- Email: [email protected]
Epoxidation of bulky organic molecules over pillared titanosilicates
G Model
ARTICLE IN PRESS
CATTOD-9149; No. of Pages 7
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today journal homepage: www.elsevier.com/locate/cattod
Epoxidation of bulky organic molecules over pillared titanosilicates Jan Pˇrech a,∗ , Pavla Eliáˇsová a , Daifallah Aldhayan b , Martin Kubu˚ a a b
J. Heyrovsk´ y Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, v.v.i., Dolejˇskova 3, CZ-182 23 Prague 8, Czech Republic Department of Chemistry, College of Science, King Saud University, B.O. Box 2455, Riyadh 11451, Saudi Arabia
a r t i c l e
i n f o
Article history: Received 4 June 2014 Received in revised form 27 June 2014 Accepted 3 July 2014 Available online xxx Keywords: Layered TS-1 zeolite Pillaring Epoxidation Cyclooctene Norbornene Verbenol.
a b s t r a c t Layered and pillared TS-1 catalysts were prepared using surfactant C18 H37 –N+ (CH3 )2 C6 H12 –N+ (CH3 )2 C6 H13 in the hydroxide form as a structure directing agent and silicon(IV) ethoxide or its mixture with titanium(IV) butoxide as a pillaring medium. The pillaring treatment significantly increases the specific surface area (BET up to 685 m2 /g) and thus the accessibility of active centres for bulky molecules. The addition of tetrabutylorthotitanate into the pillaring mixture has a positive effect on the material performance as a catalyst of cyclooctene, norbornene and ␣-pinene epoxidation under mild conditions, with hydrogen peroxide as the oxidant. Optimum composition of the pillaring medium is Si/Ti ratio 20. The highest yields of cyclooctene oxide (16.9% after 4 h) and norbonene oxide (14.8% after 4.5 h) were obtained using the above-prepared catalyst at 60 ◦ C with substrate/catalyst mass ratio being 10 and substrate/H2 O2 ratio being 2. Selective oxidation of verbenol to verbenone was observed instead of the epoxidation with the above catalysts. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Zeolites represent a rich family of porous crystalline aluminosilicates successfully applied as catalysts, adsorbents and ion exchangers [1,2]. Originally, they are powerful acidobasic catalysts; however, incorporation of other heteroelements (e.g. titanium, boron, iron, vanadium, tin, germanium and others) into the crystalline framework opened other areas of catalysis [3]. Titaniumcontaining zeolites are able to catalyse various oxidation reactions, e.g., epoxidation of olefins [4], oxidation of alkanes to alcohols and ketones [5], oxidation of aromatic hydrocarbons to phenols [6], ammoxidation of cyclohexanone [7], oxidation of amines to hydroxylamines [8] and thioethers to sulfoxides [9]. Furthermore, these catalysts have strong advantages in comparison with amorphous materials and other conventional oxidation catalysts: (i) they possess a regular microenvironment around the active centres [10]; (ii) the active sites are resistant to deactivation via oligomerization of metal-oxide species [10]; (iii) the materials exhibit enhanced stability towards leaching of the metal in comparison with the amorphous materials and other conventional oxidation catalysts [10]. TS-1 (MFI topology), Ti-BEA and Ti-MWW are the most frequently studied titanosilicates [3]. The TS-1 zeolite is currently a catalyst of choice for epoxidation of C C double bonds with
∗ Corresponding author. Tel.: +420 266 053 856. E-mail address: [email protected] (J. Pˇrech).
hydrogen peroxide providing low waste and being environmentally friendly [11]. Its discovery [12] represents a considerable breakthrough in the field of oxidation catalysis. Presently, the TS1 is industrially applied by Eni in a propylenoxide process [13]. However, TS-1 has pores of 5.5 A˚ in diameter. Therefore, bulky substrates like terpenes cannot easily access the active sites located mainly inside the pores. Although the epoxidation of terpenes over titanosilicates has been given a certain attention, it still remains a challenge [14–16]. The application of layered materials for catalysis offers the possibility to overcome the diffusion problems [17,18] as layered materials possess high external surface areas in comparison to conventional three-dimensional titanosilicates. Ryoo and co-workers [18] reported the preparation of specially designed TS-1 in the form of nanosheets. They are formed using surfactant-based template (C18 H37 N+ (CH3 )2 C6 H12 N+ (CH3 )2 C6 H13 ·OH2 − ) instead of conventional tetrapropylammonium hydroxide [19]. This material has been reported to exhibit activity similar to bulk TS-1 in epoxidation of 1-hexene; however, its activity is an order of magnitude higher in epoxidation of cyclooctene [18]. Wu et al. [20] succeeded in the preparation of Ti-MWW combining a synthesis of Ti-MWW lamellar precursor (denoted as Ti-MCM-22P) in the presence of boron with acidic treatment of the material to remove extra-framework Ti species prior to calcination. Later, direct synthesis of the Ti-MWW lamellar precursor Ti-YNU-1 was reported. Ti-YNU-1 exhibited higher activity in epoxidation of C5 –C8 cycloalkenes at 60 ◦ C than three-dimensional Ti-MWW (3D Ti-MWW, Si/Ti = 45) and Ti-BEA (Si/Ti = 35) although it had a Si/Ti ratio as high as 240 [21], demonstrating the advantages of layered
http://dx.doi.org/10.1016/j.cattod.2014.07.002 0920-5861/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model CATTOD-9149; No. of Pages 7
ARTICLE IN PRESS J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
2
catalysts. Subsequently, Kim et al. prepared a Ti-MCM-36 catalyst by swelling the Ti-MCM-22P layers with a surfactant and pillaring it with mesoporous silica. This material exhibited more than three times higher TON (Ti-MCM-36: 257, TS-1: 81) than the conventional TS-1 in 1-hexene epoxidation at 45 ◦ C [22]. Wang et al. [23] reported the preparation of interlayer-expanded Ti-MWW (Ti-IEZ-MWW), which was formed using diethoxydimethylsilane. The material expresses an order of magnitude higher TON (TiIEZ-MWW: 233, 3D Ti-MWW: 33) compared with 3D Ti-MWW in epoxidation of cyclohexene at 60 ◦ C. In this contribution, we present the synthesis, properties and catalytic performance of pillared TS-1 catalysts combining the approach of Ryoo and Kim [18,22]. Parent layered TS-1 was prepared according to the Ryoo protocol. To enhance the interlayer void volume of the layered material, the pillaring treatment was applied. Formation of either silica or a mixture of silica and titania pillars between the crystalline layers helps to keep them apart when the surfactant template is removed by calcination and to improve accessibility to the active centres. Ti-MCM-36 material is included as a material formed in the same manner but possessing a different topology of the crystalline layers. The materials are compared with conventional 3D TS-1 as catalysts in selective epoxidation with hydrogen peroxide as the oxidant. Cyclooctene, norbornene, ␣-pinene and verbenol are used as model substrates. 2. Experimental 2.1. Synthesis of two-dimensional TS-1 The syntheses of parent-layered TS-1 was carried out according to the reported procedure [18] from silicon(IV) ethoxide (TEOS; Aldrich 98%) and titanium(IV) butoxide (TBOTi; Aldrich 97%) using a surfactant template C18 H37 N+ (CH3 )2 C6 H12 N+ (CH3 )2 C6 H13 in the hydroxide form (C18-6-6 OH2 ; prepared as described in the literature [18]). Typically, TBOTi was added drop-wise into TEOS and stirred for 30 min. Then an aqueous solution of the template was added to the mixture and homogenized at 60 ◦ C for 3 h. Evaporated ethanol was replaced with the same mass amount of water. Synthesis mixture of composition 100 TEOS: 2.5 TBOTi: 6 C18-6-6 OH2 : 5000 H2 O was hydrothermally crystallized in a Teflon-lined autoclave at 160 ◦ C for 236 h under agitation. After the given time, the zeolite was filtered off, washed with water, dried at 80 ◦ C and finally calcined or subjected to the pillaring treatment. Calcination was carried out at 570 ◦ C for 8 h, using a temperature ramp of 1 ◦ C/min. No swelling of the material was necessary as the layers are swollen intrinsically by the surfactant template. The pillaring was performed as per the Kim procedure [22]. TEOS or a mixture of TEOS and TBOTi in a given Si/Ti ratio was mixed with the as-synthesized dry layered zeolite (10 g of the TEOS mixture per 1 g of the zeolite) and stirred at 65 ◦ C for 24 h. Then the mixture was centrifuged and the solid material was dried for 48 h at room temperature. Subsequently, the product was hydrolysed in water with 5% of ethanol (100 ml/1 g) at ambient temperature for 24 h under vigorous stirring. Ethanol in the mixture helps to disperse the hydrophobic titanosilicate. Finally, the solid material was centrifuged again, dried at 65 ◦ C and calcined in an airflow at 550 ◦ C for 10 h using the temperature ramp of 2 ◦ C/min. 2.2. Synthesis of Ti-MCM-36 The synthesis followed the procedure described by Tatsumi [20]. 26.1 g of piperidine (Aldrich, 99%) was added to 73.3 g of distilled water and stirred for 15 min. The solution was separated into two parts and 2.57 g of TBOTi (Si/Ti 30) was added into one part and
17.74 g of H3 BO3 (Fluka, 99%, Si/B 0.75) was added to the other part. The solutions were stirred for about 50 min. 12.86 g of Cab-OSil M5 (Havel Composites, Czech Republic) was divided into two parts and each was slowly added into the solution and stirred for another 60 min. Then, both slurries were mixed together and stirred for next 60 min. The final gel was charged into Teflon-lined autoclave. The crystallization proceeded under agitation and autogenous pressure at 130 ◦ C for first 24 h, and then the temperature was increased to 150 ◦ C. After 24 h, the temperature was increased to 170 ◦ C for the final 5 days. Solid product was collected by filtration, washed with distilled water and dried in the oven at 60 ◦ C overnight. The as-synthesized Ti-MCM-22(P) was stirred at 85 ◦ C with 2 M HNO3 for 15 h (20 ml/1 g) to remove extra-framework Ti species. The Ti-MCM-22(P) was swollen using 20 ml/1 g of 25% solution of cetyltrimenthylammonium hydroxide (C16 TMA-OH, prepared by ion-exchange from chloride form). The slurry was stirred for 16 h at ambient temperature. The product Ti-MCM-22SW was separated by centrifugation, washed with water and dried at 60 ◦ C. Pillaring of Ti-MCM-22SW was carried out with TEOS (50 ml/1 g). The mixture was stirred at 85 ◦ C for 16 h. The solid product was isolated by centrifugation and dried at ambient temperature. Subsequently, the material was hydrolysed in water (100 ml/1 g) at ambient temperature for 24 h. Finally, the product was centrifuged again, dried at 60 ◦ C and calcined in the airflow at 550 ◦ C for 10 h using the temperature ramp of 2 ◦ C/min.
