Первый свет / First Light (2018) WEBRip 720p | L2 | Bobs Burgers S07E03 HDTV x264-FLEET [eztv] | Station 19

SBA-15 catalysts

SBA-15 catalysts

G Model ARTICLE IN PRESS CATTOD-10504; No. of Pages 8 Catalysis Today xxx (2016) xxx–xxx Contents lists available at ScienceDirect Catalysis Toda...

929KB Sizes 1 Downloads 87 Views

G Model

ARTICLE IN PRESS

CATTOD-10504; No. of Pages 8

Catalysis Today xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Catalysis Today journal homepage: www.elsevier.com/locate/cattod

Propane oxidative dehydrogenation over VOx /SBA-15 catalysts Gheorghita Mitran a , Rawaz Ahmed b,1 , Emmanuel Iro b , Saeed Hajimirzaee b , Simon Hodgson b , Adriana Urd˘a a,∗ , Maria Olea b , Ioan-Cezar Marcu a,∗ a Laboratory of Chemical Technology and Catalysis, Department of Organic Chemistry, Biochemistry and Catalysis, Faculty of Chemistry, University of Bucharest, 4-12, Blv. Regina Elisabeta, 030018 Bucharest, Romania b School of Science and Engineering, Teesside University, Middlesbrough, United Kingdom

a r t i c l e

i n f o

Article history: Received 15 September 2016 Received in revised form 2 December 2016 Accepted 7 December 2016 Available online xxx Keywords: Oxidative dehydrogenation Propane Supported vanadia catalysts VOx /SBA-15

a b s t r a c t VOx /SBA-15 catalysts with five different vanadium loadings were prepared by a modified wet impregnation method, characterized using N2 adsorption, XRD, EDX, SEM, Raman and UV–vis spectroscopies and H2 -TPR techniques, and tested in the oxidative dehydrogenation of propane in the temperature range 450–600 ◦ C. For all the catalysts the propane conversion increases with both reaction temperature and vanadium loading, while the selectivity for propene decreases mainly to the benefit of carbon oxides. Several types of VOx species coexist on the catalyst surface, with monomeric and low-polymerized ones leading mainly to propene, while V2 O5 crystallites at high vanadium loadings producing more carbon oxides. Propene was determined to be the only primary product irrespective of the vanadium content. © 2016 Elsevier B.V. All rights reserved.

1. Introduction The increasing demand in unsaturated hydrocarbons for the manufacturing of polymers, synthetic fibers, detergents, elastomers and synthetic fuels, has triggered the development of inexpensive and environmentally friendly ways of their production. The lower price of alkanes compared to the corresponding alkenes makes their use as feedstock in the chemical and petrochemical industry very attractive [1–3]. However, the current alkanes conversion technology, based on the dehydrogenation process, has several disadvantages, such as high endothermicity, unfavorable equilibrium shift and difficult control of cracking side reactions at high temperatures, which contribute to the rapid coking of the catalyst [1,3,4]. Therefore, the alternative conversion route consisting in the oxidative dehydrogenation (ODH) process becomes promising from engineering and economic points of view by overcoming all these drawbacks [1,5–7]. The main problem remains avoiding side reactions leading to oxygen-containing organic molecules and carbon oxides, which severely limit the selectivity to dehydrogenation products [1,8,9]. Early attempts to use other oxidants (i.e. iodine) were abandoned due to corrosion and environmental problems

∗ Corresponding authors. E-mail addresses: [email protected] (A. Urd˘a), [email protected] (I.-C. Marcu). 1 Present address: Kurdistan Institution for Strategic Studies and Scientific Research (KISSR), Qirga, Slemani, As Sulaymaniyah, Iraq.

posed by the oxidant [1]. Recently, N2 O was also used as oxidation agent in ODH reactions with interesting results [10,11]. Many catalysts have been tested for the ODH reaction, but supported vanadia and molybdena proved to be the most active and selective ones [12]. In both cases, the most selective catalysts contain well-dispersed vanadium or molybdenum species onto the support surface, at near monolayer coverage, which is about 7.5 V nm−2 for polyvanadate [13] and about 4.8 Mo nm−2 for polymolybdate; however, vanadia has higher ODH activity than molybdena at equivalent dispersion [6]. Notably, much lower values corresponding to the monolayer coverage were more recently proposed for catalysts supported on silica (0.7 V nm−2 ) and explained by the lower density and reactivity of silica surface hydroxyl groups [8]. Vanadium oxide catalysts supported on various oxides, such as SiO2 , Al2 O3 , TiO2 , Nb2 O5 , MgO, La2 O3 , Sm2 O3 and Bi2 O3 , to mention some of them, have been extensively studied [1,7,14,15]. Their catalytic performance in the ODH of alkanes depends on the nature of the support (which, in turn, is partly responsible for the vanadium reducibility and dispersion), the vanadium loading and the preparation conditions, which influence the nature and structure of the VOx surface species [5,15]. Moreover, the acid-base character of the support affects the catalytic performance [15]. It has been found that in the ODH of propane the most selective catalysts were obtained with vanadium supported on basic, rather than on acidic metal oxides. It has also been found that VOx species with V in octahedral and tetrahedral positions are the active ones, the tetrahedral being less active but more selec-

http://dx.doi.org/10.1016/j.cattod.2016.12.014 0920-5861/© 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: G. Mitran, et al., Propane oxidative dehydrogenation over VOx /SBA-15 catalysts, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.014

