Vanadium Pentoxide Catalyst over Carbon-Based Nanomaterials for

Article pubs.acs.org/IECR

Vanadium Pentoxide Catalyst over Carbon-Based Nanomaterials for the Oxidative Dehydrogenation of Propane Moslem Fattahi,† Mohammad Kazemeini,*,† Farhad Khorasheh,† and Ali Morad Rashidi‡ †

Department of Chemical and Petroleum Engineering, Sharif University of Technology, Azadi Avenue, Tehran, Iran Nanotechnology Research Center, Research Institute of Petroleum Industry, Tehran, Iran

‡

ABSTRACT: A series of V2O5 catalysts supported on multiwall carbon nanotube (MWCNT), single wall carbon nanotube (SWCNT), and graphene were synthesized by hydrothermal and reflux methods for oxidative dehydrogenation of propane (ODHP) to propylene. The catalysts were characterized by techniques including the BET surface area measurements, XRD, FTIR, H2-TPR, NH3-TPD, FESEM, and UV−vis diffuse reflectance spectroscopy. The performance of the catalysts and the supports were subsequently examined in a fixed bed reactor. The main products were propylene, ethylene and COx. The vanadium catalyst supported on graphene with C/V molar ratio of 1:1 synthesized through the hydrothermal method had the best performance under the reactor test conditions of 450 °C, feed C3H8/air molar ratio of 0.6, and the total feed flow rate of 90 mL/min resulting in average values of 53.6% and 50.7% for propylene selectivity and propane conversion, respectively. This catalyst was further employed in a series of experiments to study the effects of operating parameters including the reaction temperature, propane to air ratio, and the total feed flow rate on conversions and product selectivities using an experimental design method utilizing the response surface methodology (RSM) with central composite design at three levels. The resulting quadratic equations properly correlated the obtained experimental data. Optimum conditions for maximizing propane conversion and propylene selectivity as well as minimizing the COx selectivity were determined at the temperature of 500 °C, C3H8/air molar ratio of 0.28, and total flow rate of 60 mL/min. Ultimately, results under optimum conditions revealed satisfactory agreement between the experimental and predicted data.

1. INTRODUCTION Oxidative conversion of light alkanes to olefins is still a challenging approach not only from a fundamental but also from an applied point of view. The wide availability and low price of light alkanes and the fact that they are generally environmentally nonaggressive products have provided incentives for their use as raw materials in the chemical industry.1,2 Investigations on the catalytic oxidative dehydrogenation of propane (ODHP) were undertaken because this route was expected to lead to lower costs of propylene production as compared with the noncatalytic and nonoxidative processes.3,4 The presence of oxygen limited coking and extended the catalysts lifetimes.1,3,5 A number of promising catalytic systems studied in order to improve the ODHP reaction. Two most prominent catalytic systems studied used the molybdenum and vanadium-based catalysts.2−4 The main challenge was to minimize the formation of carbon oxides (COx) in favor of selective products. Selectivity to olefins was increased while the COx formation suppressed, when N2O utilized as the oxidizing agent.6−10 Main reactions under the ODHP conditions were given by reactions 1−5 C3H8 + 0.5O2 → C3H6 + H 2O

(1)

C3H8 + (2 + 1.5x)O2 → 3COx + 4H 2O

(2)

C3H6 + (1.5 + 1.5x)O2 → 3COx + 3H 2O

(3)

C3H8 + (1 + 1.5x)O2 → C2H4 + COx + 2H 2O

(4)

C3H6 + (0.5 + 0.5x)O2 → C2H4 + COx + H 2O

(5)

© 2013 American Chemical Society

An important and promising area of research in carbon chemistry has been the use of nanostructured carbons, and especially carbon nanotubes (CNTs), as metal-free catalysts. Recent reports showed an outstanding performance of nanoscaled carbon materials in the oxidative dehydrogenation (ODH) of hydrocarbons.11,12 Both the inferior selectivity and the yield of alkenes so far retarded the industrial application of the ODH technology despite outstanding advantages of the oxidative pathway as compared with nonoxidative processes for alkene production.13 CNTs were promising materials for the ODH of ethylbenzene at lower temperatures and without excess supply of steam. Recently it was shown that CNTs with surface modification had a high potential for the activation of lower hydrocarbons.14,15 Nanotubes of transition metal oxides were also interesting nanostructures as they presented a number of oxidation states with different redox-active properties. Vanadium oxide nanotubes (VNT) drew increasing attention as vanadium compounds used in a wide range of catalytic and electrochemical applications.16 Supported vanadia catalysts also utilized extensively for investigating ODH of hydrocarbons. The most commonly used supports for this purpose included the silica, alumina and titania. Recently, zirconia also used as a single or mixed oxide support in this venue. The characteristics that favor zirconia over the conventional supports included a strong interaction with the Received: Revised: Accepted: Published: 16128

March 6, 2013 September 15, 2013 October 22, 2013 October 22, 2013 dx.doi.org/10.1021/ie4007358 | Ind. Eng. Chem. Res. 2013, 52, 16128−16141

Industrial & Engineering Chemistry Research

Article

the copper foils were pretreated by 25% acetic acid. The copper foils were then filtered and washed with deionized water until the filtrate pH of 7 was reached and subsequently dried in a vacuum oven at 90 °C for 2 h. In the CVD process, methane and hydrogen were introduced under a vacuum of 10 mTorr at 1050 °C, leading to formation of the graphene over the copper foils. Graphene was then purified using 50% HCl at 40 °C and then washed by distilled water and filtered. The filter cake was consequently dried in a vacuum oven at 40 °C.25 To eliminate the amorphous carbon, we calcined MWCNT, SWCNT, and graphene samples under an air atmosphere at 400 °C for 2 h and designated them as MWCNT-1, SWCNT-1, and G-1; respectively, for subsequent reactor tests. 2.2. V 2O5 with Template of Carbon Nanostructures. Vanadium pentoxide (V2O5) (supplied by the Aldrich Company) deposited on graphene, MWCNT and SWCNT (all produced inhouse)25,26 in carbon to vanadium (C/V) molar ratio of 1:1 each dissolved in 200 mL of ethanol. The solution temperature was raised to 90 °C under vigorous stirring for 120 min. This solution and 50 mL of water were added to a Teflon-lined autoclave with a stainless steel shell kept at 180 °C for 48h. After quenching to room temperature, the autoclave was heated up again to 180 °C for an additional 24h. The resulting precipitate was filtered and washed to get neutralized. This precipitate was then washed with a solution of ethanol and n-hexane then dried at 100 °C for 24 h to get vanadium oxide over the MWCNT, SWCNT and Graphene. The samples were subsequently calcined at 500 °C for 2 h under nitrogen atmosphere and under air atmosphere at 500 °C for an additional 2 h. The resulting catalysts prepared by the hydrothermal method were designated as VMWCNT-1, V-SWCNT-1, and V-G-1. For catalysts prepared by the reflux method, V2O5, and MWCNT, SWCNT, or graphene, in a carbon to vanadium (C/V) ratio of 1:1, were dissolved in 200 mL of ethanol. The solution temperature was raised to 90 °C under vigorous stirring for 120 min. This solution and 50 mL water were transferred to a round-bottom flask for reflux at 180 °C for 48 h. After cooling to room temperature, the reflux setup was again heated up to 180 °C and kept there for an additional 24 h. The resulting precipitate was filtered, washed, dried, and calcined using the same procedures outlined for the hydrothermal method. The resulting catalysts prepared by the reflux method were designated as VMWCNT-R-1, V-SWCNT-R-1, and V-G-R-1. 2.3. Catalyst Characterization Techniques. The prepared catalysts were characterized by the FESEM, XRD, TPR, FTIR, ASAP and nitrogen porosimetry, as well as UV−vis diffuse reflectance spectroscopy. Field-emission scanning electron microscope (FESEM) images obtained with a Hitachi S-4160 device. Gold was used as a conductive material for sample coating. X-ray diffraction (XRD) measurements were conducted using standard powder diffraction procedure carried out with a STOE Analytical X-ray diffractometer (Cu kα radiation, λ = 1.5406 Å). Data were collected for 2θ between 3 and 120° with a step size of 0.08°. The acidity of the samples measured by the temperature programmed desorption (TPD) of ammonia using BEL-CAT (type A, Japan) instrument with a conventional flow apparatus. A 54 mg of each sample was initially degassed at 500 °C under helium flow rate of 50 mL/min for 50 min at a heating rate of 10 °C/min. The samples were then quenched to 100 °C and saturated with 5% NH3/He for 30 min. The samples were then purged under helium flow for 15 min to remove weakly and physically adsorbed NH3 on the surface of the catalyst and subsequently heated at rate of 10 °C/ min under a 30 mL/min flow of He carrier gas from 100 to 610 °C. The amount of ammonia in the effluent was measured by a thermal conductivity detector (TCD). The TPR experiments were carried out in a BEL-CAT (type A, Japan) instrument coupled with a gas flow controller. The catalysts were heated in a 6.52% (v/v) H2 in Ar from room temperature up to 800◦C at a heating rate of 5 °C/min and flow rate of 150 mL/min. Subsequently, FTIR analyses were performed and recorded on an ABB Bomem MB-100 spectrophotometer. FTIR spectra were obtained from KBr pellets taken from 4000 to 400 cm−1 with a resolution of 4 cm−1. The pore size, textural properties and specific surface area of the catalysts were measured by nitrogen adsorption−desorption at −196 °C using the ASAP apparatus, model