2.3. Synthesis of three-dimensional TS-1 Conventional titanosilicate TS-1 (3D TS-1) was prepared from a gel with an initial Si/Ti ratio of 40 according to the procedure described in ref. [19] using tetraethyl orthotitanate (Aldrich, technical grade) and tetraethyl orthosilicate with tetrapropylammonium hydroxide (Aldrich, 20% in water) as a structure directing agent.
2.4. Characterization X-ray powder diffraction (XRD) patterns were collected using a Bruker AXS D8 Advance diffractometer equipped with a graphite monochromator and a position-sensitive detector Våntec-1 using Cu K␣ radiation in Bragg–Brentano geometry. Data were collected in continuous mode over the 2 range of 1–40◦ . The size and shape of zeolite crystals were examined by scanning electron microscopy (SEM) on a JEOL, JSM-5500LV microscope. The images were collected with an acceleration voltage of 20 kV. Nitrogen sorption isotherms were measured at liquid nitrogen temperature (−196 ◦ C) with Micromeritics Gemini volumetric instrument. Prior to the sorption measurements, individual zeolites were outgassed in a stream of helium at 300 ◦ C for 3 h. BET area was evaluated using adsorption data in the range of a relative pressure from p/p0 = 0.05 to 0.25. The t-plot method [24] was applied to determine the volume of micropores (Vmicro ) and external surface area (Sext ). The adsorbed amount of nitrogen at p/p0 = 0.95 reflects the total adsorption capacity (Vtotal ). DR-UV/Vis absorption spectra were collected using PerkinElmer Lambda 950 Spectrometer with a 5 or 2 mm quartz tube and a large 8 × 16 mm2 slit. The data were collected in the wavelength range of 190–500 nm. All the samples were analysed after calcination. Chemical composition of the materials (expressed hereafter as a Si/Ti ratio) was determined by XRF with a spectrometer Philips PW 1404 using an analytical program UniQuant. The samples were mixed with dentacryl as a binder and pressed on the surface of cellulose pellets.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model
ARTICLE IN PRESS
CATTOD-9149; No. of Pages 7
J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 1. Comparison of XRD patterns of as-synthesized, calcined and pillared TS-1 and Ti-MCM-36 catalysts.
2.5. Catalytic reactions The catalytic experiments were carried out in a Heidolph Synthesis 1 system at 60 ◦ C with alkene/catalyst mass ratio 10 and alkene/H2 O2 molar ratio 2. Acetonitrile (Fisher chemical, HPLC grade) was used as a solvent and mesitylene served as an internal standard. The shaking speed was 610 rpm. Typically 500 mg of the substrate [cis-cyclooctene (99%, Aldrich), norbornene (99%, Aldrich), ␣-pinene (Acros Organics, 95%) or verbenol (95%, Aldrich)] was mixed with 250 mg of the internal standard, 50 mg of the catalyst and 8 ml of acetonitrile. The reaction was started by adding H2 O2 aqueous solution (35 wt%, Aldrich) into the mixture. Samples were taken in regular intervals, centrifuged, cooled and analysed using an Agilent 6850 GC system with 50 m long DB-5 column, an autosampler and an FID or an MS detector. Helium was used as a carrier gas. 3. Results and discussion 3.1. Catalyst synthesis, morphology and texture The layered TS-1 material was prepared according to the Ryoo procedure. The XRD pattern is presented in Fig. 1. It is consistent with the XRD pattern presented by Ryoo for nanosheet TS-1 [18]. The first diffraction line at 2 = 1.4◦ characterizes the distance between the zeolite layers. Ryoo observed a similar position of the line (2 = 1.45◦ ) when preparing an aluminosilicate nanosheet MFI
3
zeolite according to the same synthetic protocol [25]. In the assynthesized material, the layers are kept apart regularly by the surfactant template and after calcination they stack to each other randomly causing the disappearance of the low 2 diffraction line (Fig. 1). During the pillaring treatment, the interlayer space is filled with TEOS, forming amorphous mesoporous silica pillars after hydrolysis and calcination. The pillars keep the layers apart even after template removal and thus also the diffraction line between 2 = 1.7◦ and 2.0◦ is still present. The shift of the diffraction line from 2 = 1.4◦ (as-synthesized sample) to 2 = 1.7◦ (calcined pillared sample) during calcination reflects a slight change in the interlayer distance when the template is removed by calcination and the layers are kept apart only by silica pillars. The position of the line is corresponding to a d-spacing of 4.5–5.2 nm and an interlayer distance of 2.5–3 nm (the thickness of the TS-1 layer is approximately 2 nm [18]). The textural properties of all discussed samples are listed in Table 1. Layered TS-1 samples possess BET surface areas between 473 and 522 m2 /g and total adsorption capacities (Vtot ) of 0.44–0.51 cm3 /g. The total adsorption capacity is much higher in comparison with 3D TS-1 (Vtot = 0.16 cm3 /g). Similarly, the external surface area (Sext ) is increased for layered TS-1 (226–293 m2 /g) compared with 3D TS-1 (151 m2 /g). The pillaring treatment with TEOS significantly increases the BET surface area (up to 595 m2 /g) as well as the external surface area (up to 384 m2 /g). On the other hand, the total adsorption capacity of the material remains similar (layered TS-1a 0.44 cm3 /g vs. pillared TS-1a 0.46 cm3 /g). This corresponds to the fact that although the interlayer space is in fact larger for the pillared material than for the layered material, it is partially filled with silica pillars causing an increase in the surface area but a slight decrease in the adsorption volume. This behaviour is slightly different in comparison to Ti-MCM-22 vs. Ti-MCM-36 reported by Kim et al. [22], where Ti-MCM-22 is a fully connected 3D-zeolite and therefore there are only micropores between the layers lying perfectly organized one on another and total adsorption capacity is strongly increased by the swelling and pillaring. The organization into 3D TS-1 does not occur during calcination of layered TS-1. SEMs in Fig. 2 show flake-like crystals forming agglomerates for both layered TS-1 and pillared TS-1. In the case of the pillared material, the agglomerates are partly covered with amorphous silica coming from the excess of the TEOS used for the pillaring. The pillaring treatment with TEOS has a disadvantage in diluting the titanosilicate catalytically active phase (see an increase in the Si/Ti ratio between layered and pillared samples: layered TS1a Si/Ti = 36 vs. pillared TS-1a Si/Ti = 93; layered TS-1b Si/Ti = 33 vs. pillared TS-1b Si/Ti = 55). We tried to equalize the Ti content by addition of TBOTi into the pillaring mixture. The TBOTi can react either with the TEOS forming active centres in the pillars or with
Table 1 Textural properties and Ti content of the TS-1 and Ti-MCM-36 catalysts. Titanosilicatea
BET (m2 /g)
Sext (m2 /g)
Vmic (cm3 /g)
Vtot (cm3 /g)
Si/Ti
3D TS-1 Layered TS-1a Layered TS-1b Pillared TS-1a Pillared TS-1b Ti-pillared TS-1b(10)b Ti-pillared TS-1a(20)b Ti-pillared TS-1b(20)b Ti-pillared TS-1b(60)b Ti-MCM-36
450 473 511 595 575 540 480 634 685 786
151 293 226 384 337 323 239 402 524 629
0.13 0.09 0.13 0.09 0.11 0.10 0.11 0.10 0.07 0.07
0.16 0.44 0.51 0.46 0.39 0.33 0.29 0.37 0.37 0.44
44 36 33 93 55 14 23 20 30 57
a b
Samples marked (a) and (b) were prepared from the same batch of parent nanosheet TS-1. Numbers in brackets represent Si/Ti ratio in the pillaring mixture.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model CATTOD-9149; No. of Pages 7 4
ARTICLE IN PRESS J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 2. SEM images of the layered (left) and pillared (right) TS-1.