G Model CATTOD-10504; No. of Pages 8

ARTICLE IN PRESS G. Mitran et al. / Catalysis Today xxx (2016) xxx–xxx

2

tive. The higher the concentration of VO2+ and V2 O5 species the lower is the selectivity to propene [16]. The presence of polymerized vanadium species and bulk V2 O5 is detrimental to the activity of the catalyst as well, as less vanadium atoms are accessible to the propane molecules. In order to minimize the formation of V2 O5 crystallites responsible for the formation of COx , oxide supports with high surface area and lower reducibility than vanadia are used [6,7,17]. The support plays an important role in the dispersion of the active phase, on the accessibility, reactivity and acidity of the active sites, and also on the mass/heat transfer phenomena [18]. Combining the textural and acid-base properties of the support with the redox properties of the vanadia species makes it easier to activate alkanes at lower temperatures [19]. Silica-supported vanadia catalysts were reported to be active in propane ODH [5] and, in order to increase the propene yields, mesoporous siliceous materials such as MCM-41, MCM-48, mesocellular foams (MCF) and SBA-15 were used as supports [5,20,21]. SBA-15 is a mesoporous silica, characterized by large surface area and formed by a hexagonal array of uniform tubular channels, with tunable pore diameters in the range of 5–30 nm and very narrow pore-size distribution [22,23]. It has been used as support for catalysts in the ODH of butanes [24,25], ethane [26], ethylbenzene [11], as well as for methanol oxidation [23,27]. Propane ODH on vanadia supported on mesoporous SBA-15 has been studied by several groups [17,22]. However, since both silica and vanadia possess acidic character, it is very challenging to prepare catalysts with high dispersion at high loadings. Therefore, the preparation method and conditions should be carefully chosen. Some of the preparation methods used so far are: alcoholic impregnation [20,22], wet impregnation [17,24,28], mechanical mixing [22], direct synthesis [18,29], adsorption [22] and grafting-ion exchange [27,30,31]. Although superior performances of the mesoporous VOx /SBA-15 catalysts have already been reported [22], a fundamental understanding of their behavior towards the propane ODH remains to be accomplished. The aim of this paper is to investigate further the catalytic activity of the SBA-15-supported vanadia catalysts and to gain a deeper insight into their structure-activity relationship. In order to accomplish these goals, a series of VOx /SBA-15 catalysts, with vanadium loadings between 6.5 and 11 wt%, were prepared using an improved impregnation method and their detailed characterization by N2 -adsorption, SEM/EDX, XRD, H2 -TPR, Raman and UV–vis spectroscopies as well as their catalytic performance in the oxidative dehydrogenation of propane have been performed.

2. Experimental 2.1. Support preparation The general experimental procedure followed in the preparation of SBA-15 mesoporous silica was described in detail by Zhao et al. [32]. However, this procedure was modified and the preparation conditions, such as the amount of P123, type of acid and aging time, were optimized through experiments. The optimized procedure is described below. 4 g of Pluronic 123 triblock copolymer (EO20-PO70-EO20, P123, Sigma-Aldrich) was dissolved into 120 g of deionised (DI) water at 40 ◦ C, under vigorous stirring for 6 h. Then 20 ml of 2 M HCl (Fisher Scientific) was added to the clear solution and after 10 min, 9.25 g of TEOS (Tetraethylorthosilicate, Sigma-Aldrich) was introduced dropwise, the solution being kept under vigorous stirring at 40 ◦ C for 20 h. A white precipitate was formed which was kept in an oil bath at 95 ◦ C for 24 h, without stirring, then filtered, washed with 600 ml of deionised water and dried at 60 ◦ C for 24 h. The white powder was then calcined at 550 ◦ C, under steady-state conditions,

for 6 h, at a heating rate of 10 ◦ C min−1 . After calcination, the sample was cooled down to 60 ◦ C at a cooling rate of 10 ◦ C min−1 and placed in a round bottom flask with a short condenser inside to be re-hydrated by using 50 g of DI water. The suspension was then placed in an oil bath at 105 ◦ C and left boiling under a gentle stirring. After 2 h, the suspension was filtered and 2 g of SBA-15 sample were collected by filtration and dried a 120 ◦ C for 6 h. 2.2. Catalyst preparation Five SBA-15-supported vanadia samples were prepared by wet impregnation. A solution of ammonium vanadate, NH4 VO3 (SigmaAldrich), was contacted with the SBA-15 support, under stirring, at 55–60 ◦ C for 2 h, to achieve a final V content of 6.5–11.0 wt%. The solid product was recovered by filtering and then dried in vacuum, at 100 ◦ C, for 1 h, followed by calcination in air at 600 ◦ C, for 5 h, at a heating rate of 5 ◦ C min−1 . The impregnated samples were labelled as xV, where x denotes the wt% of V. 2.3. Catalysts characterization The textural characterization was performed using the nitrogen adsorption/desorption method, with a Micromeritics ASAP 2010 automatic equipment. The surface areas were calculated using the BET method in the relative pressure, P/P0 , region 0.065–0.2, while the pore sizes were determined by the BJH method from the nitrogen desorption branch. Prior to nitrogen adsorption, the samples were degassed at 150 ◦ C for 10 h. Scanning Electron Microscopy (SEM) accompanied by X-ray energy dispersion analysis (EDX) was used to monitor the morphology and surface composition of the catalysts samples. SEM/EDX examination was performed using a Hitachi S-4500 field emission microscope with a Quartz PCI XOne SSD X-ray analyser. To obtain information on both, the mesoporous structure and the crystalline structure of the SBA-15-supported vanadia catalysts, small- and wide-angle X-ray diffraction (XRD) patterns were recorded on a Bruker D8 (25 kV, 20 mA) powder X-ray diffractometer, using Cu K␣ radiation (␭ = 0.15406 nm), a tube voltage of 40 kV, and a current of 20 mA. The data were collected from 0.5 to 4◦ (2␪) with a resolution step size of 0.01◦ (small-angle XRD patterns) and from 10 to 70◦ (2␪) with a resolution step size of 0.02◦ (conventional wide-angle XRD patterns). The type of V species formed on the surface of the SBA-15 was studied with Raman spectroscopy. The Raman spectra were recorded with a Raman Microscope Spectrometer Horiba Jobin Yvon – Labram HR UV–vis-NIR, between 100 and 1200 cm−1 , using a 633 nm wavelength for excitation. Before recording the spectra, the spectrometer was calibrated using the Raman band at 520 cm−1 of a standard SiO2 sample. The UV–VIS spectra were recorded using a UV3600 UV–vis spectrophotometer with Shimadzu ISR-3100 integrating sphere attachment having an angle of incident light of 0–8◦ , two light sources: D2 (deuterium) lamp for the ultraviolet range and WI (halogen) lamp for the visible range and a PMT (photomultiplier tube) detector. The spectra were recorded in the range of 190–800 nm (the switching wavelength of the lamps is between 282 nm and 393 nm) with a wavelength step of 2 nm, having the slit width of 8 nm. The spectra were measured using samples diluted with extra pure barium sulfate (Nacalai Tesque). The reducibility of the catalysts was studied by temperature programmed reduction under hydrogen (H2 -TPR). Experiments were performed using a CATLAB microreactor – MS system (Hiden Analytical, UK) under a flow of 5% H2 /Ar mixture through the CATLAB’s packed micro-reactor containing about 35 mg of sample, which was heated at a constant rate of 20 ◦ C min−1 up to 850 ◦ C. The system