active metal phase resulting in higher dispersion, higher thermal and chemical stability, as well as the unique combination of acid−base and reducing−oxidizing properties. The applications of the CNTs as catalysts attracted much interests in the catalysis community due to their unique porous microstructure and chemical properties. CNTs exhibited excellent thermal conductivity, high surface area, as well as high thermal and chemical stabilities. Furthermore, carbon was also active and selective for catalytic oxidation reactions (e.g.; ODH of ethylbenzene to styrene).17 Most of these reactions were highly exothermic. The energy was released to the catalyst surface causing the catalytic activity or selectivity to decline since the thermal conductivity of the active catalysts and/or the normally used supports was very low. This problem was possibly solved by employing supports with high thermal conductivity, such as CNTs, b-SiC, Si3N4, and BN18 conducting the heat from the catalyst surface to the reactor walls. In a different investigation, vanadium oxide-carbon nanotube composites were prepared by annealing a mixture of CNTs and V2O5 powder in air above the melting point of vanadium oxide.19 Such nanocomposites also synthesized by coating the surface of nitric acid-treated CNTs with vanadic acid.20 However, only few works were reported on the use of CNTs or other carbonaceous materials to catalyze the ODH of light hydrocarbons.21 Liu et al.15 reported the application of the CNTs as catalysts in the ODH of 1-butene to butadiene. The catalytic performance of the CNTs was remarkably high, stable, and superior to that of activated carbon and iron oxide. The characterization of the CNTs before and after reaction was made to test the hypothesis that oxygen functional groups were the active sites for the ODH reaction. Herein, it was demonstrated that, the catalytic activity of the V2O5 over well-nanostructured carbon, such as CNTs and graphene prepared by hydrothermal and reflux methods as well as; the pristine CNTs and graphene in the ODHP reaction. The effects of type of carbon nanostructure (e.g., MWCNT, SWCNT, or graphene) and method of preparation (hydrothermal or reflux) at 450 °C, C3H8/air = 0.6, and total feed flow rate of 90 mL/min were studied in a conventional fixed-bed reactor. Moreover, the effects of reaction conditions including temperature, C3H8/air ratio, and total flow rate were investigated by the RSM to obtain the optimum reaction conditions for the most superior catalyst namely the V2O5 over graphene prepared by the hydrothermal method. The synthesized catalysts were characterized by the BET, XRD, FTIR, H2-TPR, NH3-TPD, FESEM, and UV−vis diffuse reflectance spectroscopy. This investigation provided a simple yet effective catalyst synthesis route toward a better performance for the ODHP reaction.

2. EXPERIMENTAL SECTION 2.1. Catalyst Support Preparation. Carbon nanotubes were prepared by chemical vapor deposition (CVD) of methane over Co− Mo/MgO at 900◦C. The pristine CNTs contained catalytic particles and carbonaceous impurities such as amorphous carbon and carbon nanoparticles. There was a need to eliminate these impurities before the CNTs used as a catalyst support. The CNT samples were purified using a 37wt.% HCl aqueous solution. The resulting mixture was filtered, washed with the deionized water until the pH of the filtrate was neutral and dried at 120 °C.22−24 For further purification, the resulting CNTs were dissolved in a 6 M HNO3 solution and heated at 80 °C for 3 h and subsequently washed and dried at 120 °C for 12 h. Graphene nanosheets were grown on copper foils by catalytic decomposition in a quartz tube system by the CVD method.25 First, 16129

dx.doi.org/10.1021/ie4007358 | Ind. Eng. Chem. Res. 2013, 52, 16128−16141


Article

2010 (Belsorp mini II manufactured by BEL-Japan). The samples were first degassed under vacuum at 200 °C for 12 h. The BET and BJH methods were utilized for measuring the specific surface area, total pore volume and average pore diameter. The UV−visible spectroscopy was performed by the AvaSpec-2048 equipped with a diffusion reflectance accessory. UV−visible spectra measured in the range of 200−800 nm. It is reiterated that, 20−30 mg of the powdered material was necessary to fill up the sample cup. 2.4. Reactor Setup. Catalyst performance tests were carried out in a conventional tubular fixed-bed quartz reactor. A schematic of the experimental setup was shown in Figure 1. The reactor was 10 mm in

selectivity for component i (%) i

∑i

propane conversion (%) =

3 8

× 100

3 8

(7)

(8)