silanol groups on the surface of the layers practically impregnating them with additional Ti. Initially, Si/Ti ratio 20 in the pillaring mixture was used (Ti-pillared TS-1 (20a)). The Ti content was higher than in the parent material (Si/Ti 23 vs. 36) but the BET surface area as well as the total adsorption capacity decreased significantly in comparison with TEOS-pillared TS-1 (BET: 480 vs. 595 m2 /g; Vtot : 0.29 vs. 0.46 cm3 /g). The nitrogen sorption isotherms are presented in Fig. 4. However, the catalytic performance improved dramatically (vide infra). Encouraged by the success, we have prepared a series of Tipillared materials with different Ti content in the pillaring mixture (namely XTBOTi = 0.1, XTBOTi = 0.5 and XTBOTi = 0.0167 in Ti-pillared TS-1b(10), (20) and (60), respectively). The XRD patterns are presented in Fig. 3. No change in the XRD pattern was observed even when high Ti loading was used (XTBOTi = 0.1). A trend in the dependence of BET surface area on Ti content in the pillaring mixture can be observed in Fig. 5. The horizontal line marks the BET surface area of a non-pillared material. The higher Ti content in the pillaring mixture is used, the lower BET surface area possesses the product. We ascribe this behaviour to the fact that TBOTi is more reactive [26] than TEOS and therefore it supports formation of thicker pillars than pure TEOS (the silica-TiO2 phase between the layers is denser). A comparison of the isotherms is presented in Fig 4. Not only the titanium content but also the character of the titanium species simultaneously determines the performance of the catalyst [27]. It is generally assumed that isolated framework
Fig. 3. XRD patterns of Ti-pillared TS-1 samples compared to TEOS-pillared TS-1 and 3D TS-1.
tetrahedrally coordinated titanium species are the active sites catalysing the epoxidation [28]. The DR-UV spectra of the catalysts are presented in Fig. 6. The most intensive band in all the spectra is centred at 220 nm and in (b) series of the samples, another band at 260 nm can be observed. Zecchina et al. [29] ascribed these bands to tetrahedrally coordinated framework “TiO4 ” species and octahedrally coordinated extra-framework “TiO6 ” species, respectively. Anatase-like TiO2 phase absorbing at 330 nm may cause an unproductive decomposition of the oxidant [3]. However, no band centred at 330 nm is observed in any of the spectra, indicating that the formation of the anatase-like TiO2 phase did not occur. The intensity of the band at 260 nm is growing with increasing Ti content in the pillaring mixture, as expected, due to impregnation of the material with extra-framework Ti species either on the surface of the catalytic
Fig. 4. Nitrogen sorption isotherms of the layered and pillared titanosilicates: Tipillared TS-1a(20) (a); pillared TS-1a (b); layered TS-1a (c); layered TS-1b (d); Tipillared TS-1b(10) (e); Ti-pillared TS-1b(20) (f); Ti-pillared TS-1b(60) (g). Empty points denote desorption.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model CATTOD-9149; No. of Pages 7
ARTICLE IN PRESS J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
5
Table 2 Epoxidation of cyclooctene with H2 O2 at 60 ◦ C over different catalysts; conversion and yield after 4 h of the reaction. Catalyst
Conversion (%)
Epoxide yield (%)
S(20) a (%)
3D TS-1 Layered TS-1a Pillared TS-1b Ti-pillared TS-1b(10) Ti-pillared TS-1a(20) Ti-pillared TS-1b(20) Ti-pillared TS-1b(60) Ti-MCM-36
8.0 3.7 6.5 19.1 19.3 21.0 18.6 18.0
3.0 2.8 3.5 15.3 10.0 16.9 14.0 8.0
42 75 58 80 50 76 76 54
a
Fig. 5. Ti-pillared TS-1 BET surface area dependence on Ti content in the pillaring mixture (series of the samples: grey a; black b). The horizontal line marks the BET surface area of a non-pillared material.
layers or inside the pillars; however, it is questionable, whether the titanium in the pillars may be active. 3.2. Catalytic tests The pillared TS-1 samples were investigated in cyclooctene, norbornene, ␣-pinene and verbenol epoxidation in a liquid phase using aqueous hydrogen peroxide as the oxidant in acetonitrile at 60 ◦ C. Acetonitrile was chosen as a solvent inert to reaction with formed epoxide and it suppresses possible intramolecular rearrangements
Selectivity at 20% conversion defined as [yield of epoxide]/[consumed alkene].
of substrates and products due to its weak basicity [14]. In order to confirm the activity of the catalysts, blank experiments without any catalyst were carried out. No conversion of any of the substrates was observed, which confirms the stability of the reaction mixtures without catalysts. The results of the cyclooctene epoxidation are summarized in Table 2. Cyclooctene underwent epoxidation over 3D TS-1; however, the yield of the epoxide as well as the selectivity of the process were low (yield 3.0% after 4 h, selectivity 42% at 20% conversion). Unexpectedly, the performance of the layered TS-1 was comparable to 3D TS-1 in terms of the yield (2.8% after 4 h); however, the layered TS-1 provided a higher selectivity (75% at 20% conversion). We assume that cyclooctene is oxidized due to the flexibility of its molecule, which is able to adopt such a conformation that fits into the TS-1 channels, although our observation is in contradiction with Ryoo’s observations [18]. When using pillared TS-1, the yield
Fig. 6. DR-UV spectra of the titanosilicate samples: 3D TS-1 (a), Ti-MCM-36 (b), layered TS-1a (c), Ti-pillared TS-1a(20) (d); layered TS-1b (e), pillared TS-1b (f), Ti-pillared TS-1b(10) (g), Ti-pillared TS-1b(20) (h), Ti-pillared TS-1b(60) (i).
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model
ARTICLE IN PRESS
CATTOD-9149; No. of Pages 7
J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
6
Table 4 Oxidation of ␣-pinene over titanosilicate catalysts at 60 ◦ C; conversion and yield after 4 h of the reaction. Catalyst
Conversion (%)
Epoxide yield (%)
S10% (%)a
TS-1 Pillared TS-1a Ti-pillared TS-1a(20) Ti-MCM-36
8.8 12.3 9.6
No epoxide 0.6 1.0 1.9
7 8 20
a
Selectivity at 10% conversion defined as [yield of epoxide]/[consumed alkene].
Table 5 Oxidation of verbenol to verbenone over titanosilicate catalysts at 60 ◦ C; conversion and yield after 4 h of the reaction.
Fig. 7. Development of the cyclooctene oxide yield in time over Ti-pillared TS-1b catalysts.
only slightly increased to 3.5% after 4 h, which may be accounted for the dilution of the active component. The performance of Ti-MCM36 was better providing 8% yield after 4 h at the similar selectivity (54% at 20% conversion). Ti-pillared TS-1 (20a) provided yield of 10% at the same time and with similar selectivity as well being the best performing catalyst of the (a) series. Even better results were achieved with catalysts of the (b) series providing yields of cyclooctene oxide of 15.3, 16.9 and 14% using Ti-pillared TS-1b(10), (20) and (60), respectively, after 4 h, with selectivity of 76–80% at 20% conversion. Development of the yield in time for these catalysts is presented in Fig. 7. The difference in the yield obtained using Ti-pillared TS-1a(20) and Ti-pillared TS-1b(20) can be explained as a pure manifestation of the influence of surface accessibility. The catalysts possess the same titanium content (Si/Ti (Ti-pillared TS-1a(20)) = 23, Si/Ti (Ti-pillared TS-1b(20)) = 20); however, they differ in the BET surface area and the external surface area as well. The ratio between the yield of cyclooctene oxide y4 h (Ti-pillared TS-1b(20))/y4 h (Tipillared TS-1a(20)) = 1.69 is the same as their external surface area ratio Sext (Ti-pillared TS-1b(20))/Sext (Ti-pillared TS-1a(20)) = 1.68. Having in mind the dependence of surface area on Ti content in the pillaring mixture, we conclude there are two factors contradicting each other and thus defining the catalyst performance. The higher the content of titanium is, the more active sites are present; however, the lower specific area possesses the material. It appears that the Si/Ti ratio 20 in the pillaring mixture is close to the optimum. The only by-product observed was 2-cycloocten-1-one. The highest yield of norbornene oxide was achieved with Ti-pillared TS-1b(20) (14.8% after 4.5 h; 56% selectivity at 10% conversion), while 3D TS-1 gave the epoxide yield only 1.8% (Table 3). The order of catalyst performance is similar to the case of cyclooctene 3D TS-1 (norbornene oxide yield 1.8% after
Table 3 Epoxidation of norbornene over titanosilicate catalysts at 60 ◦ C; conversion and yield after 4.5 h of the reaction. Catalyst
Conversion (%)
Epoxide yield (%)
S10% (%)a
3D TS-1 Layered TS-1b Pillared TS-1b Ti-pillared TS-1b(20) Ti-MCM-36
2.8 12.0 10.6 23.0 24.1
1.8 6.1 5.5 14.8 8.5
65b 42 52 56 26
a b
8 h.