Please cite this article in press as: G. Mitran, et al., Propane oxidative dehydrogenation over VOx /SBA-15 catalysts, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.014

G Model CATTOD-10504; No. of Pages 8

ARTICLE IN PRESS G. Mitran et al. / Catalysis Today xxx (2016) xxx–xxx

600

2.4. Catalytic tests

500

3. Results and discussion

400

3

-1

SBA-15 6.5V 7.5V 9V 10V 11V

300

200

100

0 0.2

0

0.4

0.6

0.8

Fig. 1. Nitrogen adsorption-desorption isotherms of bare SBA-15 and VOx /SBA-15 samples.

(110) (200)

(100)

11V 10V 9V 7.5V 6.5V

3.1. Catalysts characterization All the samples showed adsorption-desorption isotherms (Fig. 1) of type IV, typical for mesoporous materials, with H1 hysteresis loops characteristic to well-defined, relatively uniform cylindrical pores, with a narrow size distribution. The amount of adsorbed nitrogen was significantly higher for the SBA-15 support than for the V-containing samples. Regarding the latter samples, it was similar for the lower three V loadings, then decreased for the 10V and 11V samples. The SBA-15 sample has narrow and bimodal pores with maxima at about 5.9 and 7.1 nm, while for the V-containing samples the pore distributions become unimodal, with an average pore diameter of 6.3–6.4 nm (Fig. S1). This suggests that V deposition blocked the low-size pores of the SBA-15 support, in line with the significantly lower surface area of the V-containing samples (Table 1). On the other hand, there were no significant changes of the average pore diameter among the V-containing samples (Table 1) suggesting a well and uniform deposition of vanadium species onto the SBA-15 walls. The SEM images (Fig. S2) show the surface morphology of the catalyst samples. They all have the same fiber-like morphology of the SBA-15 support, with primary particles coupled along their length and having a width of ca. 5 ␮m, proving that no morphological changes were produced by V deposition. The chemical composition, estimated by EDX, and textural properties of the prepared catalysts are presented in Table 1. Both the surface area (793 m2 g−1 ) and the pore volume (0.85 cm3 g−1 ) of the SBA-15 support were higher than those of the V-containing sys-

1

Relative pressure (P/P0)

Intensity (a.u.)

The catalytic ODH of propane was carried out in a fixed bed quartz tube down-flow reactor with an internal diameter of 15 mm, as previously described [33]. The catalyst bed (0.1 g) was supported by quartz wool, while the dead volumes before and after the catalyst bed were filled with quartz chips in order to minimize potential homogeneous reactions at higher temperatures [1]. The axial temperature profile was measured using an electronic thermometer placed in a thermowell centered in the catalyst bed. The propane – air mixture (with propane-to-oxygen molar ratios between 1 and 4) was fed into the reactor with a space velocity between 1000 and 2000 ml (min gcat )−1 . Catalytic reactions were performed at atmospheric pressure in the temperature range from 450 to 600 ◦ C. In a typical reaction run, the reactor was heated to the desired temperature in a reactants’ flow, and then allowed to stabilize for 1 h before the first analysis was made. Each run was carried out over a period of 2–3 h, until two consecutive measurements were identical. The reaction products were analyzed with a ThermoQuest Trace Gas Chromatograph equipped with a thermal conductivity detector (TCD) using a CTR I column, and a flame ionization detector (FID) using an alumina column. The presence of oxygenates was checked on a ThermoFinnigan Gas Chromatograph equipped with a flame ionization detector (FID) using a DB-5 column. The major products observed under our reaction conditions were propene, CO, CO2 and cracking products (methane and ethylene), with trace amounts of oxygenates (acrolein, acetaldehyde). The latter were not considered for calculating the catalytic performances as the carbon balance based on the major products was in all runs higher than 97%. Conversion of propane and product selectivities were expressed as mol% on a carbon atom basis.

Adsorbed volume (cm g )

was maintained for 1 h at 850 ◦ C under H2 /Ar flow to complete the reduction.