3. CATALYST CHARACTERIZATION RESULTS 3.1. Catalyst Structure. Figure 2a−c illustrated the FESEM micrographs for samples V-MWCNT-1, V-SWCNT1, and V-G-1, respectively. Images a and b in Figure 2 revealed a porous network of entangled CNTs with diameters of about 40 to 120 nm and lengths of up to the micrometer scale. The FESEM image in Figure 2c indicated that the obtained nanostructure was an ordered, well-structured, and homogeneous material. The growth of the V2O5 within the porous CNTs and graphene networks by the hydrothermal method led to the formation of flexible nanomaterials with a dark-green color. In Figure 2, there were not many signs of mineral clusters implying that most of the mineral impurities of the MWCNTs were removed in the purification process. It is noteworthy that, some particles of the vanadium were not either attached properly or grown on the CNTs’ axis implying clusters of vanadium formed alongside of the CNTs. Figure 2(c) confirmed that, the vanadium oxide species were properly precipitated on the surface of graphene and a rather uniform nanostructure of the VOx-graphene formed during the synthesis process. The FESEM images of catalysts synthesized by the reflux method were presented in Figure 3. The catalytic materials were essentially amorphous in nature and nonuniformly dispersed on the support. The nanostructures produced by the reflux method were continuous fibrous granule-like structures with diameters less than 180 nm. On the other hand, the morphology of the catalysts produced by the hydrothermal method was a beltlike, homogeneous, and uniform one. From Figures 2c and 3c, it was observed that the graphenebased catalysts showed more uniformity when they possessed smaller particle sizes. The lengths of the V2O5−CNTs nanorods were up to 80 μm with average width of 90−180 nm (see Figures 2a, b and 3a, b). With graphene as template however, the energy of interaction between graphene and vanadium complexes overcame the thermal activation barrier for aggregation of the V2O5 and led to rapid breakage of the graphene sheet layers causing the barlike structures to form especially for the sample synthesized through the hydrothermal method. It is reiterated that, the CNTs generally acted as geometric templates.19,20 Nonetheless, as seen through images a and b in Figure 2 as well as images a and b in Figure 3, V2O5 nuclei first grew along the axis of the CNTs, and then peeled away from the carbon nanotubes because of weak interactions between them. These observations suggested that the CNTs in some cases did not act as template27 but seemed to serve as catalysts by providing their curved surfaces to induce the nucleation of the V2O5 nanostructures (Figures 2 and 3). It should be reiterated that, the open-ended CNTs were observed through the FESEM images. These CNTs revealed

i

i

i

where i included all the components containing carbon atoms in the exit gas stream, ni was the number of carbon atoms of component i, and Fi was its molar flow rate.

diameter and 70 cm long. It was placed in an electrical furnace equipped with a temperature controller to maintain the reactor temperature to within ±1 °C of the desired set point using thermocouples placed both on the inside and on the external surface of the reactor. The catalyst loading was 0.5g mixed with 0.5g quartz beads for all ODHP experiments. Nitrogen, air and propane utilized were more than 99.9% pure. The reactor was purged at room temperature with nitrogen for 2h. The reactor temperature was then increased to the desired value with a temperature ramp of 5 °C/min. The reaction temperature was in the range of 400−500 °C. Air and propane gas flow rates were controlled by mass flow controllers. The exit gases were analyzed for light hydrocarbons, CO and CO2 using an online Agilent gas chromatograph equipped with TCD and FID detectors. In a typical experiment, product samples were analyzed after the start of the run when stable conditions achieved. Product samples were then analyzed hourly after the start of the run. Propane conversion, product selectivities, and propylene yield were defined as follows

( n3 )[Fi]out − [FC H ]out n ∑i ( 3 )[Fi ]out

( n3 )[Fi]out − [FC H ]out

propane yield (%) propane conversion × selectivity of propylene = 100

Figure 1. Schematic diagram of the experimental apparatus utilized in this research.

∑i

( n3 )[Fi]out

=

× 100 (6) 16130



Article

Figure 2. FESEM micrographs of V2O5 nanostructures prepared by the hydrothermal method (a) V-MWCNT-1, (b) V-SWCNT-1, and (c) V-G-1.

Figure 3. FESEM micrographs of V2O5 nanostructures prepared by the reflux method (a) V-MWCNT-R-1, (b) V-SWCNT-R-1, and (c) V-G-R-1.

that, the catalyst impurities grew upon the CNT’s were adequately removed during the purification process. Furthermore, the presence of particles with a size range below 120 nm on the CNT surface suggested possible chemical interactions between them and the CNTs. Moreover, coexistence of different particle sizes implied that, the CNT’s might have promoted uniform nucleation of the V2O5 nanoparticles on the CNT surfaces. Nonetheless, this interaction was insufficient in regions away from these surfaces. The above observations were consistent with other investigations.28,29 3.2. FTIR Spectra. Figure 4 showed the FTIR spectra of catalysts prepared through the hydrothermal method. The broad IR spectra indicated strong absorptions at 528 to 541 cm−1, related to symmetric and asymmetric stretching

vibrations of V−O−V, and at 915 to 924 cm−1 assigned to the stretching of the short VO bonds from the vanadyl bonds.30 The weak band appearing at 1003 to 1008 cm−1 was due to initial disorder from VO2(B) octahedral arrangement. These spectra suggested that the structure of vanadium layers were rearranged over the carbon nanostructure templates. This was also suggested by the FESEM images indicating nanotubular or belt-like structures for graphene based catalysts (Figure 2c). FTIR Spectra of catalysts synthesized by the reflux method were presented in Figure 5. The 900 to 1050 cm−1 band observed in many vanadium oxide compounds presented intermediate oxidation state between V5+ and V4+. This was attributed to the stretching of short VO bonds also appeared in the B phase of the VO2.30 In Figures 4 and 5, the spectrum of 16131



Article

Figure 4. FTIR spectra of catalysts prepared by the hydrothermal method.

Figure 6. XRD patterns for catalysts prepared by the hydrothermal method.

products were black in color after the hydrothermal process indicating that V5+ reduced to V4+.32 The XRD patterns for catalysts prepared by the reflux method were presented in Figure 7. All the diffraction peaks

Figure 5. FTIR spectra of catalysts prepared by the reflux method.

the as-prepared carbon nanotubes showed the C−C stretching bonds in the range of 1550−1650 cm−1. This was a characteristic of the nanotubes and graphene sheets thru the phonon modes. In samples for which the FTIR spectra determined, the bands with very low intensity at around 3400 cm−1 corresponded to the H2O (or OH− groups) adsorbing on the surface of the samples. All peaks slightly shifted to the right because of the interaction effects of the carbon nanostructures with the vanadium metal species.31 3.3. XRD Patterns. The X-ray diffraction (XRD) patterns for vanadium catalysts over the CNTs and graphene synthesized by the hydrothermal method were presented in Figure 6. The XRD pattern for the V-MWCNT-1 sample indicated that, the growth of the V2O5 was highly crystalline with well-defined (00l) reflections at 2θ angles of 12.8, 17.4, 19.7, 24.1, 28.2, 43.3, and 48.2, consistent with a lamellar structure. The V-SWCNT-1 and graphene composites showed similar reflections but with significantly lower intensities indicating that, the vanadium grew within the composites contained smaller crystalline domains. The XRD patterns represented the VO2(B) nanomaterials. No peaks of any other phases or impurities were observed confirming that, the VO2(B) nanomaterials with high purity synthesized. The peak separation was clearly noticed at 2θ = 17.1, 20.6, 24.7, and 29.6° for the graphene-based catalyst (V-G-1). On the other hand, for the V-MWCNT-1 sample the 2θ in the range of 1−5° emphasized the existence of the mesopore region due to the ends of the CNTs being open and their walls etched. The

Figure 7. XRD patterns for catalysts prepared by the reflux method.