Selectivity at 10% conversion defined as [yield of epoxide]/[consumed alkene]. Selectivity provided at 3% conversion; 10% conversion was not achieved until
Catalyst
Conversion (%)
Verbenone yield (%)
S10% (%)a
3D-TS-1 Layered TS-1a Ti-pillared TS-1a(20) Ti-MCM-36
30.6 46.2 10.4 30.2
24.7 44.8 10.2 14.2
71 97 98 32
a Selectivity at 10% conversion defined as [yield of verbenone]/[consumed verbenol].
4.5 h) < layered TS-1 (yield 6.1% after 4.5 h), pillared TS-1 (yield 5.5% after 4.5 h) < Ti-MCM-36 (yield 8.5% after 4.5 h) < Ti-pillared TS-1 (yield 14.8% after 4.5 h); however, the difference between 3D TS-1 and layered structures is more evident due to rigidity of the norbornene molecule. The performance of Ti-MCM-36 and Ti-pillared TS-1 exceeds the performance of Ti-CFI catalyst reported by our group previously (yield 5.8% after 4 h) [16]. Main oxidation byproduct observed is 3-hydroxy-norbornane-2-one. ␣-Pinene was oxidized using only layered catalysts because it is a bulky molecule that cannot access the narrow pores of the 3DTS-1 (Table 4). The best yield was obtained using the Ti-MCM-36 catalyst (1.9% after 4 h); however, the selectivity of the oxidation ranges from 7 to 20% at 10% conversion and many other oxidation products are observed including mainly hydroxyl and carbonyl compounds. This is in accordance with the findings of Eimer et al. using mesoporous titanosilicate prepared from TS-1 precursors [15]. Last but not least, verbenol was used as a substrate as it being one of the products of ␣-pinene oxidation. The results are summarized in Table 5. Verbenol was rapidly oxidized (conversion up to 46% after 4 h) with a high selectivity to the main product ranging 71–98% for TS-1 catalysts. However, the main product is not epoxide but verbenone. No verbenol epoxide was observed in any of the reactions. The 3D TS-1 catalyst was also active, providing higher conversion than Ti-pillared TS-1 (30.6% vs. 10.4% after 4 h). Tatsumi et al. showed that reaching the right position to the active site may be difficult for the substrates with trisubstituted double bond (e.g. 3-methyl-2-buten-1-ol) and this steric hindrance leads to the preferential oxidation of hydroxyl group if present and better accessible. Epoxide and aldehyde or ketone are formed concurrently in such cases [30]. Verbenol represents an extreme case when epoxidation reaction is suppressed and only hydroxyl group is oxidized. 4. Conclusions Pillared TS-1 catalysts were prepared from TS-1 nanosheets, using TEOS as a pillaring medium. The pillaring treatment helps preserve the mesoporosity of the material and keeps the active centres accessible even for bulky molecules. The disadvantage in the active phase dilution has been overcome by the addition of TBOTi into the pillaring mixture and thus boosting the Ti content in the material. On the other hand, addition of TBOTi causes a decrease in the surface area in comparison with TS-1 pillared with pure TEOS. Based on the catalytic testing of materials with different Ti content,
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model CATTOD-9149; No. of Pages 7
ARTICLE IN PRESS J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
we conclude that the optimum composition of the pillaring mixture is close to TEOS/TBOTi ratio 20. The pillared TS-1 and Ti-pillared TS-1 were active in cyclooctene, norbonene and ␣-pinene epoxidation with hydrogen peroxide as the oxidant under mild reaction conditions. The performance of Ti-pillared TS-1 catalysts exceeds 3D TS-1 as well as layered TS-1. Furthermore, a selective oxidation of verbenol to verbenone was observed under the reaction conditions used. In contrast to epoxidation reactions, layered TS-1 was the best performing catalyst for the verbenol oxidation. Acknowledgements The authors thank Dr. L. Brabec for SEM images, Dr. G. Sádovská for UV/Vis measurements and Dr. Martin Lamaˇc for NMR analyses. The authors acknowledge the Czech Science Foundation (P106/12/G015) and the NSTIP strategic technologies programs (12-(PET2614)02) in the Kingdom of Saudi Arabia for the support of this research. References ˇ [1] J. Cejka, A. Corma, S.I. Zones (Eds.), Zeolites and Catalysis: Synthesis, Reactions and Applications, Wiley-VCH, Weinheim, 2010. ˇ [2] J. Cejka, G. Centi, J. Perez-Pariente, W.J. Roth, Catal. Today 179 (2012) 2–15. [3] T. Tatsumi, Metal-substituted zeolites as heterogenous oxidation catalysts, in: N. Mizuno (Ed.), Modern Heterogenous Oxidation Catalysis, Wiley-WCH, Weinheim, 2009, pp. 125–153. [4] A. Corma, P. Esteve, A. Martínez, S. Valencia, J. Catal. 152 (1995) 18–24. [5] T. Tatsumi, M. Nakamura, S. Negishi, H. Tominaga, J. Chem. Soc. Chem. Commun. 476 (1990).
7
[6] G. Peregot, G. Bellussi, C. Corno, M. Taramasso, F. Buonomot, A. Esposito, In: Y. Murakami, A., Iijima, J.W. Ward (Eds.), Stud. Surf. Sci. Catal. 28 (1986) 129–136. [7] P. Birke, P. Kraak, R. Schödel, F. Vogt, in: T. Hattori, T. Yashima (Eds.) Stud. Surf. Sci. Catal. 83 (1994) 425–432. [8] J.S. Reddy, P.A. Jacobs, Catal. Lett. 37 (1996) 213–216. [9] V. Hulea, F. Fajula, J. Bousquet, J. Catal. 198 (2001) 179. [10] R.A. Sheldon, J. Dakka, Catal. Today 19 (1994) 215–246. [11] C. Perego, A. Carati, P. Ingallina, M.A. Mantegazza, G. Bellussi, Appl. Catal. A 221 (2001) 63–72. [12] M. Tamarasso, G. Perego, B. Notari, US Patent 4410501, 1983. [13] http://www.eni.com/en IT/attachments/azienda/attivita-strategie/ petrolchimica/licensing/TS1-flyer-lug09.pdf (22 May 2014). [14] J.C. van der Waal, M.S. Rigutto, H. van Bekkum, Appl. Catal. A 167 (1998) 331–342. [15] G.A. Eimer, I. Diaz, E. Sastre, S.G. Casuscelli, M.E. Crivello, E.R. Herreroa, J. PerezPariente, Appl. Catal. A 343 (2008) 77–86. ˇ ˚ J. Cejka, [16] J. Pˇrech, M. Kubu, Catal. Today 227 (2014) 80–86. ˇ [17] W.J. Roth, P. Nachtigall, R.E. Morris, J. Cejka, Chem. Rev. 114 (2014) 4807–4837. [18] K. Na, Ch. Jo, J. Kim, W. Ahn, R. Ryoo, ACS Catal. 1 (2011) 901–907. [19] M. Taramasso, G. Perego, B. Notari, in: H. Robson (Ed.), Verified Syntheses of Zeolitic Materials, Elsevier, Amsterdam, 2001, p. 207. [20] P. Wu, T. Tatsumi, T. Komatsu, T. Yashima, J. Phys. Chem. B 105 (2001) 2897–2905. [21] W. Fan, P. Wu, S. Namba, T. Tatsumi, Angew. Chem. Int. Ed. 116 (2004) 238–242. [22] S. Kim, H. Ban, W. Ahn, Catal. Lett. 113 (2007) 160–164. [23] L. Wang, Y. Wang, Y. Liu, H. Wu, X. Li, M. He, P. Wu, J. Mater. Chem. 19 (2009) 8594–8602. [24] B.C. Lippens, J.H. de Boer, J. Catal. 4 (1965) 319. [25] K. Na, M. Choi, W. Park, Y. Sakamoto, O. Terasaki, R. Ryoo, J. Am. Chem. Soc. 132 (2010) 4169–4177. [26] A. Thangaraj, S. Sivasanker, J. Chem. Soc. Chem. Commun. 123 (1992). [27] B. Notari, Adv. Catal. 41 (1996) 253. [28] F. Bonito, A. Damin, G. Ricchiardi, M. Ricci, G. Spano, R. D’Aloisio, A. Zecchina, C. Lamberti, C. Prestipino, S. Bordiga, J. Phys. Chem. B 108 (2004) 3573–3583. [29] A. Zecchina, G. Spoto, S. Bordiga, A. Ferrero, G. Petrini, G. Leofanti, M. Padovan, in: P.A. Jacobs, N.I. Jaeger, L. Kubelková, B. Wichterlová (Eds.), Stud. Surf. Sci. Catal. 69 (1991) 251–258. [30] T. Tatsumi, M. Yako, M. Nakamura, Y. Yuhara, H. Tominaga, J. Mol. Catal. 78 (1993) L41–L45.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
ARTICLE IN PRESS
CATTOD-9149; No. of Pages 7
Catalysis Today xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today journal homepage: www.elsevier.com/locate/cattod
Epoxidation of bulky organic molecules over pillared titanosilicates Jan Pˇrech a,∗ , Pavla Eliáˇsová a , Daifallah Aldhayan b , Martin Kubu˚ a a b
J. Heyrovsk´ y Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, v.v.i., Dolejˇskova 3, CZ-182 23 Prague 8, Czech Republic Department of Chemistry, College of Science, King Saud University, B.O. Box 2455, Riyadh 11451, Saudi Arabia
a r t i c l e
i n f o
Article history: Received 4 June 2014 Received in revised form 27 June 2014 Accepted 3 July 2014 Available online xxx Keywords: Layered TS-1 zeolite Pillaring Epoxidation Cyclooctene Norbornene Verbenol.