3

SBA-15 0

1

2

3

4

5

6

2-theta (°) Fig. 2. Small-angle diffractograms of the SBA-15 support and VOx /SBA-15 catalyst samples.

tems. A decrease in the BET surface areas (from 502 to 419 m2 g−1 ) and pore volumes (from 0.82 to 0.69 cm3 g−1 ) is observed as the V content increases. Changes in textural parameters upon vanadium impregnation are often reported in the literature, attributed to VOx species entering the support pores or to a partial destruction of the framework [9,17]. These changes could also be due, at least partly, to the treatment in water of SBA-15 material during the impregnation process, as previously reported by Galarneau et al. [34]. For our samples, it can be concluded that well-organized mesoporous Vcontaining SBA-15 catalysts with a high surface area were obtained for all V loadings. The VOx surface density, defined as the number of vanadium atoms per square nanometer of the catalyst, was calculated, the results obtained being tabulated in Table 1. It provides a convenient parameter for comparing catalysts with different surface areas. A more detailed discussion of the results will follow. Fig. 2 shows the small-angle XRD patterns of the calcined samples, with three resolved peaks at 0.88, 1.54 and 1.75◦ 2␪, attributed

Please cite this article in press as: G. Mitran, et al., Propane oxidative dehydrogenation over VOx /SBA-15 catalysts, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.014

G Model

ARTICLE IN PRESS

CATTOD-10504; No. of Pages 8

G. Mitran et al. / Catalysis Today xxx (2016) xxx–xxx

4

Table 1 The chemical composition and textural properties of the SBA-15 support and VOx /SBA-15 catalysts. Sample

V loading (wt%)

BET surface area (m2 g−1 )

Average pore diameter (nm)

Total pore volume (cm3 g−1 )

V density (nm−2 )

SBA-15 6.5V 7.5V 9V 10V 11V

– 6.6 7.4 8.9 10.2 10.9

793 502 487 480 468 419

5.9 and 7.1 6.4 6.4 6.3 6.4 6.3

0.85 0.82 0.81 0.81 0.76 0.69

– 1.6 1.8 2.2 2.6 3.1

146

195

(110) (301) (400)

285

407

525

698

996 1036

11V

Intensity (a.u.)

(321) (710)

(021)

(411) (600) (012) (020)

Intensity (a. u.)

(002) (102)

271

(011) (310)

(001) (101)

157

(200)

11V 10V

10V

9V 7.5V 6.5V SBA-15

9V 100

7.5V

200

300

400

500

600

700

800

900

1000

1100 1200

-1

Raman shift (cm )

Fig. 4. Raman spectra of bare SBA-15 and VOx /SBA-15 samples.

6.5V 10

20

30

40

50

60

70

2-theta (°) Fig. 3. Wide-angle diffractograms of the calcined VOx /SBA-15 catalyst samples (V2 O5 phase indexed).

to the (100), (110) and (200) reflections, respectively, features of SBA-15 with hexagonal symmetry [24,25]. As the vanadium loading increases, the intensity of the (100) peak weakens. This can be due to the partial blocking of the SBA-15 mesopores and decline of long-range order of hexagonally arranged porosity [24] or, to the deposition of vanadia onto the pore walls, as vanadium has a higher absorption factor for X-rays than silicon [35]. However, all three peaks are visible at all V loadings, showing that the SBA-15 structure is retained upon vanadium impregnation. Fig. 3 shows the wide-angle diffractograms. It can be observed that all samples have a broad peak centered at ca. 13.5◦ 2␪, attributed to the amorphous SBA-15 support. The presence of diffraction peaks at higher 2␪ angles indicates the presence of V2 O5 crystallites (PDF 03-65-0131; characteristic peaks at ca. 15.5, 20, 22, 26 and 31◦ 2␪) [36,37]. The most intense of them (2␪ = 20 and 26◦ ) can barely be observed in the 6.5V sample, accounting for the high dispersion of the vanadia species, likely as mono- and polyvanadate species. As the vanadium loading increases, the V2 O5 peaks become more intense. From the practical point of view, a monolayer surface coverage for the 6.5V sample can be considered, as it is close to the point when vanadia species start to condense to nanocrystalline oxide-like species as shown by XRD results for the samples with higher V content. From the theoretical point of view, the monolayer (a bidimensional surface vanadium oxide overlayer on the support [14]) was defined as the need for 4.98 × 1014 V2 O5 species nm−2 to completely cover the surface of the support [38], or as 0.10 wt% V2 O5 per m2 of surface [39]. Monolayer surface cov-

erage of the surface vanadia on oxide supports has been estimated from structural and experimental determinations. From the V O bond lengths of crystalline V2 O5 , monolayer coverage is estimated at 10 VOx nm−2 for a bidimensional polyvanadate layer and at 2.5 VOx nm−2 for isolated monomeric vanadia units. Monolayer surface coverage determined from Raman spectroscopy was found to be ca. 7–8 VOx nm−2 for different oxide supports (Al2 O3 , TiO2 , ZrO2 , Nb2 O5 and CeO2 ) with the exception of silica-supported vanadia, which exhibited a maximum surface coverage of only 0.7 VOx nm−2 [40 and References therein]. Other studies found 1.1 VOx nm−2 for monolayer coverage on silica [5,9], attributed to the weak interaction between the vanadium oxide and the support, leading to a poor dispersion of the precursor salt [5,14]. This is an expected result, as both V2 O5 and SBA-15 have an acidic character. It has been reported that the dispersion of the vanadium oxide as well as its structure can be understood on the basis of the different acid-base character of the different supports. The acidity of the metal oxide is related to the pH at which the surface possesses zero surface charge (point of zero charge, PZC, or isoelectric point, IE). IE of V2 O5 was measured as 1.4 [38], while for SBA-15 it was 4.2 [41]. A weaker interaction of the vanadium oxide with another acidic oxide, SBA-15 here, will favour the agglomeration of the VOx units to form V2 O5 . In this study a monolayer coverage on SBA-15 can be approximated to be 1.6 VOx nm−2 . To the best of our knowledge, this is the highest monolayer coverage of vanadia on SBA-15 reported so far. The reason for this higher coverage can be explained by the fact that the support was re-hydrated before impregnation, which led to a better dispersion of the surface species. Raman spectroscopy has a higher sensitivity to the presence of oxide-like species than XRD [9]. The Raman spectrum of the SBA-15, presented in Fig. 4, exhibits no characteristic bands, while those of the V-containing samples show bands corresponding to different vanadium species present on the surface. Thus, the 6.5V catalyst shows bands at ca. 1036 cm−1 attributed to the symmetric stretch-