attributed to the orthorhombic structure V2O5 with lattice constants of a = 11.516 Å, b = 3.5656 Å, and c = 4.3727 Å (JCPDS # 41−1426). The reflux method led to the formation of a nanostructure without a change in the oxidation reduction order of V2O5. The V2O5 peak was broad yet small at 2θ = 25.2°. 3.4. Temperature Programmed Desorption (TPD). Acidity of the catalysts was characterized by the temperature programmed desorption using ammonia (NH3-TPD).33,34 Pure carbon nanostructures showed two weak peaks. Moreover, modification of samples increased the acidity of catalysts significantly. The NH3-TPD profiles for catalysts V-G-1 and VSWCNT-R-1 presented in Figure 8. Figure 8a displayed the weak peak around 175 °C attributed to Brønsted acid sites and physisorption of NH3 molecules and the stronger peak above 325 °C corresponded to NH3 from Lewis acid sites or the decomposition of NH4+ species formed over strong Brønsted sites. In Figure 8b, however, the strong peak around 325 °C attributed to the decomposition of NH4+ materials formed over strong Brønsted sites. Compared to SWCNTs support nonetheless, vanadium species showed significant adsorption ability toward the NH3 on acidic sites. The quantitative analyses of acidic sites for the above catalysts were summarized in Table 16132



Article

Figure 9. Hydrogen temperature-programmed reduction (H2-TPR) results for V-G-1, V-SWCNT-R-1, and V-MWCNT-1.

methanation reaction of the CNTs and the second one centered at 380 °C, assigned to reduction of bulk V5+ to V3+.38,39 In the V-G-1 material, the very weak peak observed at the 350 °C related to monomeric, as well as polymeric VO4 units on the surface.40 Incorporation of vanadium to graphene caused the peak of V5+ to shift negatively by about 30 °C. TPR of the vanadium nanostructure prepared with carbon nanobased materials as template demonstrated as follows41,42 (I) At temperature of 686 °C: V2O5 → V6O13 (II) At temperature of 717 °C:V6O13 → V2O4 (III) At temperature of 853 °C: V2O4 → V2O3 3.6. UV−Vis Diffuse Reflectance Spectroscopy. The UV−vis diffuse spectra of V2O5 over carbon nanocomposites presented in Figure 10. The bands in the region around 500 nm

Figure 8. Typical NH3-TPD curves for a() V-SWCNT-R-1 and SWCNT-1 and (b) V-G-1 and G-1.

1. Characterization of the acid sites by NH3-TPD technique indicated that the number of these sites capable of modifying Table 1. NH3-TPD Results for Two of the Prepared Catalysts catalyst

V-G-1

G-1

V-SWCNT-R-1

SWCNT-1

total mmol NH3/g

0.834

0.334

6.136

4.139

the vanadium on the CNT ultimately increased, affecting the propylene yield and hence its cracking to ethylene during the ODHP reaction. 3.5. Temperature Programmed Reduction (TPR). The reduction behavior of the catalysts investigated by the H2-TPR technique and the results for some of the samples presented in Figure 9. The CNTs used as catalyst support in the present study were the further purified version of the material used in the work of Esmaeili et al.35 In the TPR curves of that investigation, two relatively sharp peaks appeared at about 413 and 678 °C corresponding to the residual catalyst impurities in the sample. The purification steps performed on the CNT’s in the present research however, resulted in the removal of such residual impurities because the aforementioned sharp peaks disappeared in the TPR curves of the current study. The strong peak in V-G-1 catalyst centered at around 570 °C might be attributed to methanation of the open carbon nanobased support.36,37 In V-MWCNT-1 and V-SWCNT-R-1 materials, there was a peak in the TPR profile of the VOx-CNTs at around 570 °C that might be deconvoluted into two different peaks; the first one of which, centered at 690 °C, related to the

Figure 10. UV−vis spectra of catalysts prepared by hydrothermal and reflux methods.

for samples prepared by the reflux method attributed to polymeric vanadium (V5+) species with VO4, or possibly isolated vanadium species VO6 (V5+) including physically adsorbed water molecules. The absorption of V5+ cation due to d-d charge transition observed around 400−600 nm confirmed the presence of V2O5 species.43 The spectra of the samples prepared by the reflux method showed an intense shift of optical diffuse shoulders in the absorption edges to the visible region in comparison with the spectra for the samples prepared by the hydrothermal method. The catalysts prepared by reflux methods examined here presented significant peak at 520 nm which, in comparison with hydrothermally prepared samples indicated the presence of three-dimensional V2O5 crystallites. The persistence of a broad ‘‘tail’’ below the 500 nm for the 16133



Article

4. CATALYST EVALUATIONS RESULTS For a comparative study of catalytic performance of the prepared catalysts including V2O5 over MWCNT, SWCNT, and graphene synthesized by the hydrothermal and reflux methods, as well as the bare supports of MWCNT, SWCNT, and graphene, were evaluated in the experimental reactor setup at the ODHP reaction temperature of 450 °C using a feed with C3H8 to air molar ratio of 0.6 and a total feed flow rate of 90 mL/min. Catalysts without calcination in air had much lower activity indicating that the active sites for the ODHP, including carbonyl-like groups were introduced on the CNTs during calcination under air atmosphere. It was expected that the vanadium surface area reductions over the entire range of the pore size distribution to occur after the thermal spreading and calcination treatments. This revealed that the surface of the CNTs and graphene were suggested to be the anchoring locations for the vanadia precursor after calcinations when these particles dispersed upon such supports. It is reiterated that, the presence of oxygen around the active sites (i.e., vanadium and carbon nanostructures) remarkably enhanced the C−H bond activation of propane causing it to exhibit strong site dependence in turn, giving rise to the reactivity of carbon nanostructures toward ODHP. GC analyses of the product gases indicated that along with the unreacted propane, the main reaction products were propylene, ethylene, CO2, and CO (together accounting for more than 99.5% of products) and no organic oxygenates were detected. GC analyses of the products were performed hourly for 6 h from start-of-run. Figure 12

reflux methods indicated that, the additional agglomerated vanadia species compared to those found in other catalyst samples were also detected. The UV spectrum for graphene containing samples showed a very weak peaks at around 350 nm corresponding to the excitation of the π-plasmon of the graphitic structure. 3.7. Surface Area Analysis. Porosimetry analyses of the prepared catalysts presented in Figure 11. Figure 11a showed

Figure 11. (a) Nitrogen sorption isotherms and (b) pore-size distributions for V-G-1 and V-SWCNT-R-1.

nitrogen sorption isotherms of V2O5 over SWCNT and graphene that were similar in shape. Both catalysts displayed the well-known type IV isotherm, which was characteristic of materials with mesoporosity and a high energy of adsorption. Figure 11b illustrated the BJH pore size distribution of the catalysts. The V-G-1 catalyst has a wider pore size distribution than the other catalyst. The SWCNTs have surface area of 480 m2 g−1 and an average pore diameter of 1.2 nm. Graphene has a lower surface area of 345 m2 g−1 and a larger average pore diameter of 12.7 nm. The morphology of the catalyst with graphene was generally less uniform in comparison with that with SWCNT. The V-G-1 and V-SWCNT-R-1 catalysts exhibited surface area of 283 and 198 m2 g−1 and average pore size of 18.2 and 13.7 nm, respectively, suggesting the catalysts were hierarchically porous (Figure 11b). Such a hierarchical structure was essential to ensure a good catalytic performance since the large pore channels allowed rapid species transport, whereas the small ones provided the composites with higher surface areas and more surface active sites.44

Figure 12. Propane conversions at 450 °C, C3H8/air = 0.6, and feed flow rate of 90 mL/min.

presented propane conversions versus time-on-stream for different catalysts indicating a stable trend. Similar stable trends were also obtained for propylene, ethylene and COx selectivities and propylene yield with time-on-stream and their mean values in a 6 h run were reported in Figure 13 for comparative purposes. Dissociative conversion of propane and propylene into ethylene and COx was observed under the reaction conditions. Because of the stable performance of the prepared carbon-based catalysts under an O2 atmosphere at the reaction temperatures employed in this work in addition with an overall check on the carbon balance between the feed and products, it was concluded that CO2 formation during the ODHP reaction was mainly originated from oxidation of the hydrocarbon feedstock and not from burning of the carbon catalysts. 16134



Article

Figure 13. Mean values of propylene, ethylene and COx selectivities and propylene yield for different catalysts at 450 °C, C3H8/air = 0.6, and feed flow rate of 90 mL/min.