a b s t r a c t Layered and pillared TS-1 catalysts were prepared using surfactant C18 H37 –N+ (CH3 )2 C6 H12 –N+ (CH3 )2 C6 H13 in the hydroxide form as a structure directing agent and silicon(IV) ethoxide or its mixture with titanium(IV) butoxide as a pillaring medium. The pillaring treatment significantly increases the specific surface area (BET up to 685 m2 /g) and thus the accessibility of active centres for bulky molecules. The addition of tetrabutylorthotitanate into the pillaring mixture has a positive effect on the material performance as a catalyst of cyclooctene, norbornene and ␣-pinene epoxidation under mild conditions, with hydrogen peroxide as the oxidant. Optimum composition of the pillaring medium is Si/Ti ratio 20. The highest yields of cyclooctene oxide (16.9% after 4 h) and norbonene oxide (14.8% after 4.5 h) were obtained using the above-prepared catalyst at 60 ◦ C with substrate/catalyst mass ratio being 10 and substrate/H2 O2 ratio being 2. Selective oxidation of verbenol to verbenone was observed instead of the epoxidation with the above catalysts. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Zeolites represent a rich family of porous crystalline aluminosilicates successfully applied as catalysts, adsorbents and ion exchangers [1,2]. Originally, they are powerful acidobasic catalysts; however, incorporation of other heteroelements (e.g. titanium, boron, iron, vanadium, tin, germanium and others) into the crystalline framework opened other areas of catalysis [3]. Titaniumcontaining zeolites are able to catalyse various oxidation reactions, e.g., epoxidation of olefins [4], oxidation of alkanes to alcohols and ketones [5], oxidation of aromatic hydrocarbons to phenols [6], ammoxidation of cyclohexanone [7], oxidation of amines to hydroxylamines [8] and thioethers to sulfoxides [9]. Furthermore, these catalysts have strong advantages in comparison with amorphous materials and other conventional oxidation catalysts: (i) they possess a regular microenvironment around the active centres [10]; (ii) the active sites are resistant to deactivation via oligomerization of metal-oxide species [10]; (iii) the materials exhibit enhanced stability towards leaching of the metal in comparison with the amorphous materials and other conventional oxidation catalysts [10]. TS-1 (MFI topology), Ti-BEA and Ti-MWW are the most frequently studied titanosilicates [3]. The TS-1 zeolite is currently a catalyst of choice for epoxidation of C C double bonds with
∗ Corresponding author. Tel.: +420 266 053 856. E-mail address: [email protected] (J. Pˇrech).
hydrogen peroxide providing low waste and being environmentally friendly [11]. Its discovery [12] represents a considerable breakthrough in the field of oxidation catalysis. Presently, the TS1 is industrially applied by Eni in a propylenoxide process [13]. However, TS-1 has pores of 5.5 A˚ in diameter. Therefore, bulky substrates like terpenes cannot easily access the active sites located mainly inside the pores. Although the epoxidation of terpenes over titanosilicates has been given a certain attention, it still remains a challenge [14–16]. The application of layered materials for catalysis offers the possibility to overcome the diffusion problems [17,18] as layered materials possess high external surface areas in comparison to conventional three-dimensional titanosilicates. Ryoo and co-workers [18] reported the preparation of specially designed TS-1 in the form of nanosheets. They are formed using surfactant-based template (C18 H37 N+ (CH3 )2 C6 H12 N+ (CH3 )2 C6 H13 ·OH2 − ) instead of conventional tetrapropylammonium hydroxide [19]. This material has been reported to exhibit activity similar to bulk TS-1 in epoxidation of 1-hexene; however, its activity is an order of magnitude higher in epoxidation of cyclooctene [18]. Wu et al. [20] succeeded in the preparation of Ti-MWW combining a synthesis of Ti-MWW lamellar precursor (denoted as Ti-MCM-22P) in the presence of boron with acidic treatment of the material to remove extra-framework Ti species prior to calcination. Later, direct synthesis of the Ti-MWW lamellar precursor Ti-YNU-1 was reported. Ti-YNU-1 exhibited higher activity in epoxidation of C5 –C8 cycloalkenes at 60 ◦ C than three-dimensional Ti-MWW (3D Ti-MWW, Si/Ti = 45) and Ti-BEA (Si/Ti = 35) although it had a Si/Ti ratio as high as 240 [21], demonstrating the advantages of layered
http://dx.doi.org/10.1016/j.cattod.2014.07.002 0920-5861/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model CATTOD-9149; No. of Pages 7
ARTICLE IN PRESS J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
2
catalysts. Subsequently, Kim et al. prepared a Ti-MCM-36 catalyst by swelling the Ti-MCM-22P layers with a surfactant and pillaring it with mesoporous silica. This material exhibited more than three times higher TON (Ti-MCM-36: 257, TS-1: 81) than the conventional TS-1 in 1-hexene epoxidation at 45 ◦ C [22]. Wang et al. [23] reported the preparation of interlayer-expanded Ti-MWW (Ti-IEZ-MWW), which was formed using diethoxydimethylsilane. The material expresses an order of magnitude higher TON (TiIEZ-MWW: 233, 3D Ti-MWW: 33) compared with 3D Ti-MWW in epoxidation of cyclohexene at 60 ◦ C. In this contribution, we present the synthesis, properties and catalytic performance of pillared TS-1 catalysts combining the approach of Ryoo and Kim [18,22]. Parent layered TS-1 was prepared according to the Ryoo protocol. To enhance the interlayer void volume of the layered material, the pillaring treatment was applied. Formation of either silica or a mixture of silica and titania pillars between the crystalline layers helps to keep them apart when the surfactant template is removed by calcination and to improve accessibility to the active centres. Ti-MCM-36 material is included as a material formed in the same manner but possessing a different topology of the crystalline layers. The materials are compared with conventional 3D TS-1 as catalysts in selective epoxidation with hydrogen peroxide as the oxidant. Cyclooctene, norbornene, ␣-pinene and verbenol are used as model substrates. 2. Experimental 2.1. Synthesis of two-dimensional TS-1 The syntheses of parent-layered TS-1 was carried out according to the reported procedure [18] from silicon(IV) ethoxide (TEOS; Aldrich 98%) and titanium(IV) butoxide (TBOTi; Aldrich 97%) using a surfactant template C18 H37 N+ (CH3 )2 C6 H12 N+ (CH3 )2 C6 H13 in the hydroxide form (C18-6-6 OH2 ; prepared as described in the literature [18]). Typically, TBOTi was added drop-wise into TEOS and stirred for 30 min. Then an aqueous solution of the template was added to the mixture and homogenized at 60 ◦ C for 3 h. Evaporated ethanol was replaced with the same mass amount of water. Synthesis mixture of composition 100 TEOS: 2.5 TBOTi: 6 C18-6-6 OH2 : 5000 H2 O was hydrothermally crystallized in a Teflon-lined autoclave at 160 ◦ C for 236 h under agitation. After the given time, the zeolite was filtered off, washed with water, dried at 80 ◦ C and finally calcined or subjected to the pillaring treatment. Calcination was carried out at 570 ◦ C for 8 h, using a temperature ramp of 1 ◦ C/min. No swelling of the material was necessary as the layers are swollen intrinsically by the surfactant template. The pillaring was performed as per the Kim procedure [22]. TEOS or a mixture of TEOS and TBOTi in a given Si/Ti ratio was mixed with the as-synthesized dry layered zeolite (10 g of the TEOS mixture per 1 g of the zeolite) and stirred at 65 ◦ C for 24 h. Then the mixture was centrifuged and the solid material was dried for 48 h at room temperature. Subsequently, the product was hydrolysed in water with 5% of ethanol (100 ml/1 g) at ambient temperature for 24 h under vigorous stirring. Ethanol in the mixture helps to disperse the hydrophobic titanosilicate. Finally, the solid material was centrifuged again, dried at 65 ◦ C and calcined in an airflow at 550 ◦ C for 10 h using the temperature ramp of 2 ◦ C/min. 2.2. Synthesis of Ti-MCM-36 The synthesis followed the procedure described by Tatsumi [20]. 26.1 g of piperidine (Aldrich, 99%) was added to 73.3 g of distilled water and stirred for 15 min. The solution was separated into two parts and 2.57 g of TBOTi (Si/Ti 30) was added into one part and
17.74 g of H3 BO3 (Fluka, 99%, Si/B 0.75) was added to the other part. The solutions were stirred for about 50 min. 12.86 g of Cab-OSil M5 (Havel Composites, Czech Republic) was divided into two parts and each was slowly added into the solution and stirred for another 60 min. Then, both slurries were mixed together and stirred for next 60 min. The final gel was charged into Teflon-lined autoclave. The crystallization proceeded under agitation and autogenous pressure at 130 ◦ C for first 24 h, and then the temperature was increased to 150 ◦ C. After 24 h, the temperature was increased to 170 ◦ C for the final 5 days. Solid product was collected by filtration, washed with distilled water and dried in the oven at 60 ◦ C overnight. The as-synthesized Ti-MCM-22(P) was stirred at 85 ◦ C with 2 M HNO3 for 15 h (20 ml/1 g) to remove extra-framework Ti species. The Ti-MCM-22(P) was swollen using 20 ml/1 g of 25% solution of cetyltrimenthylammonium hydroxide (C16 TMA-OH, prepared by ion-exchange from chloride form). The slurry was stirred for 16 h at ambient temperature. The product Ti-MCM-22SW was separated by centrifugation, washed with water and dried at 60 ◦ C. Pillaring of Ti-MCM-22SW was carried out with TEOS (50 ml/1 g). The mixture was stirred at 85 ◦ C for 16 h. The solid product was isolated by centrifugation and dried at ambient temperature. Subsequently, the material was hydrolysed in water (100 ml/1 g) at ambient temperature for 24 h. Finally, the product was centrifuged again, dried at 60 ◦ C and calcined in the airflow at 550 ◦ C for 10 h using the temperature ramp of 2 ◦ C/min.