Please cite this article in press as: G. Mitran, et al., Propane oxidative dehydrogenation over VOx /SBA-15 catalysts, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.014

G Model

ARTICLE IN PRESS

CATTOD-10504; No. of Pages 8

G. Mitran et al. / Catalysis Today xxx (2016) xxx–xxx

5

255

608 707

290 Hydrogen consumption (a.u.)

11V

Kubelka-Munk function (a.u.)

11V

385

10V

591 698

10V 578

560

668

7.5V 555

665

6.5V

9V

470 0

100

200

300

400

500

600

700

800

900

Temperature (°C)

7.5V 6.5V 200

683

9V

Fig. 6. H2 -TPR profiles for the VOx /SBA-15 samples.

300

400

500

600

Wavelength (nm) Fig. 5. UV–vis DRS spectra of the VOx /SBA-15 samples.

VO3− 4

ing mode of the vanadyl bond (V O) in isolated monomeric species [14,20,27], and at 157 and 271 cm−1 , assigned to hydrated amorphous polymeric species on the surface [27]. As the vanadium loading increases, the intensity of the 1036 cm−1 band diminishes, while those belonging to polymeric species intensify. New bands at 996, 698, 525, 407, 285, 195, and 146 cm−1 , assigned to V2 O5 crystallites on the support [6,27,42], appear and increase in intensity, in line with the XRD results. Since the ratio of scattering cross section of crystalline V2 O5 relative to isolated monovanadate and polyvanadate species was estimated to be very large (approximately 10 [6,17,20]), V2 O5 nanoparticles may exist in low amounts in the 7.5V and 9V samples. UV–vis DRS spectra give additional structural information on the surface VOx species, the energy of the oxygen-vanadium charge transfer bands indicating the coordination of the V5+ center [20]. The SBA-15 support exhibited only a very low intensity spectrum (not shown). For the VOx /SBA-15 catalyst samples (Fig. 5), the spectra show an absorption band at ca. 255 nm with a shoulder at ca. 290 nm and another one at ca. 385 nm, with intensities increasing with vanadium loading. The first band can be attributed to low-energy charge transfer transitions between O2− ligands and V5+ species in isolated tetrahedral VO3− structures on the 4 SBA-15 surface [11,18,20,43], while the shoulder arises from the charge transfer transitions in oligomeric (VO3− 4 )n structures [43]. The second band can be assigned to 2D polymeric structures of pentacoordinated vanadium ions, i.e. polyvanadates [43]. Only in the 10V and 11V samples a shoulder was observed at around 470 nm, indicating the presence of “bulk-like” vanadia crystallites [20,43]. These findings are in line with the XRD and Raman results and confirm the presence of mono- and oligomeric (VO3− 4 )n species and of polyvanadate structures on the surface of all samples, and of bulk-like crystallites at high vanadium loading. The selectivity for oxidative dehydrogenation products depends on the reactivity and availability of the surface oxygen species: when surface intermediates have a high probability to react with lattice oxygen species, a lower selectivity to propene is obtained [1]. Therefore, hydrogen temperature-programmed reduction (TPR) profiles were recorded to obtain information on the reducibility of the catalyst samples. For all the catalysts, reduction occurred

in the range 500–700 ◦ C (Fig. 6), the obtained TPR profiles being deconvoluted into two main features. The 6.5V sample shows two broad equal area peaks centered at 555 and 665 ◦ C, respectively. The first peak can be attributed to the reduction of monomeric and oligomeric (VO3− 4 )n species in tetrahedral coordination geometry [43], while the second one can be assigned to the reduction of V2 O5 clusters [43]. The presence of V2 O5 clusters at low V loadings is not unexpected as it has already been evidenced for silica-supported V2 O5 catalysts [43 and References therein]. As the V loading increases both reduction maxima shift to higher temperatures, while the peaks remain broad. Thus, for the first reduction peak the temperature of the maximum H2 consumption increases continuously from 555 ◦ C for 6.5V sample to 608 ◦ C for 11V sample, while for the second reduction peak it increases from 665 to 707 ◦ C. On the other hand, the intensity of the first reduction peak increases while that of the second one continuously decreases. A perusal of these data by analogy with previous results on the reducibility of silica-supported V2 O5 catalysts [20,43 and References therein] clearly suggests that besides the tetrahedral mono- and oligomeric (VO3− 4 )n species mainly identified in the 6.5V sample, less reducible square-pyramidal polymeric structures, i.e. polyvanadates, become predominant with increasing the V loading. At the same time, the V2 O5 clusters responsible for the second reduction peak of the 6.5V sample evolve toward less reducible V2 O5 crystallites at high V loadings. This interpretation is in line with the XRD, Raman and UV–vis data of the samples. It is noteworthy that, except for H2 -TPR, all the characterization results discussed above were obtained at room temperature. However, it has been shown that for silica supported catalysts in reaction conditions and in the presence of water (reaction product), the observed species are not the same as at room temperature [44]. Thus, Launay et al. showed that the molecular structure of VOx species supported on mesoporous silica strongly depends on the degree of hydration. Dehydration by heating at 400 ◦ C caused depolymerization of the vanadium species leading to increased dispersion and anchoring on the silica support. This dynamic change was completely reversible at rehydration. Therefore, it is highly possible that in the reaction conditions the ratio between monomeric and polymeric species changes to favor the former, the selective species in the ODH process.