Table 2. Mean Values of Propane Conversions and Product Selectivities over Pristine Nanocatalysts at T = 450°C, C3H8/air = 0.6, and Total Feed Flow Rate of 90 mL/min pristine nanocatalyst

propane conversion (%)

propylene selectivity (%)

ethylene selectivity (%)

COx selectivity (%)

propylene yield (%)

MWCNT-1 SWCNT-1 G-1

23.1 24.7 23.9

24.3 34.9 29.5

6.5 10.9 11.3

68.6 53.5 58.6

5.6 8.6 7.1

5. EFFECT OF REACTION CONDITIONS ON PRODUCT YIELDS

Both propane conversion and propylene selectivity were sensitive to the support structure and any impurity or inactive species on the surface may act as a site diluent increasing the selectivity toward dehydrogenation. Carbon support would act as an inactive species thereby improving propylene selectivity. The catalysts prepared by the reflux method had higher propylene yields in comparison with those prepared by the hydrothermal method except for the catalyst prepared from graphene. Combustion of propane and propylene to form COx (reactions 2 and 3) and reforming to ethylene and COx (reactions 4 and 5) are important side reactions that were strongly affected by the type of carbon nanostructure used as the support. Figure 13 indicated that the catalysts based on graphene prepared by both the reflux and the hydrothermal methods had the lowest COx selectivities and the highest ratio of ethylene to COx selectivities in comparison with other catalysts. Mean values of propane conversions and product selectivities during a 6h run over pristine catalysts (i.e., MWCNT-1, SWCNT-1, and G-1) are summarized in Table 2 indicating that the pristine catalysts had much lower propylene yields and propane conversions in comparison with vanadium catalysts over carbon nanostructures. As was the case for the vanadium catalysts, the highest ratio of ethylene to COx selectivities was observed for graphene followed by SWCNT and MWCNT. Among all catalysts, V-G-1 was the most effective catalyst for ODHP with mean values of 53.6 and 50.7% for propylene selectivity and propane conversion, respectively. During the preparation of V-G-1 catalyst, graphene acted as a catalyst for vanadium growth, causing rapid energy release that resulted in the breaking of graphene sheets and subsequent formation of bar-like vanadium structures.

5.1. Data Analysis by Experimental Design. To investigate the effect of operating parameters including the temperature, propane to air ratio, and total feed flow rate on propane conversion, propylene, ethylene, and COx selectivities, we utilized an experimental design approach accounting for the effects of these factors and their interactions on optimizing the process.45 Recently, modeling via the response surface methodology (RSM) based on the experimental design applied to a variety of processes.46,47 In this investigation, Design Experts software (Version 8.0.1) employed for the design of experiments, as well as mathematical modeling and optimization. Central Composite Design (CCD) at three levels with α = 1 (Face Center) used to design such experiments. Moreover, to investigate the performance of the V-G-1 catalyst, the CCD method employed for designing of the experiments. The variables and their corresponding ranges for the investigation of propylene formation in the ODHP reaction were presented in Table 3. The temperature range was chosen between 400 to 500 °C. It was expected that gasification of the CNTs and graphene could become significant at temperatures Table 3. Levels for Process Variables in Actual and Coded Values level

16135

independent variable

−1

0

+1

(A) temperature (°C) (B) propane/air (C) feed flow rate (mL/min)

400 0.2 60

450 0.6 90

500 1.0 120



Article

Table 4. Central Composite Design Experiment Matrix and Experimental Results for This Study variables

run

A: temperature (C)

B: propane/ air ratio

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

500 450 500 450 400 450 400 450 400 400 400 500 450 500 500

0.2 0.6 1 0.6 0.6 0.6 0.2 0.2 1 0.2 1 0.2 1 0.6 1

results C: Feed Flow rate (mL/min)

propane conversion (mol %)

propylene selectivity (mol %)

ethylene selectivity (mol %)

COx selectivity (mol %)

propylene yield (mol %)

60 90 60 120 90 60 120 90 120 60 60 120 90 90 120

59.4 51.8 51.1 50.2 40.4 53.2 45.5 52.9 38.2 46.4 39.0 59.3 50.3 58.7 50.1

58.1 53.6 50.3 52.2 54.4 55.9 51.5 50.3 44.3 52.8 53.1 47.5 52.2 56.9 48.4

15.4 13.0 14.6 12.1 9.6 14.3 12.2 15.9 11.4 8.2 15.5 17.9 16.9 15.8 12.9

26.2 33.3 34.7 35.2 35.8 29.2 35.8 33.3 43.5 38.6 30.9 34.0 30.1 26.9 38.1

34.5 27.8 25.7 26.2 21.9 29.7 23.5 26.6 16.9 24.5 20.7 28.1 26.2 33.4 24.2

Figure 14. Plot of predicted responses versus experimental values for (a) propane conversion, (b) propylene selectivity, (c) ethylene selectivity, and (d) COx Selectivity.

above 530 °C, whereas the catalysts were not active at 300 °C.21 To evaluate the effect of process variables on propane conversion and propylene, ethylene, and COx selectivities in the ODHP reaction, experiments were performed based on the design matrix of the CCD technique with one replicated point. The total number of experiments specified by the RSM software was 15. The design points and the average experimental results (over a 6h run) for propane conversion

and propylene, ethylene, and COx selectivities presented in Table 4. To minimize the effects of uncontrolled factors, experiments were performed in a random sequence. The main products included propylene, ethylene, CO, and CO2 with only trace amounts of methane, ethane, butane, and butylene presented in the products. To predict the curvature of responses, we described propane conversion, propylene, ethylene, and COx selectivities by 16136



Article

Figure 15. (a) Perturbation plot at T = 450 °C, propane/air = 0.6, and feed flow rate = 90 mL/min; (b−d) interaction of process variables on propane conversion at midpoints of variable ranges.

empirical second-order polynomials. The typical form of such polynomials is provided below k

Y = a0 +

k

k

Y3(ethylene selectivity) = 13.62 + 1.86A + 0.47B + 0.16C − 1.66AB

k

− 0.027AC − 1.17BC − 1.65A2 − 2.08B2

∑ aixi + ∑ aiixi 2 + ∑ ∑ aijxixj + ε i=1

i=1

i

j

(9)

(12)

Y4(COx selectivity)

where Y is the predicted response, a0 a constant, ai the ith linear coefficient, aii the ith quadratic coefficient, aij the ith interaction coefficient, xi the independent variable, k the number of factors, and ε the associated error. Quadratic and modified quadratic equations based upon the coded values for the propane conversion and propylene, ethylene, and COx selectivities are presented in eqs 10−13, respectively.