2.3. Synthesis of three-dimensional TS-1 Conventional titanosilicate TS-1 (3D TS-1) was prepared from a gel with an initial Si/Ti ratio of 40 according to the procedure described in ref. [19] using tetraethyl orthotitanate (Aldrich, technical grade) and tetraethyl orthosilicate with tetrapropylammonium hydroxide (Aldrich, 20% in water) as a structure directing agent.
2.4. Characterization X-ray powder diffraction (XRD) patterns were collected using a Bruker AXS D8 Advance diffractometer equipped with a graphite monochromator and a position-sensitive detector Våntec-1 using Cu K␣ radiation in Bragg–Brentano geometry. Data were collected in continuous mode over the 2 range of 1–40◦ . The size and shape of zeolite crystals were examined by scanning electron microscopy (SEM) on a JEOL, JSM-5500LV microscope. The images were collected with an acceleration voltage of 20 kV. Nitrogen sorption isotherms were measured at liquid nitrogen temperature (−196 ◦ C) with Micromeritics Gemini volumetric instrument. Prior to the sorption measurements, individual zeolites were outgassed in a stream of helium at 300 ◦ C for 3 h. BET area was evaluated using adsorption data in the range of a relative pressure from p/p0 = 0.05 to 0.25. The t-plot method [24] was applied to determine the volume of micropores (Vmicro ) and external surface area (Sext ). The adsorbed amount of nitrogen at p/p0 = 0.95 reflects the total adsorption capacity (Vtotal ). DR-UV/Vis absorption spectra were collected using PerkinElmer Lambda 950 Spectrometer with a 5 or 2 mm quartz tube and a large 8 × 16 mm2 slit. The data were collected in the wavelength range of 190–500 nm. All the samples were analysed after calcination. Chemical composition of the materials (expressed hereafter as a Si/Ti ratio) was determined by XRF with a spectrometer Philips PW 1404 using an analytical program UniQuant. The samples were mixed with dentacryl as a binder and pressed on the surface of cellulose pellets.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model
ARTICLE IN PRESS
CATTOD-9149; No. of Pages 7
J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 1. Comparison of XRD patterns of as-synthesized, calcined and pillared TS-1 and Ti-MCM-36 catalysts.
2.5. Catalytic reactions The catalytic experiments were carried out in a Heidolph Synthesis 1 system at 60 ◦ C with alkene/catalyst mass ratio 10 and alkene/H2 O2 molar ratio 2. Acetonitrile (Fisher chemical, HPLC grade) was used as a solvent and mesitylene served as an internal standard. The shaking speed was 610 rpm. Typically 500 mg of the substrate [cis-cyclooctene (99%, Aldrich), norbornene (99%, Aldrich), ␣-pinene (Acros Organics, 95%) or verbenol (95%, Aldrich)] was mixed with 250 mg of the internal standard, 50 mg of the catalyst and 8 ml of acetonitrile. The reaction was started by adding H2 O2 aqueous solution (35 wt%, Aldrich) into the mixture. Samples were taken in regular intervals, centrifuged, cooled and analysed using an Agilent 6850 GC system with 50 m long DB-5 column, an autosampler and an FID or an MS detector. Helium was used as a carrier gas. 3. Results and discussion 3.1. Catalyst synthesis, morphology and texture The layered TS-1 material was prepared according to the Ryoo procedure. The XRD pattern is presented in Fig. 1. It is consistent with the XRD pattern presented by Ryoo for nanosheet TS-1 [18]. The first diffraction line at 2 = 1.4◦ characterizes the distance between the zeolite layers. Ryoo observed a similar position of the line (2 = 1.45◦ ) when preparing an aluminosilicate nanosheet MFI
3
zeolite according to the same synthetic protocol [25]. In the assynthesized material, the layers are kept apart regularly by the surfactant template and after calcination they stack to each other randomly causing the disappearance of the low 2 diffraction line (Fig. 1). During the pillaring treatment, the interlayer space is filled with TEOS, forming amorphous mesoporous silica pillars after hydrolysis and calcination. The pillars keep the layers apart even after template removal and thus also the diffraction line between 2 = 1.7◦ and 2.0◦ is still present. The shift of the diffraction line from 2 = 1.4◦ (as-synthesized sample) to 2 = 1.7◦ (calcined pillared sample) during calcination reflects a slight change in the interlayer distance when the template is removed by calcination and the layers are kept apart only by silica pillars. The position of the line is corresponding to a d-spacing of 4.5–5.2 nm and an interlayer distance of 2.5–3 nm (the thickness of the TS-1 layer is approximately 2 nm [18]). The textural properties of all discussed samples are listed in Table 1. Layered TS-1 samples possess BET surface areas between 473 and 522 m2 /g and total adsorption capacities (Vtot ) of 0.44–0.51 cm3 /g. The total adsorption capacity is much higher in comparison with 3D TS-1 (Vtot = 0.16 cm3 /g). Similarly, the external surface area (Sext ) is increased for layered TS-1 (226–293 m2 /g) compared with 3D TS-1 (151 m2 /g). The pillaring treatment with TEOS significantly increases the BET surface area (up to 595 m2 /g) as well as the external surface area (up to 384 m2 /g). On the other hand, the total adsorption capacity of the material remains similar (layered TS-1a 0.44 cm3 /g vs. pillared TS-1a 0.46 cm3 /g). This corresponds to the fact that although the interlayer space is in fact larger for the pillared material than for the layered material, it is partially filled with silica pillars causing an increase in the surface area but a slight decrease in the adsorption volume. This behaviour is slightly different in comparison to Ti-MCM-22 vs. Ti-MCM-36 reported by Kim et al. [22], where Ti-MCM-22 is a fully connected 3D-zeolite and therefore there are only micropores between the layers lying perfectly organized one on another and total adsorption capacity is strongly increased by the swelling and pillaring. The organization into 3D TS-1 does not occur during calcination of layered TS-1. SEMs in Fig. 2 show flake-like crystals forming agglomerates for both layered TS-1 and pillared TS-1. In the case of the pillared material, the agglomerates are partly covered with amorphous silica coming from the excess of the TEOS used for the pillaring. The pillaring treatment with TEOS has a disadvantage in diluting the titanosilicate catalytically active phase (see an increase in the Si/Ti ratio between layered and pillared samples: layered TS1a Si/Ti = 36 vs. pillared TS-1a Si/Ti = 93; layered TS-1b Si/Ti = 33 vs. pillared TS-1b Si/Ti = 55). We tried to equalize the Ti content by addition of TBOTi into the pillaring mixture. The TBOTi can react either with the TEOS forming active centres in the pillars or with
Table 1 Textural properties and Ti content of the TS-1 and Ti-MCM-36 catalysts. Titanosilicatea
BET (m2 /g)
Sext (m2 /g)
Vmic (cm3 /g)
Vtot (cm3 /g)
Si/Ti
3D TS-1 Layered TS-1a Layered TS-1b Pillared TS-1a Pillared TS-1b Ti-pillared TS-1b(10)b Ti-pillared TS-1a(20)b Ti-pillared TS-1b(20)b Ti-pillared TS-1b(60)b Ti-MCM-36
450 473 511 595 575 540 480 634 685 786
151 293 226 384 337 323 239 402 524 629
0.13 0.09 0.13 0.09 0.11 0.10 0.11 0.10 0.07 0.07
0.16 0.44 0.51 0.46 0.39 0.33 0.29 0.37 0.37 0.44
44 36 33 93 55 14 23 20 30 57
a b
Samples marked (a) and (b) were prepared from the same batch of parent nanosheet TS-1. Numbers in brackets represent Si/Ti ratio in the pillaring mixture.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model CATTOD-9149; No. of Pages 7 4
ARTICLE IN PRESS J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
Fig. 2. SEM images of the layered (left) and pillared (right) TS-1.