Please cite this article in press as: G. Mitran, et al., Propane oxidative dehydrogenation over VOx /SBA-15 catalysts, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.014

G Model CATTOD-10504; No. of Pages 8

ARTICLE IN PRESS G. Mitran et al. / Catalysis Today xxx (2016) xxx–xxx

6

Table 2 Propane conversion and product selectivity values in the ODH of propane. Propane-to-oxygen mol ratio = 2:1, space velocity = 1500 ml (gcat min)−1 . Catalyst sample

Temperature (◦ C)

Propane conversion (%)

6.5V

450 500 550 600 450 500 550 600 450 500 550 600 450 500 550 600 450 500 550 600

0.5 2.8 12.1 16.3 0.6 3.9 12.9 17.5 0.8 5.1 13.7 18.5 1.1 6.0 14.9 19.6 1.5 6.8 16.3 21.6

7.5V

9V

10V

11V

a

Eact a (kJ mol−1 )

Selectivities (%) Propene

Cracking

COx

87 80 68 66 84 74 64 62 83 72 59 57 73 65 44 41 71 58 39 38

1 2 3 5 1 1 2 3 1 2 5 7 1 1 2 5 1 1 2 4

12 18 29 29 15 25 34 35 16 26 36 36 26 34 54 54 28 41 59 58

158

152

141

130

119

Apparent activation energy.

3.2. Catalytic study No significant reaction took place in the absence of catalysts [45]. For all the VOx /SBA-15 samples, the main products were propene and carbon oxides, with minor amounts of cracking products (methane and ethylene) and trace amounts of oxygenates. Oxygen was never completely consumed, therefore the catalytic activity and selectivity to products were not influenced by a lack of reactant. The catalytic activities strongly depend on the temperature and vanadium content. As it can be seen in Table 2, conversion values increase with both temperature and V loading, reaching ca. 22% for the 11V sample at 600 ◦ C. At the same time propene selectivity decreased mainly for the benefit of carbon oxides. The highest propene selectivities on the entire temperature range were obtained for the 6.5V catalyst, with values decreasing for higher V content. Correlating these findings with the characterization data, we conclude that the active species for ODH reaction are the isolated vanadium species on the surface, while the vanadia crystallites formed at higher loadings lead mainly to carbon oxides. Indeed, the isolated active sites favor propene formation and preclude its further oxidation to COx , while in V2 O5 crystallites this site isolation is no longer possible and, therefore, subsequent complete oxidation becomes preponderent [18,24]. Notably, these catalytic performances are comparable with [9,19] or better than [8] those already published for similar VOx /SBA-15 catalysts, especially concerning propene selectivity. This is obviously due to the presence of significant amounts of mono- and oligomeric (VO3− 4 )n species even at high vanadium loadings, as suggested by UV–vis spectroscopy. Higher propane conversions reported in other studies [18,22,29] were obtained in significantly different reaction conditions, i.e. much lower GHSV. The apparent activation energies (Eact ) corresponding to the propane transformation on the VOx /SBA-15 catalysts have been calculated (Table 2) from the Arrhenius plots presented in Fig. S3. They decreased with increasing V content, in line with the observed increase of the catalytic activity. This suggests that the nature of the catalytic site changes with increasing V content, as tetrahedral VOx species transform into square-pyramidal polymeric ones and then into V2 O5 crystallites, in line with the characterization data. Thus, based on the characterization results and the data in Table 2,

tetrahedral mono- and oligomeric (VO3− 4 )n species are less active (lower propane conversion, higher activation energies) but highly selective (higher selectivity to propene), while square-pyramidal polymeric species and even more V2 O5 crystallites are more active (higher propane conversion, lower activation energies) but less selective in ODH reaction of propane. It is noteworthy that the apparent activation energies values listed in Table 2 are in line with literature data for silica-supported V catalysts, ranging from 80 to 160 kJ mol−1 for similar V density on the surface [14]. The effect of the conversion on the selectivities has been studied for 6.5V, 9V and 10V catalysts at 550 ◦ C and a propane-to-oxygen molar ratio equal to 2, by varying the space velocity in the range 1000–2000 ml (gcat min)−1 . As expected, the selectivity to propene decreased in all cases with increasing conversion (Fig. 7). The extrapolation to zero conversion results in 100% selectivity to propene and zero selectivity for COx , showing that propene is the only primary product on SBA-15-supported V catalysts irrespective to the V content, as was proved for other V-based catalysts [1,14]. This clearly suggests that the process takes place in successive steps, the first one being the oxidative dehydrogenation of propane into propene, which is then oxidized to carbon oxides, while the direct combustion of propane does not occur in measurable amounts on the studied catalysts. The same conclusion was also reached by other studies of propane ODH on supported vanadia catalysts [14,30]. The influence of the propane-to-oxygen molar ratio on the catalytic performances was also investigated for the 9V catalyst. The reactions were performed at 550 ◦ C and the space velocity fixed at 1500 ml (gcat min)−1 , while modifying the propane-to-oxygen ratios between 1 and 4. The propane conversion strongly decreased when the propane-to-oxygen molar ratio increased from 1 to 4 (Fig. 8), while the selectivity to propene increased at the expense of carbon oxides. These results could be explained by the decrease of the available oxygen related to the increase in the propane-tooxygen ratio. 4. Conclusion VOx /SBA-15 catalysts with different vanadia loadings were prepared by a modified wet impregnation method, characterized and