= 31.90 − 2.37A + 0.65B + 2.40C + 1.69AB + 0.34AC + 0.99BC + 2.41C 2 − 2.33ABC

(13)

Statistical criteria indicated that the models were significant for describing the process. Plots of predicted versus actual values for Y1 to Y4 presented in Figure 14 indicating the model predictions were distributed evenly around the 45° line. 5.2. Effect of Process Variables on the Propane Conversion. Perturbation plots showed comparison between all factors at a selected point thru the considered design space. The perturbation plot and dual effect of process variables on propane conversion were demonstrated in Figure 15. The relatively flat line for feed flow rate indicated a fairly insignificant effect of this factor on propane conversion in the design space. Temperature and propane to air ratio, on the other hand, had pronounced effects on propane conversion. The coefficients of eq 10 indicated that the positive influence of the individual operating conditions on the propane conversion ranked first by the temperature followed by the propane to air

Y1(propane conversion) = 52.02 + 6.90A − 3.48B − 0.59C − 0.34AB + 0.056AC − 0.12BC − 2.55A2 − 0.47B2 − 0.37C 2 (10)

Y2(propylene selectivity) = 54.42 + 0.51A − 1.20B − 2.56C − 0.49A2 − 3.9B2 (11) 16137



Article

Figure 16. Perturbation plot at T = 450 °C, propane/air = 0.6, and feed flow rate = 90 mL/min for (a) propylene selectivity and (b) ethylene selectivity.

Figure 17. (a) Contour plot of temperature and propane/air at feed flow rate of 90 mL/min, and (b) the perturbation plot of the temperature, propane/air ratio, and feed flow rate on COx selectivity.

ratio. To examine the interaction between factors on a particular response, we utilized three-dimensional plots. In such plots, two factors were varied while the third one was held constant at the middle of its range. Figure 15b indicated that the influence of temperature on the propane conversion was more pronounced for lower propane to air ratios. The interactions between propane to air ratio and feed flow rate as shown in Figure 15c and temperature, as well as feed flow rate revealed in Figure 15d. This latter figure indicated that the feed flow rate had less of an influence on the propane conversion compared with the other two variables in the design space. 5.3. Effects of Process Variables on Product Selectivities. Figure 16a showed the perturbation plot for propylene selectivity. It was a foregone conclusion from this plot and eq 11 that propane to air ratio and feed flow rate had the most pronounced effects on the propylene selectivity. Propylene selectivities increased with increasing propane to air ratio and went through a maximum near the midpoint of its range as the propane and propylene oxidation competed with the main ODHP reaction. Propylene selectivity increased slightly with

enhancing temperature and the occurrence of the maximum in the propylene selectivity with propane to air ratio was observed over the entire range of reaction temperatures. The perturbation plot for ethylene selectivity was presented in Figure 16b indicating that ethylene selectivity was sensitive to propane to air ratio. Ethylene selectivity was not sensitive to the feed flow rate, whereas temperature had a positive effect on it. At higher temperatures, however the oxy-cracking of both propane and propylene to ethylene became more significant. In the case of propane, two types of radicals, isopropyl and npropyl were formed. Although the secondary carbon−hydrogen bond energy was about 10 kJ/mol lower than that of the primary carbon−hydrogen, these two radicals formed in approximately the same ratio.48 The carbon−hydrogen bond cleavage of the isopropyl radical led to the formation of propylene while the β-scission led to formation of ethylene. The absence of methane in significant amounts suggested that oxy-cracking (and not the thermal cracking) was the dominant cracking mechanism leading to the formation of ethylene and COx. 16138



Article

Table 5. Experimental and Predicted Results of ODHP Process at Optimum Point Predicted by the Software propane conversion (mol %)

optimum conditions

propylene selectivity (mol %)

ethylene selectivity (mol %)

COx selectivity (mol %)

propylene yield (mol %)

optimization approach

T (°C)

C3H8 to air ratio

feed flow rate (mL/min)

Exp.

Pred.

Exp.

Pred.

Exp.

Pred.

Exp.

Pred.

Exp.

Pred.

1 2

500 500

0.49 0.28

60 60

58.5 59.7

57.5 59.2

57.4 58.9

58.1 56.5

12.6 13.9

13.9 -

29.2 26.1

26.2 26.3

33.6 35.2

33.4 33.5

selectivity. Propane to air ratio was found to be a critical parameter for optimization of the ODHP reaction. For the low ratios of propane to air, propane was the limiting reactant, resulting in high propylene selectivity. An increase in the propane to air ratio resulted in an increase in the ethylene and COx selectivities while lowered that of the propylene. On the other hand, the propane conversion only slightly decreased with enhancing the feed flow rate. Moreover, the ethylene selectivity was not very sensitive to the feed flow rate yet the propylene and COx selectivities were significantly affected by this factor. The slight decrease in propane conversion with increasing feed flow rate suggested that, the external diffusional limitations might have been affecting the overall observed kinetics under conditions employed in this study. External diffusion limitations vanished only at high enough feed flow rates. At lower feed rates, where external diffusion of propane might have been significant, propylene selectivities were rather high. Therefore, with increasing feed rate, the propane became more readily available to the catalyst surface and propane oxidation enhanced compared with the main dehydrogenation reaction. These led to lower propylene and higher COx selectivities. Such observations were complementary with those revealed upon investigation of effects of the propane to air ratio over the product selectivities discussed above.

The contour plot showing effects of the temperature and propane to air ratio on the COx selectivity was presented in Figure 17a indicating that at a given propane to air ratio, COx selectivity decreased with enhancing temperature. The perturbation plot in Figure 17b revealed that temperature and feed flow rate had linear effects on the COx production while the propane to air ratio had a nonlinear effect. Besides, lower temperatures favored oxidation reactions leading to a higher COx selectivity. Low propane to air ratios, as well as high temperautres were also the two key factors enhancing the formation of propylene and the oxy-cracking for formation of ethylene.

6. PARAMETER OPTIMIZATION AND MODEL VALIDATION To determine the optimum conditions for the ODHP process, the optimization tool of Design Expert 8.0.1 was utilized. The strategy of the program was to optimize multiple responses for the maximization of desirability functions ranging between zero and one. With all individual desirability functions set to one, the program searched for the conditions at which desirability reached a maximum level. To achieve the optimum conditions, all factors were selected “within the range”. First propane conversion and propylene selectivity were targeted as a maximum and ethylene and COx selectivities as a minimum. In the second approach, propane conversion and propylene selectivity were selected as maximum and only the COx selectivity targeted as a minimum. The optimum process variables were obtained for each case obtaining 0.84 and 0.95 for the desirability values, respectively, and the related determined responses were presented in Table 5. To confirm the correlations developed by the CCD, we performed verification of experiments under the predicted optimum conditions and present the results in Table 5. A fairly good agreement between the predicted and the experimental data was established. The first optimization approach resulted in values for propane conversion, propylene, and COx selectivities slightly lower than those of the experimental ones reported in Table 4. For the second optimization approach with one less constraint than the first approach, however, the results were superior to those reported in Table 4. The superior optimum for the second approach was expected, as the ethylene and COx were produced in a common reaction. In general, the performance of the prepared catalyst for the ODHP reaction in this investigation was superior compared with those reported in the open literature41,42,49 in terms of both the propane conversion and propylene selectivity. Results of the experimental design for the catalyst evaluation indicated that the propane conversions and product selectivities were more influenced by the temperature and propane to air ratio than the feed flow rate. Increasing reaction temperature enhanced both the main dehydrogenation as well as the oxycracking reactions. The latter resulted in enhanced ethylene