silanol groups on the surface of the layers practically impregnating them with additional Ti. Initially, Si/Ti ratio 20 in the pillaring mixture was used (Ti-pillared TS-1 (20a)). The Ti content was higher than in the parent material (Si/Ti 23 vs. 36) but the BET surface area as well as the total adsorption capacity decreased significantly in comparison with TEOS-pillared TS-1 (BET: 480 vs. 595 m2 /g; Vtot : 0.29 vs. 0.46 cm3 /g). The nitrogen sorption isotherms are presented in Fig. 4. However, the catalytic performance improved dramatically (vide infra). Encouraged by the success, we have prepared a series of Tipillared materials with different Ti content in the pillaring mixture (namely XTBOTi = 0.1, XTBOTi = 0.5 and XTBOTi = 0.0167 in Ti-pillared TS-1b(10), (20) and (60), respectively). The XRD patterns are presented in Fig. 3. No change in the XRD pattern was observed even when high Ti loading was used (XTBOTi = 0.1). A trend in the dependence of BET surface area on Ti content in the pillaring mixture can be observed in Fig. 5. The horizontal line marks the BET surface area of a non-pillared material. The higher Ti content in the pillaring mixture is used, the lower BET surface area possesses the product. We ascribe this behaviour to the fact that TBOTi is more reactive [26] than TEOS and therefore it supports formation of thicker pillars than pure TEOS (the silica-TiO2 phase between the layers is denser). A comparison of the isotherms is presented in Fig 4. Not only the titanium content but also the character of the titanium species simultaneously determines the performance of the catalyst [27]. It is generally assumed that isolated framework
Fig. 3. XRD patterns of Ti-pillared TS-1 samples compared to TEOS-pillared TS-1 and 3D TS-1.
tetrahedrally coordinated titanium species are the active sites catalysing the epoxidation [28]. The DR-UV spectra of the catalysts are presented in Fig. 6. The most intensive band in all the spectra is centred at 220 nm and in (b) series of the samples, another band at 260 nm can be observed. Zecchina et al. [29] ascribed these bands to tetrahedrally coordinated framework “TiO4 ” species and octahedrally coordinated extra-framework “TiO6 ” species, respectively. Anatase-like TiO2 phase absorbing at 330 nm may cause an unproductive decomposition of the oxidant [3]. However, no band centred at 330 nm is observed in any of the spectra, indicating that the formation of the anatase-like TiO2 phase did not occur. The intensity of the band at 260 nm is growing with increasing Ti content in the pillaring mixture, as expected, due to impregnation of the material with extra-framework Ti species either on the surface of the catalytic
Fig. 4. Nitrogen sorption isotherms of the layered and pillared titanosilicates: Tipillared TS-1a(20) (a); pillared TS-1a (b); layered TS-1a (c); layered TS-1b (d); Tipillared TS-1b(10) (e); Ti-pillared TS-1b(20) (f); Ti-pillared TS-1b(60) (g). Empty points denote desorption.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model CATTOD-9149; No. of Pages 7
ARTICLE IN PRESS J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
5
Table 2 Epoxidation of cyclooctene with H2 O2 at 60 ◦ C over different catalysts; conversion and yield after 4 h of the reaction. Catalyst
Conversion (%)
Epoxide yield (%)
S(20) a (%)
3D TS-1 Layered TS-1a Pillared TS-1b Ti-pillared TS-1b(10) Ti-pillared TS-1a(20) Ti-pillared TS-1b(20) Ti-pillared TS-1b(60) Ti-MCM-36
8.0 3.7 6.5 19.1 19.3 21.0 18.6 18.0
3.0 2.8 3.5 15.3 10.0 16.9 14.0 8.0
42 75 58 80 50 76 76 54
a
Fig. 5. Ti-pillared TS-1 BET surface area dependence on Ti content in the pillaring mixture (series of the samples: grey a; black b). The horizontal line marks the BET surface area of a non-pillared material.
layers or inside the pillars; however, it is questionable, whether the titanium in the pillars may be active. 3.2. Catalytic tests The pillared TS-1 samples were investigated in cyclooctene, norbornene, ␣-pinene and verbenol epoxidation in a liquid phase using aqueous hydrogen peroxide as the oxidant in acetonitrile at 60 ◦ C. Acetonitrile was chosen as a solvent inert to reaction with formed epoxide and it suppresses possible intramolecular rearrangements
Selectivity at 20% conversion defined as [yield of epoxide]/[consumed alkene].
of substrates and products due to its weak basicity [14]. In order to confirm the activity of the catalysts, blank experiments without any catalyst were carried out. No conversion of any of the substrates was observed, which confirms the stability of the reaction mixtures without catalysts. The results of the cyclooctene epoxidation are summarized in Table 2. Cyclooctene underwent epoxidation over 3D TS-1; however, the yield of the epoxide as well as the selectivity of the process were low (yield 3.0% after 4 h, selectivity 42% at 20% conversion). Unexpectedly, the performance of the layered TS-1 was comparable to 3D TS-1 in terms of the yield (2.8% after 4 h); however, the layered TS-1 provided a higher selectivity (75% at 20% conversion). We assume that cyclooctene is oxidized due to the flexibility of its molecule, which is able to adopt such a conformation that fits into the TS-1 channels, although our observation is in contradiction with Ryoo’s observations [18]. When using pillared TS-1, the yield
Fig. 6. DR-UV spectra of the titanosilicate samples: 3D TS-1 (a), Ti-MCM-36 (b), layered TS-1a (c), Ti-pillared TS-1a(20) (d); layered TS-1b (e), pillared TS-1b (f), Ti-pillared TS-1b(10) (g), Ti-pillared TS-1b(20) (h), Ti-pillared TS-1b(60) (i).
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model
ARTICLE IN PRESS
CATTOD-9149; No. of Pages 7
J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
6
Table 4 Oxidation of ␣-pinene over titanosilicate catalysts at 60 ◦ C; conversion and yield after 4 h of the reaction. Catalyst
Conversion (%)
Epoxide yield (%)
S10% (%)a
TS-1 Pillared TS-1a Ti-pillared TS-1a(20) Ti-MCM-36
8.8 12.3 9.6
No epoxide 0.6 1.0 1.9
7 8 20
a
Selectivity at 10% conversion defined as [yield of epoxide]/[consumed alkene].
Table 5 Oxidation of verbenol to verbenone over titanosilicate catalysts at 60 ◦ C; conversion and yield after 4 h of the reaction.
Fig. 7. Development of the cyclooctene oxide yield in time over Ti-pillared TS-1b catalysts.
only slightly increased to 3.5% after 4 h, which may be accounted for the dilution of the active component. The performance of Ti-MCM36 was better providing 8% yield after 4 h at the similar selectivity (54% at 20% conversion). Ti-pillared TS-1 (20a) provided yield of 10% at the same time and with similar selectivity as well being the best performing catalyst of the (a) series. Even better results were achieved with catalysts of the (b) series providing yields of cyclooctene oxide of 15.3, 16.9 and 14% using Ti-pillared TS-1b(10), (20) and (60), respectively, after 4 h, with selectivity of 76–80% at 20% conversion. Development of the yield in time for these catalysts is presented in Fig. 7. The difference in the yield obtained using Ti-pillared TS-1a(20) and Ti-pillared TS-1b(20) can be explained as a pure manifestation of the influence of surface accessibility. The catalysts possess the same titanium content (Si/Ti (Ti-pillared TS-1a(20)) = 23, Si/Ti (Ti-pillared TS-1b(20)) = 20); however, they differ in the BET surface area and the external surface area as well. The ratio between the yield of cyclooctene oxide y4 h (Ti-pillared TS-1b(20))/y4 h (Tipillared TS-1a(20)) = 1.69 is the same as their external surface area ratio Sext (Ti-pillared TS-1b(20))/Sext (Ti-pillared TS-1a(20)) = 1.68. Having in mind the dependence of surface area on Ti content in the pillaring mixture, we conclude there are two factors contradicting each other and thus defining the catalyst performance. The higher the content of titanium is, the more active sites are present; however, the lower specific area possesses the material. It appears that the Si/Ti ratio 20 in the pillaring mixture is close to the optimum. The only by-product observed was 2-cycloocten-1-one. The highest yield of norbornene oxide was achieved with Ti-pillared TS-1b(20) (14.8% after 4.5 h; 56% selectivity at 10% conversion), while 3D TS-1 gave the epoxide yield only 1.8% (Table 3). The order of catalyst performance is similar to the case of cyclooctene 3D TS-1 (norbornene oxide yield 1.8% after
Table 3 Epoxidation of norbornene over titanosilicate catalysts at 60 ◦ C; conversion and yield after 4.5 h of the reaction. Catalyst
Conversion (%)
Epoxide yield (%)
S10% (%)a
3D TS-1 Layered TS-1b Pillared TS-1b Ti-pillared TS-1b(20) Ti-MCM-36
2.8 12.0 10.6 23.0 24.1
1.8 6.1 5.5 14.8 8.5
65b 42 52 56 26
a b
8 h.
