversion of styrene is relatively lower, the catalytic reactivity tests for Ti–Fe–SBA-15–T series (see Table 6) give similar dependence of styrene conversion and benzaldehyde selectivity on Ti and Fe contents as the Ti–Fe–MCM-41–T series. These results

Y. Wu et al. / Microporous and Mesoporous Materials 162 (2012) 51–59

indicated that the conversion of styrene depended on Ti and Fe contents, and the decreasing of Ti content and the increasing of Fe content diminished the conversion. But the selectivity of benzaldehyde was mainly controlled by Fe content. The contrasting performance of the catalysts synthesized from different titanium sources Ti(SO4)2 and TBOT demonstrates the important role the synthetic methodology may play in the functionality of the resultant mesoporous materials, even for the mesoporous catalysts composed by same chemical species. An appropriate Ti and Fe contents in catalysts could approach the optimal catalytic performance. 4. Conclusions By employing traditional hydrothermal method, we have successfully synthesized series Fe–Ti bi-metal doped MCM-41 and SBA-15 mesoporous silica with various Fe/Ti molar ratios. The extensive characterization study demonstrates that the introduction of the Fe and Ti metallic species did not damage the mesopore-regularity and the hexagonal mesoporous structure. Meanwhile, a certain amount of Fe and Ti atoms were incorporated into the mesoporous framework as expected. The Fe/Ti molar ratio in the mesoporous materials may be affected by the employed titanium source in the synthesis procedure. It was found that TBOT actually enhances the hydrolysis of Fe species via attractive electrostatic interaction, while Ti(SO4)2 turns out to suppress the hydrolysis processes via the repulsive electrostatic interaction with the Fe(OH)2+ cation. The relevant electrostatic interaction thus decides the actual Fe/Ti molar ratio for the Fe and Ti atoms that can be incorporated into the mesoporous framework. We further tested the catalytic activity for the resultant Fe–Ti bi-metal doped mesoporous silica to study the effect of Fe/Ti molar ratio on the catalytic activity of the styrene oxidation reaction. It was observed that all the resultant mesoporous silica exhibit satisfactory catalytic activity and selectivity. The higher content of Fe component in the framework may enhance the selectivity for the benzaldehyde product and the conversion of styrene was suppressed, while the conversion of styrene is mainly dependent on the Ti content in the framework. Extra metal content may yield metal oxides out of the framework, which would cause both the selectivity for the benzaldehyde and the conversion of styrene decreased. Acknowledgements This work was financially supported by The National Natural Science Foundation of China (20876077), the Basic Research Program of Jiangsu Province of China (BE2008142), the Key Program Educational Commission of Jiangsu Province of China (10KJA530015) and the Priority Academic Program Development of Jiangsu Higher Education Institutions. References [1] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992) 710. [2] P.T. Tanev, T.J. Pinnavaia, Science 267 (1995) 865. [3] S.A. Bagshaw, E. Prouzet, T.J. Pinnavaia, Science 269 (1995) 1242.