7. CONCLUSIONS This research demonstrated a facile hydrothermal and reflux synthesis of V2O5 over carbon-based nanostructures for the ODHP reaction. The bare carbon-based nanocatalysts were also tested and their performance compared with those of the synthesized catalysts. Prepared catalysts characterized through the BET, XRD, FTIR, TPR, TPD, FESEM, and UV−vis diffuse reflectance spectroscopy. These materials produced through the hydrothermal method were more stable and resulted in higher propylene yields in comparison with those prepared by the reflux method. Propylene yields of about 34% at propane conversions of more than 58% were obtained over the vanadium on graphene catalyst prepared by the hydrothermal method. Oxidation and oxy-cracking reactions also competed with the main ODHP reaction led to the formation of ethylene and COx as the side products. Oxidation reactions were more significant at lower temperatures and oxy-cracking reactions were more pronounced at higher temperature values. To investigate effects of the process operating variables including the temperature, propane to air molar ratio (C3H8/Air), and total feed flow rate on the formation of propylene and other side products for the ODHP reaction, we utilized an experimental design method combined with the RSM technique. These were done for the selected catalyst of vanadium over the graphene with the C/V molar ratio of 1:1. Results of catalyst evaluation indicated that the propane conversions and product selectivities were more influenced by temperature and propane to air ratio than by the feed flow rate. 16139



Article

(16) Liu, A.; Ichihara, M.; Honma, I.; Zhou, H. Vanadium-oxide Nanotubes: Synthesis and Template-related Electrochemical Properties. Electrochem. Commun. 2007, 9, 1766. (17) Mestl, G.; Maksimova, N. I.; Keller, N.; Roddatis, V. V.; Schlögl, R. Carbon Nanofilaments in Heterogeneous Catalysis: An Industrial Application for New Carbon Materials? Angew. Chem., Int. Ed. 2001, 40, 2066. (18) Ledoux, M. J; Crouzet, C.; Pham-Huu, C.; Turines, V.; Kourtakis, K.; Mills, P. L; Lerou, J. J. High-Yield Butane to Maleic Anhydride Direct Oxidation on Vanadyl Pyrophosphate Supported on Heat-Conductive Materials: β-SiC, Si3N4, and BN. J. Catal. 2001, 203, 495. (19) Ajayan, P. M.; Stephan, O.; Redlich, Ph.; Colliex, C. Carbon Nanotubes as Removable Templates for Metal Oxide Nanocomposites and Nanostructures. Nature 1995, 375, 564. (20) Satishkumar, B. C.; Govindaraj, A.; Nath, M.; Rao, C. N. R. Synthesis of Metal Oxide Nanorods Using Carbonnanotubes as Templates. J. Mater. Chem. 2000, 10, 2115. (21) Sui, Z.-j.; Zhou, J.-h.; Dai, Y.-c.; Yuan, W.-k. Oxidative Dehydrogenation of Propane over Catalysts Based on Carbon Nanofibers. Catal. Today 2005, 106, 90. (22) Arasteh, R.; Masoumi, M.; Rashidi, A. M.; Moradi, L.; Samimi, V.; Mostafavi, S. T. Adsorption of 2-Nitrophenol by Multi-wall Carbon Nanotubes from Aqueous Solutions. Appl. Surf. Sci. 2010, 256, 4447. (23) Rashidi, A. M.; Akbarnejad, M. M.; Khodadadi, A. A.; Mortazavi, Y.; Ahmadpourd, A. Single-wall Carbon Nanotubes Synthesized Using Organic Additives to Co-Mo Catalysts Supported on Nanoporous MgO. Nanotechnology 2007, 18, 315605 (5pp). (24) Rashidi, A.; Lotfi, R.; Fakhrmosavi, E.; Zare, M. Production of Single-walled Carbon Nanotubes from Methane over Co-Mo/MgO Nanocatalyst: A Comparative Study of Fixed and Fluidized Bed Reactors. J. Nat. Gas Chem. 2011, 20, 372. (25) Ghozatloo, A.; Shariaty-Niasar, M.; Rashidi, A. M. Preparation of Nanofluids from Functionalized Graphene by New Alkaline Method and Study on the Thermal Conductivity and Stability. Int. Commun. Heat Mass Transfer 2013, 42, 89. (26) Sadeghian, Z.; Rashidi, A. M. Synthesis Optimisation and Characterisation of Multiwalled Carbon Nanotubes Produced by Spray Pyrolysis of Hexane. Mater. Sci. Technol. 2010, 26, 1191. (27) Chen, X.-W.; Zhu, Z.; Havecker, M.; Su, D. S.; Schlogl, R. Carbon Nanotube-induced Preparation of Vanadium Oxide Nanorods: Application as a Catalyst for the Partial Oxidation of n-Butane. Mater. Res. Bull. 2007, 42, 354. (28) Tang, S.; Cao, Z. Site-dependent Catalytic Activity of Graphene Oxides Towards Oxidative Dehydrogenation of Propane. Phys. Chem. Chem. Phys. 2012, 14, 16558. (29) Cao, Z.; Wei, B. V2O5/single-walled Carbon Nanotube Hybrid Mesoporous Films as Cathodes with High-rate Capacities for Rechargeable Lithium Ion Batteries. Nano Energy 2013, 2, 481. (30) Valmalette, J.-C.; Gavarri, J.-R. High Efficiency Thermochromic VO2(R) Resulting from the Irreversible Transformation of VO2(B). Mater. Sci. Eng., B 1998, 54, 168. (31) Huang, B.; Huang, R.; Jin, D.; Ye, D. Low temperature SCR of NO with NH3 over Carbon Nanotubes Supported Vanadium oxides. Catal. Today 2007, 126, 279. (32) Reddy, Ch.V. S.; Walker, E. H., Jr.; Wicker, S. A., Sr.; Williams, Q. L.; Kalluru, R. R. Synthesis of VO2 (B) Nanorods for Li Battery Application. Curr. Appl. Phys. 2009, 9, 1195. (33) Boehm, H. P. Surface Oxides on Carbon and Their Analysis: A Critical Assessment. Carbon 2002, 40, 145. (34) Figueiredo, J. L; Pereira, M. F. R; Freitas, M. M. A; Ó rfão, J. J. M. Modification of the Surface Chemistry of Activated Carbons. Carbon 1999, 37, 1379. (35) Esmaeili, E.; Mortazavi, Y.; Khodadadi, A. A.; Rashidi, A. M.; Rashidzadeh, M. The Role of Tin-promoted Pd/MWNTs Via the Management of Carbonaceous Species in Selective Hydrogenation of High Concentration Acetylene. Appl. Surf. Sci. 2012, 263, 513.