Selectivity at 10% conversion defined as [yield of epoxide]/[consumed alkene]. Selectivity provided at 3% conversion; 10% conversion was not achieved until
Catalyst
Conversion (%)
Verbenone yield (%)
S10% (%)a
3D-TS-1 Layered TS-1a Ti-pillared TS-1a(20) Ti-MCM-36
30.6 46.2 10.4 30.2
24.7 44.8 10.2 14.2
71 97 98 32
a Selectivity at 10% conversion defined as [yield of verbenone]/[consumed verbenol].
4.5 h) < layered TS-1 (yield 6.1% after 4.5 h), pillared TS-1 (yield 5.5% after 4.5 h) < Ti-MCM-36 (yield 8.5% after 4.5 h) < Ti-pillared TS-1 (yield 14.8% after 4.5 h); however, the difference between 3D TS-1 and layered structures is more evident due to rigidity of the norbornene molecule. The performance of Ti-MCM-36 and Ti-pillared TS-1 exceeds the performance of Ti-CFI catalyst reported by our group previously (yield 5.8% after 4 h) [16]. Main oxidation byproduct observed is 3-hydroxy-norbornane-2-one. ␣-Pinene was oxidized using only layered catalysts because it is a bulky molecule that cannot access the narrow pores of the 3DTS-1 (Table 4). The best yield was obtained using the Ti-MCM-36 catalyst (1.9% after 4 h); however, the selectivity of the oxidation ranges from 7 to 20% at 10% conversion and many other oxidation products are observed including mainly hydroxyl and carbonyl compounds. This is in accordance with the findings of Eimer et al. using mesoporous titanosilicate prepared from TS-1 precursors [15]. Last but not least, verbenol was used as a substrate as it being one of the products of ␣-pinene oxidation. The results are summarized in Table 5. Verbenol was rapidly oxidized (conversion up to 46% after 4 h) with a high selectivity to the main product ranging 71–98% for TS-1 catalysts. However, the main product is not epoxide but verbenone. No verbenol epoxide was observed in any of the reactions. The 3D TS-1 catalyst was also active, providing higher conversion than Ti-pillared TS-1 (30.6% vs. 10.4% after 4 h). Tatsumi et al. showed that reaching the right position to the active site may be difficult for the substrates with trisubstituted double bond (e.g. 3-methyl-2-buten-1-ol) and this steric hindrance leads to the preferential oxidation of hydroxyl group if present and better accessible. Epoxide and aldehyde or ketone are formed concurrently in such cases [30]. Verbenol represents an extreme case when epoxidation reaction is suppressed and only hydroxyl group is oxidized. 4. Conclusions Pillared TS-1 catalysts were prepared from TS-1 nanosheets, using TEOS as a pillaring medium. The pillaring treatment helps preserve the mesoporosity of the material and keeps the active centres accessible even for bulky molecules. The disadvantage in the active phase dilution has been overcome by the addition of TBOTi into the pillaring mixture and thus boosting the Ti content in the material. On the other hand, addition of TBOTi causes a decrease in the surface area in comparison with TS-1 pillared with pure TEOS. Based on the catalytic testing of materials with different Ti content,
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002
G Model CATTOD-9149; No. of Pages 7
ARTICLE IN PRESS J. Pˇrech et al. / Catalysis Today xxx (2014) xxx–xxx
we conclude that the optimum composition of the pillaring mixture is close to TEOS/TBOTi ratio 20. The pillared TS-1 and Ti-pillared TS-1 were active in cyclooctene, norbonene and ␣-pinene epoxidation with hydrogen peroxide as the oxidant under mild reaction conditions. The performance of Ti-pillared TS-1 catalysts exceeds 3D TS-1 as well as layered TS-1. Furthermore, a selective oxidation of verbenol to verbenone was observed under the reaction conditions used. In contrast to epoxidation reactions, layered TS-1 was the best performing catalyst for the verbenol oxidation. Acknowledgements The authors thank Dr. L. Brabec for SEM images, Dr. G. Sádovská for UV/Vis measurements and Dr. Martin Lamaˇc for NMR analyses. The authors acknowledge the Czech Science Foundation (P106/12/G015) and the NSTIP strategic technologies programs (12-(PET2614)02) in the Kingdom of Saudi Arabia for the support of this research. References ˇ [1] J. Cejka, A. Corma, S.I. Zones (Eds.), Zeolites and Catalysis: Synthesis, Reactions and Applications, Wiley-VCH, Weinheim, 2010. ˇ [2] J. Cejka, G. Centi, J. Perez-Pariente, W.J. Roth, Catal. Today 179 (2012) 2–15. [3] T. Tatsumi, Metal-substituted zeolites as heterogenous oxidation catalysts, in: N. Mizuno (Ed.), Modern Heterogenous Oxidation Catalysis, Wiley-WCH, Weinheim, 2009, pp. 125–153. [4] A. Corma, P. Esteve, A. Martínez, S. Valencia, J. Catal. 152 (1995) 18–24. [5] T. Tatsumi, M. Nakamura, S. Negishi, H. Tominaga, J. Chem. Soc. Chem. Commun. 476 (1990).
7
[6] G. Peregot, G. Bellussi, C. Corno, M. Taramasso, F. Buonomot, A. Esposito, In: Y. Murakami, A., Iijima, J.W. Ward (Eds.), Stud. Surf. Sci. Catal. 28 (1986) 129–136. [7] P. Birke, P. Kraak, R. Schödel, F. Vogt, in: T. Hattori, T. Yashima (Eds.) Stud. Surf. Sci. Catal. 83 (1994) 425–432. [8] J.S. Reddy, P.A. Jacobs, Catal. Lett. 37 (1996) 213–216. [9] V. Hulea, F. Fajula, J. Bousquet, J. Catal. 198 (2001) 179. [10] R.A. Sheldon, J. Dakka, Catal. Today 19 (1994) 215–246. [11] C. Perego, A. Carati, P. Ingallina, M.A. Mantegazza, G. Bellussi, Appl. Catal. A 221 (2001) 63–72. [12] M. Tamarasso, G. Perego, B. Notari, US Patent 4410501, 1983. [13] http://www.eni.com/en IT/attachments/azienda/attivita-strategie/ petrolchimica/licensing/TS1-flyer-lug09.pdf (22 May 2014). [14] J.C. van der Waal, M.S. Rigutto, H. van Bekkum, Appl. Catal. A 167 (1998) 331–342. [15] G.A. Eimer, I. Diaz, E. Sastre, S.G. Casuscelli, M.E. Crivello, E.R. Herreroa, J. PerezPariente, Appl. Catal. A 343 (2008) 77–86. ˇ ˚ J. Cejka, [16] J. Pˇrech, M. Kubu, Catal. Today 227 (2014) 80–86. ˇ [17] W.J. Roth, P. Nachtigall, R.E. Morris, J. Cejka, Chem. Rev. 114 (2014) 4807–4837. [18] K. Na, Ch. Jo, J. Kim, W. Ahn, R. Ryoo, ACS Catal. 1 (2011) 901–907. [19] M. Taramasso, G. Perego, B. Notari, in: H. Robson (Ed.), Verified Syntheses of Zeolitic Materials, Elsevier, Amsterdam, 2001, p. 207. [20] P. Wu, T. Tatsumi, T. Komatsu, T. Yashima, J. Phys. Chem. B 105 (2001) 2897–2905. [21] W. Fan, P. Wu, S. Namba, T. Tatsumi, Angew. Chem. Int. Ed. 116 (2004) 238–242. [22] S. Kim, H. Ban, W. Ahn, Catal. Lett. 113 (2007) 160–164. [23] L. Wang, Y. Wang, Y. Liu, H. Wu, X. Li, M. He, P. Wu, J. Mater. Chem. 19 (2009) 8594–8602. [24] B.C. Lippens, J.H. de Boer, J. Catal. 4 (1965) 319. [25] K. Na, M. Choi, W. Park, Y. Sakamoto, O. Terasaki, R. Ryoo, J. Am. Chem. Soc. 132 (2010) 4169–4177. [26] A. Thangaraj, S. Sivasanker, J. Chem. Soc. Chem. Commun. 123 (1992). [27] B. Notari, Adv. Catal. 41 (1996) 253. [28] F. Bonito, A. Damin, G. Ricchiardi, M. Ricci, G. Spano, R. D’Aloisio, A. Zecchina, C. Lamberti, C. Prestipino, S. Bordiga, J. Phys. Chem. B 108 (2004) 3573–3583. [29] A. Zecchina, G. Spoto, S. Bordiga, A. Ferrero, G. Petrini, G. Leofanti, M. Padovan, in: P.A. Jacobs, N.I. Jaeger, L. Kubelková, B. Wichterlová (Eds.), Stud. Surf. Sci. Catal. 69 (1991) 251–258. [30] T. Tatsumi, M. Yako, M. Nakamura, Y. Yuhara, H. Tominaga, J. Mol. Catal. 78 (1993) L41–L45.
Please cite this article in press as: J. Pˇrech, et al., Epoxidation of bulky organic molecules over pillared titanosilicates, Catal. Today (2014), http://dx.doi.org/10.1016/j.cattod.2014.07.002