59

[4] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky, Science 279 (1998) 548. [5] S. Inagaki, Y. Sakamoto, Y. Fukushima, O. Terasaki, Chem. Mater 8 (1996) 2089. [6] F.L. Chen, M.L. Liu, Chem. Commun. (1999) 1829. [7] C.G. Wu, T. Bein, Chem. Mater. 6 (1994) 1109. [8] L. Zhang, T. Sun, Y. Ying, Chem. Commun. (1999) 1103. [9] S. Biz, M.L. Occelli, Catal. Rev-Sci. Eng. 40 (1998) 329. [10] Y. Kong, H. Zhu, G. Yang, X. Guo, W. Hou, Q. Yan, M. Gu, C. Hu, Adv. Funct. Mater. 14 (2004) 816–820. [11] A. Corma, M.T. Navarro, J.P. Pariente, J. Chem. Soc. Chem. Commun. 2 (1994) 147–148. [12] P.T. Tanev, M. Chibwe, T.J. Pinnavaia, Nature 368 (1994) 321–323. [13] U. Oberhagemann, I. Kinski, I. Dierdorf, B. Marler, H. Gies, J. Non-cryst. solids 197 (1996) 145. [14] K. Okumura, K. Nishigaki, M. Niwa, Microporous Mesoporous Mater. 509 (2001) 44–45. [15] R.C. Vasant, K.J. Suman, S.P. Nilesh, Tetrahedron Lett. 43 (2002) 1105. [16] A. Corma, M.T. Navarro, J. Pérez-Pariente, Chem. Commun. (1994) 147. [17] C. Berlini, M. Guidotti, G. Moretti, R. Psaro, N. Ravasio, Catal. Today 60 (2000) 219. [18] L.A. Rios, P. Weckes, H. Schuster, W.F. Hoelderich, J. Catal. 232 (2005) 19. [19] A. Corma, M. Iglesias, F. Sanchez, Catal. Lett. 39 (1996) 153. [20] A. Gervasini, C. Messi, P. Cartini, A. Ponti, N. Ravasio, F. Zaccheria, J. Catal. 262 (2009) 224. [21] N. He, S. Bao, Q. Xu, Appl. Catal. A 169 (1998) 29–36. [22] Y. Wang, Q. Zhang, T. Shishido, K. Takehira, J. Catal. 209 (2002) 186–196. [23] H. Lim, J. Lee, S. Jin, et al., Chem. Commun. (2006) 463–465. [24] C. Wu, Y. Kong, J. Wang, Microporous Mesoporous Mater. 113 (2008) 163–170. [25] G. Calleja, J. Aguado, A. Carrero, J. Moreno, Appl. Catal. A: Gen. 316 (2007) 22. [26] A.M. Balu, A. Pineda, K. Yoshida, J.M. Campelo, P.L. Gai, R. Luque, A.A. Romero, Chem. Commun. 46 (2010) 7825. [27] M. Selvaraj, A. Pandurangan, K.S. Seshadri, P.K. Sinha, K.B. Lal, Appl. Catal. A. 242 (2003) 347. [28] M. Selvaraj, B.R. Min, Y.G. Shul, T.G. Lee, Microporous Mesoporous Mater. 74 (2004) 143. [29] M. Selvaraj, B.R. Min, Y.G. Shul, T.G. Lee, Microporous Mesoporous Mater. 74 (2004) 157. [30] S. Telalovic, A. Ramanathan, J.F. Ng, R. Maheswari, C. Kwakernaak, F. Soulimani, H.C. Brouwer, G.K. Chuah, B.M. Weckhuysen, U. Hanefeld, Chem. Eur. J. 17 (2011) 2077. [31] Y. Li, H.A. Xia, F.T. Fan, et al., Chem. Commun. (2008) 774–776. [32] P. Kalita, N.M. Gupta, R. Kumar, J. Catal. 245 (2007) 338. [33] M. Popova, Á. Szegedi, Z. Cherkezova-Zheleva, I. Mitov, N. Kostova, T. Tsoncheva, J. Hazard. Mater. 168 (1) (2009) 226. [34] M. Popova, Szegedi, K. Lazar, A. Dimitrova, Catal. Lett. 141 (9) (2011) 1288. [35] Á. Szegedi, M. Popova, K. Lázár, React. Kinet. Mech. Catal. 104 (2) (2011) 291. [36] Y. Wang, Q. Zhang, T. Shishido, K. Takehira, J. Catal. 209 (2002) 186. [37] L. Zhang, Z. Hua, X. Dong, L. Li, H. Chen, J. Shi, J. Mol. Catal A: Chem. 268 (2007) 155. [38] G. Calleja, R. van Grieken, R. García, J.A. Melero, J. Iglesias, J. Mol. Catal A: Chem. 182–183 (2002) 215. [39] B. Samran, S. Aungkutranont, Timothy J. White, S. Wongkasemjit, J. Sol–Gel. Sci. Technol. 57 (2011) 221. [40] Y. Kong, X.F. Guo, F.M. Zhang, S.Y. Jiang, J. Wang, Y. Lu, Q. Yan, Mater. Lett. 59 (2005) 3099. [41] Y. Kong, S. Jiang, J. Wang, S. Wang, Q. Yan, Y. Lu, Microporous Mesoporous Mater. 86 (2005) 191–197. [42] S. Brunauer, L.S. Deming, W.S. Deming, E. Teller, J. Am. Chem. Soc. 62 (1940) 1723. [43] T. Tsoncheva, J. Rosenholm, C.V. Teixeira, M. Dimitrov, M. Linden, C. Minchev, Microporous Mesoporous Mater. 89 (2006) 209. [44] T. Tsoncheva, J. Rosenholm, M. Linden, L. Ivanova, C. Minchev, Appl. Catal. A: Gen. 318 (2007) 234. [45] Y.H. Zhang, F. Gao, Y. Kong, et al., Microporous Mesoporous Mater. 113 (2008) 393–401. [46] Y. Li, Z.C. Feng, H.C. Xin, F.T. Fan, et al., J. Phys. Chem. B. 110 (2006) 26114. [47] A.A. Romero, M.D. Alba, W. Zhou, J. Klinowski, J. Phys. Chem. B. 101 (1997) 5294. [48] R. Takahashi, S. Sato, T. Sodesawa, M. Kawakita, K. Ogura, J. Phys. Chem. B. 104 (2000) 12184. [49] T. Kawabata, Y. Ohishi, S. Itsuki, N. Fujisaki, T. Shishido, K. Takaki, Q. Zhang, Y. Wang, K. Takehira, J. Mol. Catal. A 236 (2005) 99–106. [50] X. Wang, W. Lian, X. Fu, J. Basset, F. Lefebvre, J. Catal. 238 (2006) 13–20. [51] A. Bensalem, J.C. Muller, F. Bozon-Verduraz, J. Chem. Soc. Faraday Trans. 88 (1992) 153. [52] C. Beck, T. Mallat, T. Burgi, A. Baiker, J. Catal. 204 (2001) 428.