Please cite this article in press as: G. Mitran, et al., Propane oxidative dehydrogenation over VOx /SBA-15 catalysts, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.014

G Model

ARTICLE IN PRESS

CATTOD-10504; No. of Pages 8

G. Mitran et al. / Catalysis Today xxx (2016) xxx–xxx

100 6.5V 9V 10V

90

Propene selectivity (%)

80 70 60 50 40 30 20 10

(a)

0 5

0

10

15

20

Propane conversion (%)

100 6.5V 9V 10V

90

COx selectivity (%)

80

their reactivity studied in the propane ODH. A monolayer coverage vanadium density of 1.6 VOx nm−2 was obtained for the sample containing 6.6 wt% V on SBA-15 (6.5V catalyst), the highest reported so far on this support. At low vanadium content tetrahedral monoand oligomeric (VO3− 4 )n species are predominant on the catalyst surface where they coexist with vanadia clusters. As vanadium content increases, square-pyramidal polyvanadate species become predominant and coexist with V2 O5 crystallites, with higher activity but lower selectivity to propene. Propane conversion increases with both temperature and V content, with concomitant decreasing of propene selectivity mainly to the benefit of COx . Increasing the propane-to-oxygen molar ratio from 1 to 4, the propane conversion decreased from 33 to 9%, but the propene selectivity increased from 20 to 69% for the 9V catalyst. Conversion-selectivity plots show that propene is the only primary product at all V loadings. The catalytic performance of the studied catalysts is comparable with or even better than other similar VOx /SBA-15 catalysts, especially concerning propene selectivity. This feature is related to the presence of significant amounts of mono- and oligomeric (VO3− 4 )n species even at high vanadium loadings. Aknowledgements This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

70 60 50

Appendix A. Supplementary data

40

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cattod.2016.12. 014.

30 20 10

References

(b)

0 5

0

10

15

20

Propane conversion (%) Fig. 7. Effect of propane conversion on the propene selectivity (a) and COx selectivity (b) in the propane ODH over VOx /SBA-15 catalysts. Reaction temperature of 550 ◦ C, propane-to-oxygen molar ratio of 2 and space velocity varied from 1000 to 2000 ml (gcat min)−1 .

[1] [2] [3] [4] [5]

[6] [7] [8] [9]

100 Conversion Propene COx

90 Conversion & Selectivities (%)

7

80

[10] [11] [12] [13]

70 60 50 40

[14]

30

[15] [16]

20 10

[17]

0 0

1

2

3

4

5

Propane-to-oxygen molar ratio Fig. 8. Effect of propane-to-oxygen molar ratio on the oxidative dehydrogenation of propane over the 9V catalyst. Reaction temperature of 550 ◦ C, space velocity of 1500 ml (gcat min)−1 .

[18] [19] [20] [21] [22]

H.K. Kung, Adv. Catal. 40 (1994) 1–38. R. Grabowski, Catal. Rev. 48 (2006) 199–268. Z. Nawaz, Rev. Chem. Eng. 31 (2015) 413–436. W. Daniell, A. Ponchel, S. Kuba, F. Andrele, T. Weingand, D.H. Gregory, H. Knözinger, Top. Catal. 30 (2002) 65–74. J. Santamaria-Gonzales, J. Luque-Zambrana, J. Merida-Robles, P. Maireles-Torres, E. Rodriguez-Castellon, A. Jimenez-Lopez, Catal. Lett. 68 (2000) 67–73. H. Dai, A.T. Bell, E. Iglesia, J. Catal. 221 (2004) 491–499. J.M. Lopez Nieto, Top. Catal. 41 (2006) 3–15. P. Gruene, T. Wolfram, K. Pelzer, R. Schlögl, A. Trunschke, Catal. Today 157 (2010) 137–142. ˇ cmanec, H. Sheng-Yang, P. Knotek, L. Capek, ˇ R. Bulánek, P. Ciˇ M. Setniˇcka, Appl. Catal. A: Gen. 415–416 (2012) 29–39. E.V. Kondratenko, M. Cherian, M. Baerns, D. Su, R. Schlögl, X. Wang, I.E. Wachs, J. Catal. 234 (2005) 131–142. P. Ku´strowski, Y. Segura, L. Chmielarz, J. Surman, R. Dziembaj, P. Cool, E.F. Vansant, Catal. Today 114 (2006) 307–313. R.B. Watson, U.S. Ozkan, J. Catal. 191 (2000) 12–29. H. Arakawa, M. Aresta, J.N. Armor, M.A. Barteau, E.J. Beckman, A.T. Bell, J.E. Berkaw, C. Creutz, E. Dinjus, D.A. Dixon, K. Domen, D.L. DuBois, J. Eckert, E. Fujita, D.H. Gibson, W.A. Goddard, D.W. Goodman, J. Keller, G.J. Kubas, H.H. Kung, J.E. Lyons, L.E. Manzer, T.J. Marks, K. Morokuma, K.M. Nicholas, R. Periana, L. Que, J. Rostrup-Nielson, W.M.H. Sachtler, L.D. Schmidt, A. Sen, G.A. Somorjai, P.C. Stair, B.R. Stults, W. Tumas, Chem. Rev. 101 (2001) 953–996. C.A. Carrero, R. Schloegl, I.E. Wachs, R. Schomaecker, ACS Catal. 4 (2014) 3357–3380. J.C. Védrine, I. Fechete, C. R. Chim. 19 (2016) 1203–1225. A. Corma, J.M. Lopez Nieto, N. Paredes, M. Perez, Y. Shen, H. Gao, S.L. Suib, Stud. Surf. Sci. Catal. 72 (1992) 213–220. S.A. Karakoulia, K.S. Triantafyllidis, G. Tsilomelekis, S. Boghosian, A.A. Lemonidou, Catal. Today 141 (2009) 245–253. M. Piumetti, B. Bonelli, P. Massiani, S. Dzwigaj, I. Rossetti, S. Casale, M. Armandi, C. Thomas, E. Garrone, Catal. Today 179 (2012) 140–148. ˇ cmanec, J. Mayerová, Catal. R. Bulánek, A. Kaluˇzová, M. Setniˇcka, A. Zukal, P. Ciˇ Today 179 (2012) 149–158. Y.-M. Liu, W.-L. Feng, T.-C. Li, H.-Y. He, W.-L. Dai, W. Huang, Y. Cao, K.-N. Fan, J. Catal. 239 (2006) 125–136. J. Xu, M. Chen, Y.-M. Liu, Y. Cao, H.-Y. He, K.-N. Fan, Microporous Mesoporous Mater. 118 (2009) 354–360. Y.-M. Liu, Y. Cao, S.-R. Yan, W.-L. Dai, K.-N. Fan, Catal. Lett. 88 (2003) 61–67.