Moreover, an increase in the molar propane to air ratio resulted in an enhancement of the ethylene and COx selectivities while lowered the propylene selectivity. Optimal conditions (T = 500 °C, propane/air = 0.28, and feed flow rate = 60 mL/min) to maximize propane conversion and propylene selectivity as well as to minimize the COx selectivity were determined. Ultimately, the performance of the aforementioned catalyst prepared for the ODHP reaction undertaken in the present study were demonstrated to be superior to those reported in the literature.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel.: +98 21 6616 5425. Fax: +98 21 6602 2853. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Cavani, F.; Trifiro, F. The Oxidative Dehydrogenation of Ethane and Propane as an Alternative Way for the Production of Light Olefins. Catal. Today 1995, 24, 307. (2) Madeira, L. M.; Portela, M. F. Catalytic Oxidative Dehydrogenation of n-Butane. Cat. Rev. - Sci. Eng. 2002, 44, 247. (3) Bettahar, M. M.; Costeintin, G.; Savary, L.; Lavalley, J. C. On the Partial Oxidation of Propane and Propylene on Mixed Metal Oxide Catalysts. Appl. Catal. A 1996, 145, 1. (4) Mamedov, E. A.; Corberan, V. C. Oxidative Dehydrogenation of Lower Alkanes on Vanadium Oxide-Based Catalysts. The Present State of the Art and Outlooks. Appl. Catal. A 1995, 127, 1. (5) Fuchs, S.; Leveles, L.; Seshan, K.; Lefferts, L.; Lemonidou, A.; Lercher, J. A. Oxidative Dehydrogenation and Cracking of Ethane and Propane over LiDyMg Mixed Oxides. Top. Catal. 2001, 15, 169. (6) Pérez-Ramírez, J.; Kondratenko, E. V. Steam-activated FeMFI Zeolites as Highly Efficient Catalysts for Propane and N2O Valorisation Via Oxidative Conversions. Chem. Commun. 2003, 17, 2152. (7) Dury, F.; Centeno, M. A.; Gaigneaux, E. M.; Ruiz, P. An Attempt to Explain the Role of CO2 and N2O as Gas Dopes in the Feed in the Oxidative Dehydrogenation of Propane. Catal. Today 2003, 81, 95. (8) Kondratenko, E. V.; Cherian, M.; Baerns, M.; Su, D.; Schlögl, R.; Wang, X.; Wachs, I. E. Oxidative Dehydrogenation of Propane over V/ MCM-41 Catalysts: Comparison of O2 and N2O as Oxidants. J. Catal. 2005, 234, 131. (9) Bulánek, R.; Wichterlová, B.; Novoveská, K.; Kreibich, V. Oxidation of Propane with Oxygen and/or Nitrous Oxide over FeZSM-5 with Low Iron Concentrations. Appl. Catal. A 2004, 264, 13. (10) Ovsitser, O.; Cherian, M.; Kondratenko, E. V. In-Situ UV/vis and Transient Isotopic Analysis of the Role of Oxidizing Agent in the Oxidative Dehydrogenation of Propane over Silica-Supported Vanadia Catalysts. J. Phys. Chem. C 2007, 111, 8594. (11) Zhang, J.; Liu, X.; Blume, R.; Zhang, A.; Schlögl, R.; Su, D. S. Surface-Modified Carbon Nanotubes Catalyze Oxidative Dehydrogenation of n-Butane. Science 2008, 322, 73. (12) Su, D. S.; Maksimova, N.; Delgado, J. J.; Keller, N.; Mestl, G.; Ledoux, M. J.; Schlögl, R. Nanocarbons in Selective Oxidative Dehydrogenation Reaction. Catal. Today 2005, 102−103, 110. (13) Cavani, F.; Ballarini, N.; Cericola, A. Oxidative Dehydrogenation of Ethane and Propane: How far from Commercial Implementation? Catal. Today 2007, 127, 113. (14) Zhang, J.; Su, D.; Zhang, A.; Wang, D.; Schlögl, R.; Hébert, C. Nanocarbon as Robust Catalyst: Mechanistic Insight into CarbonMediated Catalysis. Angew. Chem., Int. Ed. 2007, 119, 7460. (15) Liu, X.; Su, D. S.; Schlögl, R. Oxidative Dehydrogenation of 1Butene to Butadiene over Carbon Nanotube Catalysts. Carbon 2008, 46, 547. 16140



Article

(36) Yang, X.; Wang, X.; Qiu, J. Aerobic Oxidation of Alcohols over Carbon Nanotube-Supported Ru Catalysts Assembled at the Interfaces of Emulsion Droplets. Appl. Catal. A 2010, 382, 131. (37) Bahome, M. C.; Jewell, L. L.; Padayachy, K.; Hildebrandt, D.; Glasser, D.; Datye, A. K.; Coville, N. J. Fe-Ru Small Particle Bimetallic Catalysts Supported on Carbon Nanotubes for Use in Fischer− Tröpsch synthesis. Appl. Catal. A 2007, 328, 243. (38) Caero, L. C.; Jorge, F.; Navarro, A.; Gutiérrez-Alejandre, A. Oxidative Desulfurization of Synthetic Diesel Using Supported Catalysts: Part II. Effect of Oxidant and Nitrogen-compounds on Extraction-oxidation Process. Catal. Today 2006, 116, 562. (39) Taufiq-Yap, Y. H.; Wong, Y. C.; Zainal, Z.; Hussein, M. Z. Synthesis of Self-assembled Nanorod Vanadium Oxide Bundles by Sonochemical Treatment. J. Nat. Gas Chem. 2009, 18, 312. (40) Gómez-Bernal, H.; Cedeño-Caero, L.; Gutiérrez-Alejandre, A. Liquid Phase Oxidation of Dibenzothiophene with Alumina-Supported Vanadium Oxide Catalysts: An Alternative to Deep Desulfurization of Diesel. Catal. Today 2009, 142, 227. (41) Koranne, M. M.; Goodwin, J. G.; Marcelin, G. Characterization of Silica- and Alumina-Supported Vanadia Catalysts Using Temperature Programmed Reduction. J. Catal. 1994, 148, 369. (42) Arena, F.; Frusteri, F.; Parmaliana, A. Structure and Dispersion of Supported-Vanadia Catalysts. Influence of the Oxide Carrier. Appl. Catal. A 1999, 176, 189. (43) Higashimoto, S.; Tanihata, W.; Nakagawa, Y.; Azuma, M.; Ohue, H.; Sakata, Y. Effective Photocatalytic Decomposition of VOC under Visible-light Irradiation on N-doped TiO2 Modified by Vanadium Species. Appl. Catal. A 2008, 340, 98. (44) Chen, Z.; Qin, Y.; Weng, D.; Xiao, Q.; Peng, Y.; Wang, X.; Li, H.; Wei, F.; Lu, Y. Design and Synthesis of Hierarchical Nanowire Composites for Electrochemical Energy Storage. Adv. Funct. Mater. 2009, 19, 3420. (45) Montgomery, D. C. Design and Analysis of Experiments, 4th Ed.; John Wiley & Sons: New York, 1996. (46) Tabandeh, F.; Khodabandeh, M.; Yakhchali, B.; Habib-Ghomi, H.; Shariati, P. Response Surface Methodology for Optimizing the Induction Conditions of Recombinant Interfere on Beta During High Cell Density Culture. Chem. Eng. Sci. 2008, 63, 2477. (47) Hosseinpour, V.; Kazemeini, M.; Mohammadrezaee, A. Optimisation of Ru-promoted Ir-Catalysed Methanol Carbonylation Utilising Response Surface Methodology. Appl. Catal. A 2011, 394, 166. (48) Perrin, D.; Martin, R. The Hetero-homogeneous Pyrolysis of Propane, in the Presence or in the Absence of Dihydrogen, and the Measurement of Uptake Coefficients of Hydrogen Atoms. Int. J. Chem. Kinet. 2000, 32, 340. (49) Sediri, F.; Gharbi, N.; Nanorod, B. Phase VO2 Obtained by Using Benzylamine as a Reducing Agent. Mater. Sci. Eng., B 2007, 139, 114.

16141


Vanadium Pentoxide Catalyst over Carbon-Based Nanomaterials for

Recommend Documents