Please cite this article in press as: G. Mitran, et al., Propane oxidative dehydrogenation over VOx /SBA-15 catalysts, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.014

G Model CATTOD-10504; No. of Pages 8 8

ARTICLE IN PRESS G. Mitran et al. / Catalysis Today xxx (2016) xxx–xxx

[23] C. Hess, I.J. Drake, J.D. Hoefelmeyer, T.D. Tilley, A.T. Bell, Catal. Lett. 105 (2005) 1–8. [24] W. Liu, S.Y. Lai, H. Dai, S. Wang, H. Sun, C.T. Au, Catal. Lett. 113 (2007) 147–154. [25] G. Wang, L. Zhang, J. Deng, H. Dai, H. He, C.T. Au, Appl. Catal. A: Gen. 355 (2009) 192–201. ˜ [26] M.A. Banares, X. Gao, J.L.G. Fierro, I.E. Wachs, Stud. Surf. Sci. Catal. 110 (1997) 295–304. [27] G. Du, S. Lim, M. Pinault, C. Wang, F. Fang, L. Pfefferle, G.L. Haller, J. Catal. 253 (2008) 74–90. [28] W. Liu, S.Y. Lai, H. Dai, S. Wang, H. Sun, C.T. Au, Catal. Today 131 (2008) 450–456. [29] F. Ying, J. Li, C. Huang, W. Weng, H. Wan, Catal. Lett. 115 (2007) 137–142. [30] C. Hess, U. Wild, R. Schlögl, Microporous Mesoporous Mater. 95 (2006) 339–349. [31] A. Dinse, S. Khennache, B. Frank, C. Hess, R. Herbert, S. Wrabetz, R. Schlögl, R. Schomäcker, J. Mol. Catal. A: Chem. 307 (2009) 43–50. [32] D. Zhao, Q. Huo, J. Feng, B.F. Chmeka, G.D. Stucky, J. Am. Chem. Soc. 120 (1998) 6024–6036. [33] G. Mitran, T. Cacciaguerra, S. Loridant, D. Tichit, I.-C. Marcu, Appl. Catal. A: Gen. 417–418 (2012) 153–162. [34] A. Galarneau, M. Nader, F. Guenneau, F. Di Renzo, A. Gedeon, J. Phys. Chem. C 111 (2007) 8268–8277.

[35] L. Vradman, M.V. Landau, M. Herskowitz, V. Ezersky, M. Talianker, S. Nikitenko, Y. Koltypin, A. Gedanken, J. Catal. 213 (2003) 163–175. [36] O. Schwarz, B. Frank, C. Hess, R. Schomäcker, Catal. Commun. 9 (2008) 229–233. [37] V.N. Shevchuk, Y.N. Usatenko, P.Y. Demchenko, O.T. Antonyak, R.Y. Serkiz, Chem. Metals Alloys 4 (2011) 67–71. [38] T. Blasco, J.M. Lopez Nieto, Appl. Catal. 157 (1997) 117–142. [39] G.C. Bond, J. Perez Zurita, S. Flamerz, P.J. Gellings, H. Bosh, J.G. Van Ommen, B.J. Kip, Appl. Catal. 22 (1986) 361–378. [40] I.E. Wachs, B.M. Weckhuysen, Appl. Catal. A: Gen. 157 (1997) 67–90. [41] M. Kosmulski, J. Colloid Interface Sci. 337 (2009) 439–448. ˜ [42] M.V. Bosco, M.A. Banares, M.V. Martínez-Huerta, A.L. Bonivardi, S.E. Collins, J. Mol. Catal. A: Chem. 408 (2015) 75–84. [43] F. Arena, F. Frusteri, G. Martra, S. Coluccia, A. Parmaliana, J. Chem. Soc. Faraday Trans. 93 (1997) 3849–3854. [44] H. Launay, S. Loridant, A. Pigamo, J.L. Dubois, J.M.M. Millet, J. Catal. 246 (2007) 390–398. [45] G. Mitran, A. Urda, N. Tanchoux, F. Fajula, I.-C. Marcu, Catal. Lett. 131 (2009) 250–257.

Please cite this article in press as: G. Mitran, et al., Propane oxidative dehydrogenation over VOx /SBA-15 catalysts, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2016.12.014